Accenture_Technology_Vision_2014

Accenture Technology Vision 2014Every Business Is a Digital BusinessFrom DigitallyDisrupted toDigital Disrupter

CONTENTS2FOREWORD 3INTRODUCTION 4TRENDS01Digital–physical blur:Extending intelligence to the edge 1202From workforce to crowdsource:The rise of the borderless enterprise 2803Data supply chain:Putting information into circulation 4204Harnessing hyperscale:Hardware is back (and never reallywent away) 5605The business of applications:Software as a core competencyin a digital world 7206Architecting resilience:“Built to survive failure” becomes themantra of the nonstop business 84CONCLUSION 98RESEARCH METHODOLOGY 102END NOTES 104CONTACTS 111#techvision2014

FOREWORDForewordWhen we declared “every business is a digital business”in the Accenture Technology Vision 2013, we didn’t seeit as a trend for last year or this year. We saw it as thefuture. The future of technology. The future of business.The future of our increasingly interconnected andinterdependent world.Enterprises are embracing technology in the way they dobusiness and also as a catalyst to create something new—new markets, new products, and new areas of growthLast year, we saw the beginnings of business transformationand revenues.based on a digital model. Organizations looking toreimagine themselves in a technology-driven world setforth on their journey to becoming digital businesses.Many organizations were experimenting, while otherswere making larger investments. But all were countingon technology to fuel their next waves of growth.The change is revolutionary. Industrial companies arebecoming customer service companies. Consumerproducts companies are becoming Internet companies.Energy companies are becoming information companies.And media and entertainment companies are becominglogistics companies.Pierre NantermeChairman & CEO3This year, we see a marked uptick in digital. The AccentureTechnology Vision 2014 lays out bold trends that arebecoming characteristic of larger enterprises, which havebeen perceived by some as lagging in converting to digitalbusinesses. While social, mobile, analytics, and cloud stilldrive these trends, the focus now is on new ways thatthese technologies are being woven into the nextFor our clients and for any organization, the AccentureTechnology Vision 2014 points toward an exciting timeof new opportunities driven by the power of technology.We hope that you find the Accenture Technology Visioninsightful as you continue on the journey to become adigital business.Paul DaughertyChief Technology Officergeneration of business strategies across every industry.ACCENTURE TECHNOLOGY VISION 2014

INTRODUCTION4#techvision2014

INTRODUCTIONINTRODUCTIONEvery Business Is a Digital Business:From Digitally Disrupted to DigitalDisrupterBig companies are back in the digital game. Procter & Gamble,Tesco, Disney, GE—these are just a few of the global 2,000 thatare now in a race to become digital. Those that get there first willbe able to disrupt their existing markets and penetrate new ones.They will be in control of their new digital destinies.5This year’s report represents the latest iteration of Accenture’sTechnology Vision, which declares that “every business is adigital business.” In last year’s report, we laid out the imperativefor every business to reimagine itself in this digital world, andwe charted the increasing appetite of leading enterprises forexploring the opportunities that emerging technologies provide.ACCENTURE TECHNOLOGY VISION 2014

INTRODUCTION6Big is the next big thingThis year we see something more. We see the first waveof large traditional companies starting to use technologyas a driving force—in some situations, the driving force—for how they grow strongly and effectively. If the lastdecade has been seen as the playground of the digitalstartups—overnight sensations such as Instagram andTwitter, Zipcar and YouTube, TripAdvisor and Airbnb—then the coming decade will see the emergence of thetraditional companies as digital giants. Backed by theirdeep resources, enormous scale, and process discipline,these new digerati are about to rewrite much of thedigital playbook.For business leaders everywhere, the next three yearswill be about determining their organizations’ pace inthis digital race—and their place in the new world ofdigital. Large companies are making a push to transformthemselves from followers to leaders in digital. The firstcontenders out of the gate are already poised to takeadvantage of the many recent technology advances inways that promise to upend expectations of industryobservers and consumers alike.Tesco, GE, and Disneyshow howConsider Tesco. In the last two years, the globalsupermarket chain not only has created interactivegrocery stores in airports and subway stations but alsohas expanded into new industries. Tesco now offersmovie streaming, e-books, and even its own seveninchtablet priced as low as $91. 1 The grocery retaileris well on its way to becoming a truly digital business.More than 20 percent of Tesco’s online sales now comethrough smartphones, and 10 percent of all orders fromTesco Direct come through its mobile website. 2In the manufacturing arena, GE is betting on theindustrial Internet, building cloud-based services withintelligent analytics so that it can collect and combinevast amounts of industrial-machine data and equipmentdata, extracting unique insights that it can use to setnew performance standards in major industries such asenergy and aviation.#techvision2014

INTRODUCTIONIn the business-to-consumer realm, Disney is introducinga collection of tools including a wireless trackingwristband to create an entirely new personalized andenriched experience for visitors to its amusement parks.For visitors wearing the MagicBand, Disney offers abetter guest experience with much less waiting. But thetechnology goes beyond just giving preferential placesin queues: for example, it allows Disney park-goers tosimply “touch to pay” for food and merchandise and tomake and share plans with family and friends in a travelgroup. The technology is also transforming the diningexperience, enabling patrons to preorder food, which isthen prepared and served as they walk into a restaurant. 3And of course, Disney captures a comprehensive digitalrecord (or storybook) of its customers’ activities.Much more than aninvestor relations gameDuring their years of experimentation, industrialpowerhouses such as GE, Disney, P&G, Tesco, Walmart,and Shell have learned by doing—steadily gainingthe skills and the competencies to pull ahead of theircompetitors, blunting the startups’ advantages, andproving to their stakeholders that they absolutely havewhat it takes to excel in the new digital economy.Of course, the big players know it’s far more than aninvestor relations game. They are acutely aware thatdigital expertise can confer exceptional strategicadvantage. They realize that not only can they controltheir own markets, but they can actually disrupt andestablish footholds in others, often creating new businessand market models in the process. For example, GeneralMotors is already raising anxiety in the car rentalbusiness by partnering with RelayRides, a startup whoseuse of mobile phones makes ride sharing easy. And AT&Tis entering the home security market with its Digital Lifeservice, allowing customers to remotely control everythingin the home, from alarms to lights to door locks.7ACCENTURE TECHNOLOGY VISION 2014

INTRODUCTIONWhen boundaries blur8A huge shift is under way. Boundaries are blurring inmany dimensions—not just between IT leaders and theirbusiness colleagues, but between digital assets andphysical resources and between enterprises and theircustomers, suppliers, and other stakeholders. IT strategyand business strategy are no longer separate; they havebecome inseparable. It’s this kind of mindset that we seeacross the digital spectrum.In many ways, digital is unlocking a renaissance for largecompanies, allowing them to reconnect with what madethem industry leaders in the first place. There is a vastvariety of opportunity before them. Just as technologyleveled the field for the newcomers of the last decade, sothe incumbents can now leverage those same forces ofdigital democratization to charge back. Lower barriers toentry for one are lower barriers to entry for all. Many willuse their new digital prowess as a potent differentiator—resetting the bar for consumer interactions, for instance,or providing unprecedented levels of supply chaineffectiveness, or perhaps developing groundbreaking newpricing models.Hallmarks of tomorrow’sdigital disruptersThe concept of the “fast follower” is becoming obsolete.In a world of nonstop change, there is no time to catchup. Instead, we expect the emerging digital leaders tobe master orchestrators—uniquely able to strive forconvergence on a huge scale as they capitalize on a widearray of technology breakthroughs to rapidly advanceand innovate with their systems and strategies.In turn, this new pace will force corporate leaders tothink about new operating models that might yieldfurther advantage. Here too, many experiments are underway—from new cross-functional roles for data scientiststo chief digital officers whose influence extends beyondthe four walls of the enterprise.#techvision2014

INTRODUCTIONFIGURE 1:Every Business Isa Digital Business:The Evolution–10-#techvision2014

INTRODUCTIONAnd throughout every organization, we anticipate are-estimation of the skills that will be most valuable,at all levels.This year’s Technology Vision report highlights sixemerging themes that reflect the shifts being seen nowamong the digital power brokers of tomorrow: largeenterprises such as Tesco, GE, and Disney that recognize,as they leverage their scale to redefine digital in theirindustries, that big is the next big thing.Collectively, these themes represent the newestexpression of Accenture’s stance that “every business isa digital business.” They provide additional componentsto Accenture’s multiyear perspective on technology’stectonic shifts and its impact on the strategies andoperational priorities for organizations worldwide(see Figure 1).Individually, each theme, from each year, highlights theevolution of a key technology, some of which are alreadycentral to the digital explorations of many leadingenterprises. Viewed in aggregate, the themes representa fundamental shift in the assumptions that companiesnow must make as they plan for success in the decadesto come. They provide a richly detailed view from whichbusiness leaders in every industry can draw insight,inspiration, and, we hope, excitement about where digitaltechnologies can take their organizations tomorrow.Becoming a digital business is no longer simply abouthow we incorporate technology into our organizations;it’s about how we use technology to reinvent thoseorganizations to get out in front of the dramaticchanges that technology is creating. For large enterprisesespecially, the opportunity to shift from disrupted todisrupter cannot be overstated. The question thesebusinesses must now ask is how they will use the nextthree years to redefine their places in this new world.11ACCENTURE TECHNOLOGY VISION 2014

TREND 1: DIGITAL–PHYSICAL BLUR12#techvision2014

TREND 1: DIGITAL–PHYSICAL BLURTREND 1Digital–physical blur:Extending intelligence to the edgeThe physical world is coming online as objects, devices, andmachines acquire more digital intelligence. What’s emergingis more than just an “Internet of Things”; it’s a new layer ofconnected intelligence that augments the actions of individuals,automates processes, and incorporates digitally empoweredmachines into our lives, increasing our insight into and controlover the tangible world. There are benefits for consumers andbusinesses. Consumers become better informed and betterequipped to influence the ways they experience everythingaround them. And businesses get real-time connections to thephysical world that allow machines as well as employees to actand react faster—and more intelligently.13ACCENTURE TECHNOLOGY VISION 2014

TREND 1: DIGITAL–PHYSICAL BLUR14#techvision2014Why now?Explosion of connected devices: The installed base ofthe Internet of Things is estimated to reach approximately212 billion in 2020. This will include 30 billion “connected(autonomous) things” that same year. 1Increased bandwidth: Global IP traffic is expected to nearlydouble between 2013 and 2016, and broadband is expected tospeed up more than twofold. 2Advanced robotics: From agriculture to oil fields, advancesin robotics are empowering human-robot collaborationin industries beyond the factory floor. Several leading carmanufacturers have committed to bringing autonomous cartechnologies to market by 2020.Rise of real-time analytics: Data sources are growing at anunprecedented velocity, and the ability to loop insightsimmediately back into the decision process is supportingautomating responsive actions like never before. By 2017,more than 50 percent of analytics implementations will makeuse of event data streams generated from instrumentedmachines, applications, and/or individuals. 3Wearing a smartwatch yet? Tracking yourworkouts with your phone? Noticed adriverless car in your neighborhood?From wearable computers to autonomous drones, theways in which we experience the world are changing fast.Intelligent interfaces are emerging that allow decisions tobe made “on the edge”—at the point where digital andphysical worlds converge—rather than in a centralizedmanner. These decisions can be made exactly when they’reneeded in informed, social, easy-to-use ways, allowingcompanies and governments to reimagine the possibilitiesfor engaging with their customers and citizens.Smartphones have turned their owners into digitallyaugmented versions of themselves—able to catalog andquantify actions throughout the day and access, create,and share an astonishing array of pertinent informationthat can enable faster, better decisions. Several carmodels can now self-park with ease—making “edge”decisions about available space, proximity to the curb,and more. Google and Nissan claim their driverless carsare just five to six years away. 4 And autonomous drones—once the sole province of the military—are being used bypolice precincts across the United States.

TREND 1: DIGITAL–PHYSICAL BLURBut what does the average smartphone subscriber havein common with the person riding in a driverless car?They are both using “edge devices”—devices whose edgesborder other devices in that they have unprecedentedcapabilities to connect with other devices. These“cyber physical” systems sense their environments andrespond—appropriately—in real time, making possiblebetter-informed decisions within windows of opportunitythat can create competitive advantage. The pivotalpoint is that the sheer quantity of these edge devices isincreasing as dramatically as their prices are dropping.At the same time, their ability to sense environmentalvariables, share data with other edge devices via thecloud, and have deep analytics performed by the cloudadds up to create rich user experiences that inform muchmore intelligent, real-time decision-making.As the line between the digital and the physicalcontinues to blur, a vast new window of opportunity isopening for the enterprise. By leveraging and enhancingtheir physical assets, traditional companies are looking atthis opportunity as a way to leapfrog online competitors,create immersive real-world experiences for consumers,and gain market share. In addition, every company nowhas the opportunity not just to gather insight to makesmart business decisions but to turn those decisions intoactions, in real time, in the real world. The enormousexpansion in intelligent capabilities is rapidly reshapingestablished operations, paving the way for industrydisruption on a massive scale.But to chase these new opportunities, business andtechnology leaders alike must rethink how they bothengage customers and run their businesses in a digitalphysicalworld. For industrial Internet leaders like GE, thismeans pushing sensing technologies outside of industrialapplications and further to the edge of operations.For companies such as Cisco, which predicts that theindustrial Internet market will be worth $14.4 trillion overthe next decade, this means focusing on the “Internet ofEverything”—or, as Cisco describes it, “bringing togetherpeople, process, data, and things to make networkedconnections more relevant and valuable than ever before.” 515ACCENTURE TECHNOLOGY VISION 2014

TREND 1: DIGITAL–PHYSICAL BLUR16A world of new userexperiencesThe way people interact with the world around them ischanging. Digital technologies offer new decision-makingexperiences—from selecting a restaurant in a newneighborhood to making a critical maintenance decisionon a gas pipeline. The power of these decision spaces isthat they give users real insights, not just information;by providing valuable insight, users are one step closer totaking action.Technology has been evolving to enable this for the lastdecade. The ubiquity of network connectivity and theproliferation of smart devices (such as sensors, signs,phones, tablets, lights, and drones) have created platformsupon which every enterprise can innovate. In terms ofcreating new consumer experiences, this is perhaps mostvisible in the surging popularity of the “quantified self”movement. Consumer wearables such as Nike’s FuelBand,Adidas’s miCoach, and Fitbit track exercise and physicalactivity in ways that allow users to easily gain insightinto their performance—often in real time—giving themthe information necessary to make decisions aboutpicking up the pace, going for another lap, or pushing fora personal best.In the urban context, the Copenhagen Wheel, areplacement bicycle wheel, is able to augment the cyclingexperience by sensing the pressure and effort exertedon the bike’s pedals and adding motorized help when therider needs it; in doing so, it’s able to quantify physicalactivity for the rider.These quantified measurements are not isolated toathletes or individuals. Osakidetza, the public healthsystem in the Basque region of Spain, is using MicrosoftKinect devices to enable telemedicine treatment ofchronic patients. Using the Kinect devices, patients aregiven not only a more natural way to interface withtechnology but also new ways to experience medicalcare. Using insights from the Kinect devices, physicaltherapists get access to a wealth of granular datathat can be used to offer remote consultations and toquantifiably gauge progress. In doing so, therapy sessionscan be more frequent, more targeted, and shorter, thusreducing costs, improving outcomes, and reducingpatients’ recovery times. 6#techvision2014

TREND 1: DIGITAL–PHYSICAL BLURMore and more, these experiences are expanding beyonda user’s personal set of devices. Embedded intelligenceisn’t limited to smartphones and tablets. Everyday objectsare becoming smart and expanding the definition of whatit means to experience the real world. In London, BritishAirways has unveiled two digital billboards that activelytrack flights in real time, animate an image of a childpointing to the plane, and display the flight number andinformation about its destination or origin—inspiringonlookers to dream about where they could be escapingto on a British Airways flight. 7 In San Francisco, parkingmeters are smart and connected; drivers see a colorcodedmap denoting varying levels of parking difficulty.Once a car is parked, the system can alert the driverwhen his meter time has nearly expired and allow him to“feed the meter” virtually, from his smartphone. Thetechnology doesn’t allow drivers to reserve a parkingspot, but it does give them an interface to make adecision on the edge. 8Similarly, Waze, a mobile phone app, allows drivers tomake decisions at the point of action. It does this byenabling users to use their phones to share insightson current road conditions and offering suggestionsabout alternate routes around traffic obstacles. In short,Waze users share rich, unstructured insights that createactionable decision spaces.These decision spaces are not the exclusive domainof digital companies. Brick-and-mortar retailerscan differentiate themselves from their online-onlycounterparts by mastering the digital–physical blur.Retailers such as Tesco, Neiman Marcus, and Staplesare continually experimenting with new ways to deliverunique and meaningful consumer experiences. Tesco andStaples are transforming their in-store technology andservice offerings to better align with consumer lifestyles:Tesco is rolling out face scanning digital signage at all450 of its UK petrol stations to tailor engaging andon-screen content to the audience of five millionplusadults who pass through its stations each week;Staples is piloting stores with less merchandise, morekiosks (with free next-day delivery), and meeting spacesfor busy small-business owners. 9 For its part, NeimanMarcus has piloted digital-physical solutions that provideboth sales associates and customers with actionableintelligence—in the form of store events, product arrivals,17ACCENTURE TECHNOLOGY VISION 2014

TREND 1: DIGITAL–PHYSICAL BLUR18and even personal touches such as the knowledge ofwhen a favorite sales associate is working. 10 With Apple’sintroduction of iBeacon indoor positioning technologyand competitors such as Estimote, brick-and-mortarretailers will find abundant opportunities to incorporatedigital transformations on the retail floor that cascadethroughout their operations.What all of these examples have in common are the waysthey enhance users’ experiences of the world, improvingtheir ability to share insights and take action. There’s apowerful secondary effect at work here too: the morethat users have access to such amplifying technology,the more they want access to real-time analytics toinform their next experiences, everywhere.Capabilities at the edgerefresh traditionalindustriesHaving access to instantly actionable decisions hasserved the industrial world well. Over the last twodecades, industrial companies have enjoyed radicalefficiency improvements by integrating digitaltechnologies throughout their operations—from RFIDtags in supply chains to robotics and remote monitoringand control in oilfield and pipeline operations.What’s changing to put digital-physical systems onboardroom agendas today is their infiltration into moreindustries and their ability to disrupt so many sectorsof global economies. The same improvements thatmanufacturers have made to drastically improve safetyand operational efficiency and, in some instances, toaugment the ability to scale are now expanding to everyindustry. Today’s digital-physical systems range fromchef robots that can serve a custom gourmet burgerevery 10 seconds to smart grid technologies that areable to identify individual appliances and their discreteenergy consumption—by simply installing a single deviceon a smart meter that can read, analyze, and deciphercomplex electrical frequencies. 11Building owners, property managers, and industrialequipment vendors are benefiting from the reduced costand increased capabilities of digital-physical systems todrive new experiences with building operations. Theyare adopting integrated solutions such as continuous#techvision2014

TREND 1: DIGITAL–PHYSICAL BLURcommissioning systems that make use of sensorsthroughout a building’s workspaces and mechanicalequipment to collect data with millisecond resolution onbuilding performance. Four years ago, Microsoft reliedon a team of reactive operations staff traversing theenormous corporate campus in trucks responding tohot and cold calls. Today, the company uses a softwareoverlay which visualizes the existing 500 million datatransactions every 24 hours and is managed by a staff ofhighly skilled engineers who is harnessing big data andusing it to improve Microsoft employees’ experience anddrive down energy costs. 12 Schneider Electric is goingone step further by incorporating some of the world’smost advanced weather modeling data into its buildingautomation systems, allowing facilities managers tofactor external climate conditions into their decisions andmaking it easier to automate many of those decisions. 13Industries from warehousing to agriculture are seeingsimilar advances. With robotic systems for orderfulfillment enabling big reductions in shipment times,warehouses can process more orders in less time withfewer employees. Taking robots to the skies, the use casesfor autonomous drones are abundant and are disrupting“The more that users have access toamplifying technology, the more theywant access to real-time analyticsto inform their next experiences,everywhere.”19ACCENTURE TECHNOLOGY VISION 2014

TREND 1: DIGITAL–PHYSICAL BLUR20a diverse group of industries. Agricultural applicationsuse infrared cameras to pinpoint crops that are receivingtoo much or too little water, first responders use dronesto detect survivors of accidents and natural disastersand to deliver supplies to those emergency zones, androad surfaces and traffic congestion are monitoredby drones. The EU has identified robotics as a highgrowthindustry and is committing resources to ensure astrategic leadership position. Together with a consortiumof ten European companies led by Shell, the EuropeanCommission has committed €3.7 million to a €6.2 millionproject to develop robots which can replace humans ininspections of pressure vessels and storage tanks usedextensively in the oil, gas, and petrochemical industries.When deployed, these robots will increase worker safety,reduce exposure to hazardous conditions, and increaseeconomic development by creating new jobs and openingnew markets for the European robotics industry. 14Individuals are seeing the benefits of augmentation aswell. Philips is piloting the use of Google Glass to allowphysicians wearing the display to simultaneously monitora patient’s vital signs and react to surgical proceduraldevelopments without having to turn away from thepatient or procedure. 15 Gartner predicts that the fieldservice industry alone stands to save $1 billion annuallyby 2017 by using smartglasses—the field now beingpioneered by Google Glass. The research firm notesthat smartglasses will enable field service technicians to“diagnose and fix problems more quickly and withoutneeding to bring additional experts to remote sites.” 16Disruptions aheadSupported by their abundant resources and their abilityto scale, large companies can now use digital-physicalsystems to disrupt their industries—and other industriestoo. To act disruptively, they must not simply use digitalphysicalsystems to improve today’s processes andservices; they have to re-imagine the end-to-end deliveryand experience of those processes and services. In doingso, there’s ample opportunity not only to disrupt existingindustries but to define new markets.Etihad Airways has reimagined what it means to operatean airline in a digital world. Working with Taleris, aprovider of intelligent operations services, Etihad Airwayswill tap the industrial Internet and use sophisticatedsoftware to harvest and analyze data generated by#techvision2014

TREND 1: DIGITAL–PHYSICAL BLURhundreds of sensors working inside its planes. The toolswill allow Etihad to monitor planes in real time, reducefuel costs, manage plane maintenance, and even spotproblems before they happen. 17Enabling many of these advances is the concept thatdata can be acquired, analyzed, and acted upon in realtime. The underlying concept is hardly novel: it applied tothe earliest anti-lock brake systems on cars, and it’s trueof many more complex systems today. The difference,however, is that today’s digital-physical systems oftenhave orders of magnitude more data to help make moreinformed decisions within a window of opportunitythat matters.This “iteration capital” is a force multiplier at Ford Motor,where 3-D printers radically reduce iteration cycles inthe design process, saving an average of one month ofproduction time for new engine parts, for instance. 18 Atmotorcycle maker Ducati, design cycle time has beenhalved by leveraging 3-D printing. 19 And the U.S. Armyis experimenting with 3-D printers to eliminate the needto carry so many spare parts. The precision enabled bytoday’s additive manufacturing technologies is leveragedby GE to make parts that were simply not possible withprevious manufacturing technologies. 20 In all of theseexamples, companies are pushing decisions to the edgein ways that were never before possible.The entire transportation ecosystem is ripe for disruptionby digital-physical systems. When driverless carsbecome common, not only will they change commuters’experiences, they are expected to reduce the incidenceof traffic accidents, improve the density of road use,smooth subsequent planning for maintenance and newroad construction, ease long-term planning for othertransportation systems such as light rail, and muchmore. Overall, driverless cars will radically disrupt theshipping and logistics industries, fleet services, publictransportation, taxicabs, rental cars, agriculture, andmining industries. What’s more, insurance liabilitymarkets will likely undergo dramatic changes as carmanufacturers move to self-insure, offsetting theirdistributed risk via reinsurance markets.In case there is any doubt that such technologies arein our future, public policy—often the biggest laggardin innovation economies—is keeping pace. In the U.S.,California, Nevada, Michigan, and Florida have alreadypassed legislation allowing driverless cars on the roads. 2121ACCENTURE TECHNOLOGY VISION 2014

TREND 1: DIGITAL–PHYSICAL BLUR22“Organizations that understand thepotential of behavior shaping andrespect its privacy implications willenjoy consumer loyalty and employeeengagement.”The U.S. federal government has instructed its FederalAviation Administration to publish rules that will permitdrones to use public airspace by September 2015. Nearlya dozen EU member states have no laws preventing smalldrones from flying below 150 meters. 22Decisions on the edge are powering innovations forconsumers and businesses alike. To lead, companies needto identify the decisions they would like to shape and theplaces where they can build decision spaces to informthose decisions. From there, companies will have newopportunities to shape behaviors in new ways.Shaping consumerbehaviorWhat’s perhaps most intriguing about digital’samplification effect is that it’s directly allowingbusinesses and the enterprise to identify and replicateproductive behaviors. Furthermore, it’s creating theexpectation that every experience can and should beoptimized. There is significant potential to shift employeebehaviors away from preparing for every permutation of#techvision2014

TREND 1: DIGITAL–PHYSICAL BLURan experience and toward developing new and simplerexperiences altogether. These experiences prove to bebetter for the end user and to be more cost-efficient forthe enterprise to implement.Royal Caribbean International cruise line has shown itsgrasp of technology’s ability to empower travelers andoptimize customer experiences. Eager to ease long linesfor shipboard restaurants—a perennial complaint frompassengers—Royal Caribbean uses sensors to relay topassengers the real-time seating availability at eachof the restaurants on its ships. In other words, RoyalCaribbean moved dining decisions to the edge by givingits passengers the data they need to decide where andwhen to have a meal. Passengers are happier and thecruise line was able to discard cumbersome schedulingprocesses, making it a win-win situation for everyone. 23This behavior-shaping capacity can be tuned by the waysin which information is presented to a user. For example,if a city wants to reduce its collective carbon emissions,it might offer a transit planning app that provides routeplanning guidance on the least carbon-intense mode orroute in addition to the quickest.Organizations that understand the potential of behaviorshaping and respect its privacy implications will enjoyconsumer loyalty and employee engagement. The keyis to notify individuals about how their activities arebeing monitored, give them the choice of opting in,and explicitly share with them the choices of actionableinformation at the decision point.These aspects are crucial. Privacy issues are likely tokeep making headlines as privacy watchdogs jump into defend against unauthorized tracking of citizens andconsumers. The company that can build a reputationfor providing valuable services while using consumers’personal data in trustworthy ways will have bigadvantages over competitors. Its brand will be morevaluable, it will have more opportunities to attract andretain lifetime customers, and it can become a preferredpartner in a larger value chain of goods and services.23ACCENTURE TECHNOLOGY VISION 2014

TREND 1: DIGITAL–PHYSICAL BLUR24Next steps for businessesAre consumers already buying smartwatches? They are.Is there a long waiting list for Google Glass? You bet. It’sfair to say that there will always be individuals who areeager to acquire technology to become better informedand to be able to enjoy different and hopefully betteruser experiences.Businesses are likely to be more circumspect about themigration of decisions to the edge. Accenture expectsthat business-to-consumer organizations will be quitequick to respond to consumers’ eagerness for andopenness to new experiences, offering everything fromnew mobile couponing opportunities to new ways tomonitor and improve their health.In the broader business world, there are enormousopportunities to move employees’ decisions closer towhere they can take action. That much is apparent whenso many employees keep their mobile phones close tothem in the workplace. But businesses’ embrace of thenext generation of automation will be cautious.Accenture anticipates three phases of uptake. The firstimpact will be on making current ways of doing thingsmuch more efficient. The second phase will see digitalphysicalsystems start to create industry disruptions.Disruption will begin as it always does—by changingusers’ expectations of what is acceptable, normal. Thebusinesses that proactively alter users’ experienceswill be the disrupters. The third phase will be in howorganizations respond. They will need to ask questionsabout how truly intelligent automation will changeinteractions with and expectations of their customersand other stakeholders. Will it open up new businessopportunities? Will it change the productivity equationin the workplace? Will it materially change how we planour use of resources? Will it simplify our organization’sstructure?The businesses that arrive at the best answers—andthat can properly leverage the strengths of machines(precision and scale) alongside the strengths of people(insight and decisions)—will be setting themselves upwith market-leading advantages.#techvision2014

TREND 1: DIGITAL–PHYSICAL BLURYour 100-day planIn 100 days, promote decisions at the edge by completingthe following:• Take an inventory of devices at the edge of your network;segregate them by those used by humans and thosethat act on their own, such as sensors and embeddedintelligence.• Catalog how data is currently being collected in yourorganization to drive business decisions. Understand howhaving more data about daily operations could improvebusiness outcomes.• Define and prioritize both the ways in which consumersengage with your products or services and the locationswhere they engage. Brainstorm ways to deliver compellinguser experiences that offer new insights into theirdecision making.• Consider how you can influence behaviors or decisionmaking to help consumers arrive at a favorable outcomefor your mission or business.• Look to early adopters to learn what businesses in differentindustries are doing to enhance consumer experiences,enable field workers, and embed intelligence into theirphysical assets.• Organize a cross-functional mobility team between your ITand business organizations. Their objective will be to pilotrelevant hardware innovations and test new consumer andemployee digital-physical experiences.• Collaborate with your customer-facing business unitsto capture the types of edge decisions they often make.Determine how they will benefit by adding data with realtime analytics at the point of action and create a strategyto deliver that solution.• Reevaluate your corporate privacy policy to address thenew digital-physical interactions for your business. Datacollection, usage, transparency, and user control (opt in/optout) guidelines should be clearly addressed.• Uncover the types of decisions that can alleviate oversightobligations from middle management and start to builddecision spaces for front-line workers to take autonomousactions.25ACCENTURE TECHNOLOGY VISION 2014

TREND 1: DIGITAL–PHYSICAL BLUR26This time next yearIn 365 days, you should step up your business agility bypushing decisions to the edge:• Develop a portfolio of pilots to deliver actionable insights toemployees and customers and that considers hardware andsoftware solutions. Aim for increased data resolution overwhat is provided today. Monitor outcomes related to thosedecisions, and consumer experiences.• Extend your infrastructure to support enterprise mobilityfor core business functions.• Develop a real-time data analytics infrastructure to supportthe data velocity and insight needs of digital–physical projects.• Develop a governance strategy to act on real time feedbackloops to enhance decisions at the edge.• Proactively address potential data privacy issues as newpilots and projects are developed. Urge leaders to gobeyond compliance, giving end users transparency andcontrol in an effort to mitigate corporate risk and liability.• Start planning for known technology disruptions comingdown the pipeline. Example: your business will make use ofautonomous vehicles and aerial drones. Be disruptive. Planon aerial drones being available for use in late 2015 anddriverless cars in 2020.#techvision2014sidebarDigital augmentation makes everyworker an information workerTo-date, corporate technology investments have focusedon improving the efficiency of only certain staff: highcostknowledge workers. That is changing. Intelligentdevices are now sufficiently abundant, inexpensive, andconnected, empowering workers at every level to performtheir jobs with greater efficiency, productivity, and safety.Front-line workers are becoming information workerswith some of the same augmentation technologies thatimprove the performance of knowledge workers.These new information workers are able to makeinformed decisions in real time, acting in the tangibleworld with the right information at the right time—oftenusing their own mobile devices rather than technologysupplied by the organization, yet interacting with theorganization’s back-end IT systems.For example, paramedics in Champaign, Illinois, andGrand Rapids, Michigan, are using their own smartphonesand tablets to increase access to medical information,find drug dosages and interactions, and share insights

TREND 1: DIGITAL–PHYSICAL BLURwith the destination hospital while a patient is in transit.In the past, if this research happened at all, it wouldhappen when the paramedics arrived at the hospital orreturned to their depots. Now, the research can happenin near real-time and have meaningful impact on thecourse of treatment. 24 The paramedics are newly enabledby information that provides critical insights within anarrow window of opportunity; those insights can spur adecision “on the edge” that can possibly save a life.In similar ways, developments with wearable technology,such as Google Glass, may provide field-serviceprofessionals in the oil and gas industry with access toreal-time information and deep expertise—improvingtheir effectiveness at fixing remote pipelines andmaintaining highly sensitive infrastructure.As front-line workers have their capabilities augmentedby digital technologies, they are emboldened to makemore informed, real-time decisions and encouraged tobecome more engaged with the organization. This drivesoperating efficiency and revenue growth. Studies haveshown that companies with high employee engagementfrequently demonstrate higher levels of operating incomeand growth in earnings per share than those found tohave low levels of employee engagement.The implications of digital workforce transformations gobeyond updating bring-your-own-device (BYOD) policies.Realizing the efficiencies from digitally augmentedworkforces will require business leaders to fundamentallyrethink how employees do their jobs and reassess theback-end systems that support them. Leaders will alsoneed to be comfortable with decentralized decisionmakingas business decisions get pushed closer to theedges of their business. Chief information officers(CIOs) will need to be more cognizant of potential risksassociated with data privacy and security. Real-timeinformation systems will become a higher priority.Getting information to workers as they need it willallow businesses to uncover a workforce that is moreproductive, eager for new challenges, and more effectiveand efficient at making and sharing decisions bycollaborating with others. The benefits can rise abovequantity and quality of work output by contributing tothe flattening of organizations—meaning that a smallernumber of supervisors and middle managers can managelarger and more dispersed teams.ACCENTURE TECHNOLOGY VISION 201427

TREND 2: FROM WORKFORCE TO CROWDSOURCE28#techvision2014

TREND 2: FROM WORKFORCE TO CROWDSOURCETREND 2From workforce to crowdsource:The rise of the borderless enterprisePicture a workforce that extends beyond your employees: onethat consists of any user connected to the Internet. Cloud, social,and collaboration technologies now allow organizations to tapinto vast pools of resources across the world, many of whom aremotivated to help. Channeling these efforts to drive businessgoals is a challenge, but the opportunity is enormous: it cangive every business access to an immense, agile workforce thatis not only better suited to solving some of the problems thatorganizations struggle with today but in many cases will do itfor free.29ACCENTURE TECHNOLOGY VISION 2014

TREND 2: FROM WORKFORCE TO CROWDSOURCE30Why now?Accelerated pace of IT change: The increasing pressure torapidly deploy new technology is accentuating some of anenterprise’s biggest pain points: market insight, innovation,and a need for highly specialized skills. These are areas forwhich crowdsourcing solutions are well suited.Maturation of crowdsourcing platforms: Communities ofshared interest have organically formed or are forming aroundalmost every product, service, and idea that can be imagined.Crowdflower, Spigit, and Mechanical Turk are just a few of thecollaboration platforms that are rapidly evolving to enableand orchestrate efficient solutions.Strong case studies from early adopters: Some of thebiggest market disrupters, such as Facebook and largeenterprises including GE, are currently using crowdsourcingservices to solve their most complex problems, and everyoneis taking notice.Where do leading companies like GE,MasterCard, and Facebook go to solvetheir toughest data science problems?They don’t always turn to the professionals on their ownpayrolls. Rather, they are beginning to turn to companieslike Kaggle—a global network of computer scientists,mathematicians, and data scientists who compete tosolve problems ranging from airline flight optimizationto retail-store location optimization. This is just oneglimpse of how the concept of the corporate workforceis changing.Over the last decade, organizations have been usingincreasingly advanced tools and processes to boostcollaboration among their employees. Videoconferencing,instant messaging, blogs, wikis, and activity streams haveall become the norm as large companies push to connecttheir employees across groups, skills, and geographicboundaries. Outside of the enterprise, these patterns areeven more pronounced: tools such as Twitter and Yelpand Wikipedia connect huge swaths of the populationto discuss and collaborate on everything from new-carreviews to health care.#techvision2014

TREND 2: FROM WORKFORCE TO CROWDSOURCEGiven that so much collaboration happens throughdigital channels, there is the potential for almost limitlesscollaboration with everyone else who is connected to theInternet—regardless of whether they are “our” employeeor not. Which raises a crucial question for business andIT leaders: “Are we missing out by not connecting to this‘expanded workforce,’ everywhere and in all directions?”The short answer is “yes.”Digital connectionplatformsMore and more digital platforms are available that makeit easier to connect to what Accenture calls the expandedworkforce. Kaggle is just one of these; others includeAmazon’s Mechanical Turk, along with services such asElance and oDesk. All of these can be considered onlinelabor markets. They help companies that have tasks thatneed to be accomplished connect with resources thathave the right experience, ability, and time.Although many organizations are experiencing thebenefits of digitally enabled forums such as innovationexchanges and crowdsourcing platforms, few executivesyet fully grasp the idea of being able to access a trulyliquid workforce—pools of premier talent gathered invirtual communities and coalescing around specificbusiness problems. This expanded workforce likely offersnot only expertise that is not immediately available inhousebut also real scale. It can be leveraged to solveproblems that may be too large or too expensive to solveinternally.Name almost any challenge—early detection of driverdrowsiness or the predictability of drug targets orelectric-only updates to hybrid cars—and there areoften already communities of experts that companiescan leverage to competently address it. The individualsinvolved may be around the corner or on the other sideof the world; what they have in common is not only theexperience and expertise to solve the problem but themotivation—in many cases the passion—to do so.The tasks involved may be as simple as data entry oras complex as industrial design. The individuals—theproblem solvers—may work on a project or just partof a project. They may be paid; they may compete forprizes. But whatever their incentives and their spheres ofinterest, the unifying feature is that their contributionsare made possible with digital tools.ACCENTURE TECHNOLOGY VISION 201431

TREND 2: FROM WORKFORCE TO CROWDSOURCE32The expanded workforcein actionTake the automotive startup Local Motors for example.Its Rally Fighter car can’t easily be called beautiful, withits scrawny jellybean body perched high above ganglystruts and monstrous off-road wheels. Its real beauty isin how it came to be—not how it looks.A harbinger of the future, Local Motors created a globalcommunity of car enthusiasts that included engineers,mechanics, and industrial designers and broke downthe creation of the vehicle into a set of tasks thatwere widely distributed, via the cloud, to this eclecticworkforce. In just 18 months, those individuals—workingclosely with the company’s employees—designed,manufactured, and delivered a car that this usercommunity loves. 1Local Motors provides a provocative example ofleveraging an unconventional workforce to accomplishsomething previously thought impossible. So intriguingis Local Motors’ social-plus-collaborative take on carmaking that it has attracted some blue-chip partners,such as BMW. BMW hopes to use the startup’s expandedworkforce model to move beyond typical focus groupsand inference, developing and testing new automobileinteriors that reflect users’ true desires rather thanincomplete and veiled answers. 2Local Motors’ real innovation is conceptual. Its leadersimagine and leverage limitless talent instead of beingboxed in by what has traditionally been a predeterminedworkforce. They see talent that may already be waitingto solve some of the challenges the organization has yetto confront. “There are two ways to build things,” LocalMotors’ co-founder and chief executive Jay Rogers toldan industry publication. “You can hire the relevant peopleto solve a problem—or you can organize in the cloud toget better ideas faster.” 3A crucial point to make is that the use of the expandedworkforce is not another form of labor arbitrage. It is notto be confused with employing contractors or temporarylabor or moving to an outsourcing arrangement. Thechannels, structures, and transactions are entirelydifferent—far more fluid and versatile than any familiarforms of accessing human resources.#techvision2014

TREND 2: FROM WORKFORCE TO CROWDSOURCEIt will be essential that borderless enterprises work toharness the energies and enthusiasm of the expandedworkforce. The current workforce is not going away,of course; not every problem will be well suited tocrowdsourced solutions. However, it is no longer enoughto rely only on groups of in-house individuals to drivemarket research, innovation, and product-developmentactivities. Digital technology has brought a global voiceto those functions. It is pushing out the boundaries thatpreviously defined the enterprise workforce. It’s not anoverstatement to say that business leaders now neednew perspectives on the nature of work itself.Let’s walk through a few of the ways in which leadingbusinesses are starting to harness the expandedworkforce.Marketing moves towardperfect informationThe expanded workforce is already changing theway companies market their products and services.Technology platforms that promote comments, userinteraction, and even consumer investment are givingconsumers a direct voice with which to communicatewith the marketing department. In doing so, theseplatforms create a porous membrane between paidemployees and the expanded workforce. Consumers areproviding richly detailed information, giving companiesan unprecedented level of insight into their products,how they’re used, and the consumers that buy them.With the right digital tools in place, producers canpredict better than ever how the market will react totheir products and who will buy them. They can segmentmarkets more discretely and test premium feature setsto see who will pay for them. The new approach tradesabstract market projections for data-driven marketdecisions; consider the difference in outcomes between33ACCENTURE TECHNOLOGY VISION 2014

TREND 2: FROM WORKFORCE TO CROWDSOURCE34running small focus groups city-by-city and runninga crowdsourcing contest to solicit customer input.Responses from 125,000 real-world consumers certainlyare more statistically sound—and arguably more honest—than responses from 125 focus-group attendees.The benefits of expanded-workforce input go furtherstill. Direct consumer feedback can furnish data onpre-order sales that enables not only sales projectionsbut also raising of working capital that can be used tobegin manufacture of the actual product. In this way,the expanded workforce becomes an invaluable tool forreducing product inventory risks. This is market insightthat is not confined by four walls or hidden behindartificial representations of consumers’ wishes.A few examples stand out. Walmart Labs, the digitaltechnology division of the world’s largest retailer, isembracing the power of the crowd to determine whichitems the company should stock in its stores and on itswebsite. Using a contest titled “Get on the Shelf,” whichwas heavily promoted on Facebook, Google+, and Twitter,the company offered a way to democratize the job of thestore buyer and bring shelf selections in line with shoppers’expectations—and wants. The effort was intended onlyas a “fun experiment”—which it was, but it also providedWalmart with a clear view of the demand for productsnot currently on its shelves. Consumers voted online forthe products they wanted to see sold at Walmart; winningentries for new products won prizes and a grand-prizewinner was chosen from the top three contestants. 4Similarly, snack maker Lay’s has used consumers’ creativityto launch new flavors. The company has run flavornaming and defining contests in more than 15 countriessince 2008. One recent U.S. initiative that used theslogan “Do us a flavor” drew nearly 4 million proposals.The winning flavor—“cheesy garlic bread”—was chosenfrom the votes of the company’s fans and by a panel ofexperts who included a popular TV actress and severaltop chefs. The new flavor is now widely available for sale. 5Of course, there is no stronger indicator of marketinterest than customers’ willingness to part with theirmoney. These days, more and more businesses are using“crowdfunding” platforms such as Kickstarter to helpvalidate product development. Kickstarter is the granddaddyof crowdfunding sites—new businesses that use Webbasedcollaboration, social media, and microfinancingtechniques to raise money for everything from new film#techvision2014

TREND 2: FROM WORKFORCE TO CROWDSOURCEprojects to seed funding for small businesses. SinceKickstarter launched, in 2009, over 5 million people haveused the platform to pledge more than $900 million,funding more than 50,000 creative projects. 6Companies are using Kickstarter and similar platformsnot just for fundraising but to provide market insightand product viability assessments for a fraction ofwhat those services typically cost. By determining whatpeople will actually pay for, the process influences andvalidates products and pricing strategies and sometimesleads to initial consumption and to product advocates.Accenture has developed initial models that show thatcrowdsourcing, when used correctly, leads to higherprofits for producers.Tesla Motors is an example of a company that has takenthis idea to the next level. The maker of electric cars hasasked for advance reservation fees from customers—$5,000 per car—not only confirming the extent of demandbut providing Tesla with working capital to the tune of$130 million. These “funds” are willingly provided byeager buyers for zero percent interest. Otherwise, Teslamight have to pay 10 to 15 percent to a bank. 7“With the right digital tools in place,producers can predict better than everhow the market will react to theirproducts and who will buy them.”35ACCENTURE TECHNOLOGY VISION 2014

TREND 2: FROM WORKFORCE TO CROWDSOURCE36Taking innovationoff its leashInnovation is now at or near the top of the C-suite agendain every organization. But it remains difficult to execute—difficult to scale up and to ramp up fast, and hard toensure that the results are of the quality expected. Yetinnovation is happening organically everywhere, whetherbusiness leaders are aware of it or not; communities ofshared interest have formed or are forming around almostevery product, service, or idea that can be imagined.There are new platforms, such as Spigit, that canhelp to propel the innovation process quickly andeffectively. Another example is TopCoder, a mechanismfor running computer-programming competitions. Suchplatforms capitalize on the very human urges to create,solve problems, and pursue knowledge. Many peopleare powerfully energized by the idea of developinginnovations that solve big, intractable problems. The zealof those working on the Human Genome Project, anambitious undertaking to fully map the human genome,is ample evidence of that.Organizations can’t depend solely on existing oremerging innovation solutions. They should be willing tocreate comparable platforms and communities. Roughlya decade ago, pharmaceutical company Eli Lilly spunoff InnoCentive, an innovation exchange that now hasmore than 300,000 registered problem solvers from 200countries. Essentially, the company marries this expandedworkforce to cloud-based technology to solve problemsposed by customers. Winners receive substantial cashprizes. To date, more than 1,500 awards have beengiven, with the size of the award ranging from $5,000to more than $1 million, based on the complexity of theproblem and nature of the challenge. 8Similarly, companies such as Facebook and Twitter areleveraging the intellect and energy of the crowdsource byusing application-programming interfaces to open up theirplatforms to the development community at large. Thisopen approach encourages individual developers andcompanies to innovate and create products that solveproblems and issues that might never surface in theinteractions of those companies’ conventional workforces.The open approach not only motivates the users of theplatforms; it creates substantial additional value for theplatform and subsequently for the platform owner itself.#techvision2014

TREND 2: FROM WORKFORCE TO CROWDSOURCEFast-tracking productdevelopmentThe open-source community is the original expandedworkforce—the trailblazers, united by ideas and interests,who helped write the rules and define the tools forfreeform ways of developing software. Collectively,these impassioned individuals make up a global armyof developers who are creating and improving freesoftware, bringing worldwide benefits server by server,device by device, for free.This expanded workforce has touched every organizationin some way. Witness the widespread use of two outputsof open-source projects: Hadoop and Linux. The first isthe engine that is powering the Big Data era; the secondis the kernel operating system embedded in 23 percentof installed enterprise servers, over 80 percent ofsmartphones shipped, and countless other systemsranging from embedded sensors to supercomputers. 9It can be argued that the open-source movement haschanged the face of software development just as muchas the move from punch cards to hard disks did. Opensourceinitiatives are foundational for companies suchas Yahoo and Google. Android is based on Linux; it isprojected to constitute nearly half of all mobile operatingsystems (smartphones and tablets) by 2017. 10Nowadays, the largest technology players all havesignificant roles in open-source development. Companiesincluding IBM, Microsoft, and Oracle—once stronglyopposed to all things open-source—are official, vital,and prolific contributors to the open-source community.The open-source paradigm has now moved to hardwaredevelopment, as seen with Facebook’s Open ComputeProject and Hewlett-Packard’s Pathfinder InnovationEcosystem Solutions initiatives. But it shouldn’t belimited to just tech companies. Businesses of all kindsshould engage in these open-source communities. Bycontributing ideas, time, and code, companies help setthe overall direction of product development, allowingthem to leverage not just software that is free butsoftware that targets their specific needs and problems.Software is also just one example of using the expandedworkforce for product development. From smart watchesto smart vehicles, the expanded workforce is beingemployed by bold companies to energize their product37ACCENTURE TECHNOLOGY VISION 2014

TREND 2: FROM WORKFORCE TO CROWDSOURCE38development. The expanded workforce is becoming aninvaluable filter for reducing product development risks,improving time to market, and determining receptivityto new-product introductions. For example, the Pebblewatch Kickstarter campaign not only raised more than$10 million in 2012 but allowed the company to gathervaluable product intelligence about pricing and productdemand. By interacting directly with its target customerbase, Pebble was able to determine that consumers werewilling pay $150 for one of its watches and that 85,000watches, once funded, would be mailed in the first month. 11Using the crowd—and its assetsInnovating, marketing, and developing products are byno means the limit of what the expanded workforce canhelp with. The concept is expanding to include people’sassets as well as their time. Just one snapshot: LocalMotors quite literally uses its customers as free labor,and the buyers of its Rally Fighter car actually pay for theexperience of participating in the car’s assembly.In fact, the use of the crowd’s assets can launch newbusinesses that are well positioned to attack longestablishedsectors. Airbnb is a classic example. Billed asa trusted community marketplace where people can list,discover, and book unique accommodations around theworld, Airbnb uses the crowd as the untapped source ofplaces for travelers to stay. In effect, everyone can offertheir own home or apartment as a kind of hotel. Airbnbis possible because digital tools—particularly mobilephones—make it very easy to find, select, and obtainaccommodation. Not surprisingly, the startup—it wasfounded in 2008—is seen as a significant threat by thehotel industry.Similarly, Uber crowdsources assets (in this case, cars)by connecting drivers with people needing a ride, amodel that the taxicab trade views with alarm. AndRelayRides—through which car owners rent out their ownvehicles—presents competition to the established rentalcarbusiness.Current markets are being disrupted and new marketsare being discovered by companies that are employinglatent talents or assets made available through digitaltechnology. Businesses can no longer be on the sidelines#techvision2014

TREND 2: FROM WORKFORCE TO CROWDSOURCEwatching and waiting to see what will happen nextand hoping to grab the coattails of the next big idea.The enterprise needs to be out there experimenting,discovering, and creating the next big idea.From doingto orchestratingSo, there is no shortage of people willing to participatein online experiments, contests, challenges, and more.Individuals are surprisingly ready to work for little orno money if they get other rewards: prizes, recognition,fame, the sense of pride in getting to create something.But how do business leaders then effectively manage theflow of talent and ideas? How do they effectively keepcontrol when their workforce moves from hundreds orthousands to tens of thousands?Given the relative immaturity of these crowd-basedservices and platforms, there’s still much to be learned.The strengths of these services can also be theirlimitations. However, although there are no clear linesor absolutes when employing the expanded workforce,there are already some useful guiding principles forplanning and implementing platforms for crowdsourcing.The use of the expanded workforce demands very diligentplanning. There must be clear, shared objectives for thecrowdsourced exercise, whether the requirement is fornew solutions to tough engineering challenges or forfunds for new-product development. And complex tasksmust be clearly and logically broken down into a seriesof independent subtasks that can be parceled out to thecrowd in a way that allows the crowdsourced efforts tobe reintegrated into the overall product or project. In thecase of labor market exchanges, there can be no subtletyin the work description; amplified across hundreds orperhaps even thousands of “workers” in the expandedworkforce, inexactness will produce meaningless resultsor work that cannot be re-aggregated.This need for structure requires more than just planning;typically, it will also require a technology solution to driveusers through the process. A superb example of this is theDARPA Language Challenge, which leveraged Amazon’sMechanical Turk platform. The challenge was to translate,from Arabic to English, messages that were communicatedon social platforms, where traditional grammar rules arenot used and context is paramount. Using the correctplatform, precise planning, and the expanded workforce,ACCENTURE TECHNOLOGY VISION 201439

TREND 2: FROM WORKFORCE TO CROWDSOURCE40DARPA was able to translate more than 200,000 wordsper week from Arabic to English. This project requiredpartitioning the work up meticulously, with explicitquality-assurance steps, and ranked incentives for thosewho provided higher-quality translations. By using acrowdsourced model, DARPA was able to shift its effortfrom finding, recruiting and hiring the correct resourcesto where it will do the most good, innovation arounddeveloping the project, planning and incentives. 12It is also crucial to properly engage the expandedworkforce. With crowdsourcing, adoption is paramount;lacking a properly motivated and incented community,nothing gets done. InnoCentive provides a great example.The knowledge exchange has not become a powerfulplatform for innovation without paying close attentionto every aspect of its operations—from setting thesize of prizes for “solvers” to how it communicates thechallenges. One more snapshot: Quirky, a co-creationcompany whose crowdsourced product innovations aresold through mainstream retailers such as Target, has adisciplined weekly voting process—online as well as onpremise—toselect the product ideas that it will put intoproduction.It’s still early days for the use of an expanded workforce,and being a borderless enterprise brings its own challenges.For instance, the inherent transience and anonymity ofthe expanded workforce places sharp limits on traditionalhuman resources activities such as job training. Andit raises many thorny questions about the security ofintellectual property.Despite these challenges, the opportunity is immensefor enterprises that are willing to step up. The companiesthat get it right will find themselves with better insightinto their customers, more-innovative products andservices, and an increased agility to retool themselveswith the skills necessary to respond to the changingtechnology landscape. The question is this: how are yougoing to position your enterprise to reap the rewards?#techvision2014

TREND 2: FROM WORKFORCE TO CROWDSOURCEYour 100-day planIn 100 days, learn about the variety of options that contributeto the borderless enterprise and begin to create a strategy forhow you can harness the crowd moving forward.• Identify any existing enterprise connections to establishedexpanded workforce platforms.• Determine how your competitors are using crowdsourcing.• Evaluate the benefits your market research, productdevelopment and innovation functions could reap fromusing expanded workforce platforms.• Develop an initial strategy to engage existing onlinecommunities in support of your core functions. Where italigns with your core business, create a catalog of existingonline communities that are specific to market-research andproduct-development functions.• Consider assets from the open-source software communitythat are usable for your core IT functions and beginplanning how to integrate them.• Design and implement a pilot to leverage the establishedexpanded workforce that most aligns with the nature ofyour business.This time next yearIn 365 days, you should be familiar with the various types ofcrowdsourcing platforms that apply to your business and havestarted an integration strategy.• Identify which tasks are most easily broken into smallerindependent tasks and conduct pilots with existing orcustom-built expanded workforce platforms to resolve them.• For market-research and product-development functions,start to figure out how to engage consumer communities.Where communities are not yet mature or specific enough,develop a strategy to build them.• After completing a few pilots, test your orchestrationstrategies. Aim to seamlessly bring solutions, no matter howsmall, back into the enterprise.• Create and implement governance and cost-controlstrategies to respond to and manage infinite capacity.• Understand the types of specialized skills that cause surgesin demand for your organization. Determine whetheradoption of expanded workforces can resolve these surgesand reevaluate your hiring structure in response.• Develop an authorized and trusted talent cloud with anestablished expanded workforce community.41ACCENTURE TECHNOLOGY VISION 2014

TREND 3: DATA SUPPLY CHAIN42#techvision2014

TREND 3: DATA SUPPLY CHAINTREND 3Data supply chain:Putting information into circulationYes, data technologies are evolving rapidly, but most have beenadopted in piecemeal fashion. As a result, enterprise data isvastly underutilized. Data ecosystems are complex and litteredwith data silos, limiting the value that organizations can getout of their own data by making it difficult to access. To trulyunlock that value, companies must start treating data more as asupply chain, enabling it to flow easily and usefully through theentire organization—and eventually throughout each company’secosystem of partners too.43ACCENTURE TECHNOLOGY VISION 2014

TREND 3: DATA SUPPLY CHAIN44Why now?Corporate data silos: Data is the lifeblood of every digitalorganization, but businesses are struggling to access, share,and analyze much of the data they already have. Through2015, 85 percent of Fortune 500 organizations will be unableto exploit big data for competitive advantage. 1Rising data volumes: In addition to the data thatorganizations already collect, new external data sources areavailable, providing new opportunities for data insights. Thedigital universe is doubling every two years and is expected togrow to 40 trillion gigabytes (more than 5,200 gigabytes forevery man, woman, and child in 2020). 2Maturing data technology: The tools and technologyrequired to build a data platform, ensuring data access andvelocity, are available and in use. For example, a reported20 percent of enterprises are already using NoSQL. 3 With thefoundation of these technologies, the integrated, end-to-enddata supply chain is possible.For years now, data has been talkedabout as the premier strategic IT asset.Today, it’s so much more than that.Business leaders now view data as among their mostvaluable assets too—some even call it the lifeblood oftheir organization. That’s why they’re implementing thenewest big data tools, investing in advanced analyticsapplications, and purchasing the latest data visualizationsoftware.Yet the reality is that these easily become one-off datafixes that contribute to data silos rather than providean end-to-end data solution. Few companies havemastered the concepts at the foundation of modern datamanagement—ideas such as the mobility and portabilityof data, its structure and velocity, data as a “saleable”product, and its valuation in open data exchanges.Fewer still are comfortable with these concepts at scale.Most efforts to properly manage data are ad hoc at best.No wonder half of all companies are concerned about theaccuracy of their data and a majority of executives areunclear about their analytics programs’ business outcomes. 4#techvision2014

TREND 3: DATA SUPPLY CHAINAccording to research by Gartner, 85 percent of Fortune500 organizations will be unable to exploit big data forcompetitive advantage through 2015. 5The high performers, however, will embark on a journeyto ROI; they will liberate their data, generate value fromit, and operationalize insights to drive strategic decisionsthrough the organization. Key to their success will bemanaging their data like any core product—in the contextof a supply chain. The supply chain begins when data iscreated, imported, or combined with other data. The datathen moves, flows, and transforms through the supplychain, incrementally acquiring value. And while theremay be diversions along the way, such as when a data“product” is removed for repairs (in other words, datacleansing), the supply chain ends with a valuable insightas its output. Guiding this movement is a data servicesplatform. Analogous to the blueprint of a factory floor,this platform provides the structure for the intelligenttransportation of data throughout the organization. Itenables the effective supply chain—fit to strategy anddesigned to drive outcomes.This means that business leaders now need to developan end-to-end view of data in order to achieve theirbusiness goals. Accenture’s earlier Technology Visionreports have dived deeply into the keystone data topics,highlighting the value of data platforms, the importanceof industrializing data services, and the need to think interms of the velocity of data as well as its variety andvolume. Now, all of those themes must come together inan end-to-end “supply chain” perspective that can helpbreak down the data silos—usually built and “owned” by asingle department—and enable data to flow freely for thebenefit of the whole organization.Working in this way, however, will change how theenterprise works with data at each stage. New sourcesof data can be leveraged. How data is managed andanalyzed will evolve. And companies will begin to explorenew ways to monetize their data. Understanding theimplications at each of these stages will be critical forbusiness leaders if they are to capitalize on the valueof their data now and in the future and, importantly,at scale.45ACCENTURE TECHNOLOGY VISION 2014

TREND 3: DATA SUPPLY CHAIN46Enabling the datasupply chain with a dataservices platformThe data supply chain must enable data movement. Andin order for data to move, it must be made visible andaccessible to those who need it when they need it. Assuch, the first step is to create a data services platformor federated data access layer, which provides a standardmethod of access to an organization’s curated andtrusted (albeit varied and siloed) data in a time-relevantmanner. Currently, only one out of five organizationsintegrates data across the enterprise. 6 But those few arerealizing great benefits. For instance, a major bank hasbeen able to use Palantir’s Capital Markets solution tointegrate 15 data sources into a single point of access.This has allowed hundreds of analysts and less-technicalbusiness users to do a variety of tasks, includingperforming regression analysis, developing hedges, andassessing investment risks. 7In order to realize their data platforms, organizationsmust first make their disparate data sources accessiblethrough data services—ranging from LightweightDirectory Access Protocol (LDAP) to Web services orapplication-programming interfaces (APIs). Regardlessof the specific type, the power of these data services isin their ability to mask back-end complexity and exposedata in defined ways. Once all of the data sources havebeen opened up, one common solution is to make themaccessible through a virtualized data layer, which unifieseverything into a single view. Then, users can interactwith this abstracted data platform in a standardized way.(Behind the scenes, they are actually being rerouted tothe data sources.)Many newer data platform strategies now depend onopening up each data source separately, but througha common standard access protocol. Increasingly,companies are turning to APIs to achieve this, and they’reoften aided by API management platforms. Walgreens,for example, has opened up its prescription API to enablethird-party developers to incorporate the technology toscan barcodes on prescription bottles, in order to make iteasier for customers to refill their prescriptions. 8#techvision2014

TREND 3: DATA SUPPLY CHAINThese are just two of many data platform solutionsin a market of vendors offering a wide variety ofmethods to create this federated access layer. Traditionalmiddleware providers such as Tibco Software and RedHat have evolved their messaging solutions to providethat abstraction layer. Companies such as Apigee aretouting their API management software as the solution.Additionally, Platform-as-a-Service (PaaS) providers suchas Amazon and Windows Azure are offering readymadesolutions already built into their clouds. And largedatabase vendors—such as Oracle and Microsoft—arecreating access layers to allow connections to datasilos in order to better manage the flow of informationthroughout an organization.In the end, there’s no one-size-fits-all solution; mostenterprises will end up with a hybrid set combiningmany of these tools. But no matter what the solution,it’s important to understand that data access and dataacceleration make the data services platform bothpossible and necessary—and thus help to realize the datasupply chain at scale.Accelerating datathrough the supply chainData access on its own isn’t enough—velocity is needed.Importantly, this does not mean that all data needs tomove at top speeds at all times but rather that dataneeds to be prioritized on the data services platform—such that important, time-critical data is acceleratedthrough the supply chain, while stale, less relevant datamoves more slowly but still meets the demands of thebusiness. Quick access to valuable data means thatanalyses can be performed, insights can be gained, andactions can be taken in the sometimes very small windowof opportunity available to businesses. Historically, ITprofessionals have addressed this problem by givingprecedence to “hot” data—data that is accessed frequentlyand saved onto high-performance systems that can storeand retrieve it very quickly. For its counterpart “cold”data—tax records, say—they have used slower diskhardware or even tape backups in legacy systems.47ACCENTURE TECHNOLOGY VISION 2014

TREND 3: DATA SUPPLY CHAIN48While these strategies have worked, newer prioritizationpractices improve data acceleration by adding manymore gradations of “data temperature”—or data tiers. Thismakes sense when businesses consider the wide range ofhow, when, and how fast users need to consume data.The next step is to enable dynamic movement of datathrough these tiers, meaning they can be “heated up”or “cooled down” at any time. The ability to seamlesslychange priority over time based on business needimproves data velocity—but it’s also an efficient andcost-effective capability.Facebook shows how. Not long ago, the social networkdiscovered that 8 percent of all Facebook photosaccounted for 82 percent of its network traffic. It turnsout that there is a marked drop-off in accessing photosas they age, meaning that the photos’ data may be toppriority at some point but not always. This led toFacebook’s development of its own three-tiered datastorage solution. Its software categorizes photos andstores them in the appropriate tier, which has dedicatedhardware to increase savings—the lower tiers (forcolder data) can store more photos and use less energy.While the hot, tier-one data can be accessed almostinstantaneously, one of the unique aspects of Facebook’ssolution is that even cold data can be retrieved quickly(most solutions require several hours) to better meet users’expectations. In this way, Facebook is both increasing itsphotos’ velocity and saving on storage costs. 9Tiered data solutions allow for time-critical andcommonly accessed data to be stored in data-centriccaching structures, optimized for quick transportthrough the supply chain. But in all likelihood, mostdata will eventually end up in the “data lake”—used tostore the vast quantities of an organization’s data thatare less time-sensitive or used less frequently. Althoughhistorically the “data lake” was tape, technologies such asApache Hadoop and Amazon Redshift have been able tokeep costs low while maintaining data accessibility. Witha greater number of data tiers, more data can be storedlonger and at a lower cost, without interfering withaccess to important, time-relevant data.#techvision2014

TREND 3: DATA SUPPLY CHAINMore uses formore data sourcesThe supply chain process starts with ingesting data.These days, companies can use a wide variety of newdata sources—including, notably, data that they do notcontrol or own. For example, if a regional grocery chainwants to analyze its daily transactions over the pastmonth, it should look beyond the data in its databaseto supplement its findings. Many companies are alreadytracking sentiment on social media sites; data can alsobe analyzed in the context of weather, characteristicsof shoppers, events in the news, or virtually any newdata dimension imaginable—if the relevant data can belocated. Whether this “external” data is to be obtainedfrom partners, Data-as-a-Service providers, or open datasources (free for anyone to use in any way they want),companies should capitalize on the business value thatthese new sources provide.Collaborate.org is one example of an open datasource, containing 5 petabytes-plus of data. 10 It is aglobal collaboration platform (meaning that users areencouraged to contribute data back to the platform) thatallows users to view data geospatially—such as satelliteimagery and air quality. The U.S. state of Hawaii is usingCollaborate.org to share data across organizations, andit is the platform for the Exemplary State Initiative,which monitors environmental efforts and enables earlyidentification of natural disasters. 11In similar ways, Beiersdorf, a global provider of skincare products, is using their own internal data alongwith syndicated data from research companies suchas Nielsen to provide board members with marketshare development information across a variety ofproducts, brands, and countries. Using SAP DemandSignal Management powered by HANA, Beiersdorf plansto automate its data integration process, leading tomore accurate and accelerated market share insights. 12These examples demonstrate the new opportunity thatbusinesses have to look beyond their four walls fordata that will grant more informed data insights and,ultimately, more value.49ACCENTURE TECHNOLOGY VISION 2014

TREND 3: DATA SUPPLY CHAIN50Advancing data discoveryThe process of discovering new insights to answerbusiness questions is changing fundamentally as usersget faster access to more data. Now, when data ismanipulated as it moves through the supply chain, valuecan be added to and obtained from it in ways that werepreviously impossible. This is because data discoveryallows businesses to discover answers to questions thatthey might never have known to ask in the first place.Previously, traditional business intelligence (BI) methodswere the only way to answer prescribed businessquestions; they require multiple lengthy steps before asolution is possible. Now, however, data discovery helpsdiscern the very questions that companies should beasking by uncovering insights in a visually interactiveand rapidly iterative manner. Effectively, data discoveryempowers users to “communicate” with data at close tothe speed of thought—accelerating businesses’ time toinsight. Companies can and should be investing in thispractice today.So that businesses can better “communicate” with andanalyze data, analytics are being embedded in datadiscovery tools (as they are in applications)—effectivelyenabling data scientists and less-technical business usersalike to do data discovery more easily and intuitively.As an example, the Teradata Aster Discovery Platformenables data scientists to do data discovery andadvanced analytics themselves, but it also allows them toextend those capabilities to business users by leveraginganalytic functions exposed through BI or other tools.When business users, and even business leaders, are ableto discover and answer their own business questions ina matter of minutes, intelligent insights quickly lead tointelligent actions.For instance, Texas Medical Center (TMC) understands theimportance of health care data to advance patient care.However, this data can also be extremely complex anddifficult for anyone—from medical experts to datascientists—to analyze. Therefore, TMC is using Ayasdi’stechnology—one that specializes in the analysis of highvolume,high-dimensional datasets—to empower users tofind insights in their data. Using a visual representationof the data and built-in statistical tools, users can more#techvision2014

TREND 3: DATA SUPPLY CHAINeasily perform “semi-supervised analysis” of the data. Theability to interact with data in this quick and iterative wayallows users to identify emerging patterns and acceleratetheir time to insight. This approach first proved its powerwhen analyzing a breast cancer dataset (one that hadbeen analyzed many times already); within just a fewminutes, a new subset of survivors was identified. 13TMC plans to use Ayasdi for a variety of applications,from analyzing clinical and genomic data to drugrepurposing. 14 The truth is, every organization hasunexpected insights waiting to be unearthed by datadiscovery methods. Now, it’s a matter of determiningthose insights and accelerating their time to insight.The next step:cognitive computingAs the volume and variety of data grow, so too do thescale and complexity of the data supply chain, making itincreasingly difficult to add to and get value from dataas it is manipulated. Imagine it this way, on a standardsupply chain: everyday, more and more raw goods(some of them new) are being delivered. Initially, thismight sound great—more supplies mean more products—but it’s not what the machines were designed to handle,and workers can’t keep up with the maintenancerequired. The supply chain becomes clogged, hinderingthe creation of valuable products. But what if, instead,machines could be taught to leverage data, learn fromit, and, with a little guidance, figure out what to do withit? That’s the power of machine learning—which is amajor building block of the ultimate long-term solution:cognitive computing. Rather than being programmed forspecific tasks, machine learning systems gain knowledgefrom data as “experience” and then generalize whatthey’ve learned in upcoming situations. Cognitivecomputing technology builds on that by incorporatingcomponents of artificial intelligence to convey insightsin seamless, natural ways to help humans or machinesaccomplish what they could not on their own. At itsmost advanced, cognitive computing will be the trulyintelligent data supply chain—one that masks complexityby harnessing the power of data to help business usersask and answer strategic questions in a data-driven way.51ACCENTURE TECHNOLOGY VISION 2014

TREND 3: DATA SUPPLY CHAIN52Although complex, large-scale cognitive computing maybe beyond the reach of most companies, there are somecognitive computing capabilities that can be put to workin practical and affordable ways. Companies should focuson tackling well-defined problems on a smaller scale—where machine learning techniques can be leveragedto accomplish practical cognitive computing goals. Forexample, Tempo, the calendar application found oniPhones, uses data contained on the phone—from socialmedia to email, location, and more—to “learn” aboutevents and display relevant information to the user whenrequested. This smart personal assistant applicationmasks the complexity behind the data supply chain asraw data is aggregated, analyzed, and turned into anactual event with value. Only then does the user naturallyinteract with the calendar through the phone’s interface.One interesting cognitive computing example comesfrom U.S. food company McCormick. Machines are nowstarting to use data to “sense” the world as humansdo, and this extends to taste—with obvious benefits forthe food industry. Using Enterra Solution’s CognitiveReasoning Platform, McCormick’s FlavorPrint site askscustomers to rate a variety of flavors in order to learntaste and, from that, creates unique taste preferenceprofiles—or what it calls FlavorPrints. If customersprovide additional information, such as cookingpreferences, equipment, and typical pantry items,they can receive better personal product and reciperecommendations. As far as these customers can tell,they’re providing just a few raw facts in return for agreat deal of personalized value about taste—somethingalmost everyone feels strongly about yet finds hard toquantify or specify. From McCormick’s point of view,learning customers’ taste preferences leads to betterinsights, product decisions, and, ultimately, ability toserve its customers. 15Cognitive computing can, and will, bring benefits tomany industries, and it will fundamentally change theways in which many businesses operate. It flips theproblem of data volume and variety on its head andinstead leverages it to enable the smart, interactivedata supply chain. The ultimate goal is for any businessuser—from a CEO to a field worker—to be able to ask anybusiness question and immediately get a data-drivenanswer from the masked data supply chain.#techvision2014

TREND 3: DATA SUPPLY CHAINAlthough this technology may seem far off, thereare already cases that prove its relevance. And by itsvery definition, with more data over time, cognitivecomputing technology will only learn more, adaptquicker, and improve. It’s important for business leadersto familiarize themselves with this technology now.Realizing data valueBy the final stages of the supply chain, a significantinvestment has been made in the ingestion,transformation, and analysis of data—and now that datais both accessible and sharable, companies have newopportunities to capitalize on its value. Of course, datais important internally, but companies must now realizethat the value of data extends outside the organizationas well. From forging new partnerships to creatingnew revenue streams, or even entering new markets,businesses now have more potential than ever to realizethe true value latent in their data.When companies open up their data for external access,they can profit by leveraging external developer talent toextend their platforms. It’s a win for developers, too; theycan use these platforms to add value to their productsmuch more easily and at a higher quality than if theyhad attempted to do so on their own. For example, oneof the reasons Google Maps is so widely known andsuccessful is that more than 800,000 websites use itsdata, accessed through its API. 16Now, companies can take advantage of the opportunitiesfor data monetization—to sell data insights directly, sharethem through partnerships, or develop entire ecosystemsaround them (see the chapter “From Workforce toCrowdsource”). There are risks and rewards for each,of course, and some methods will be better suited forcertain companies and industries than others, but theyare all important strategies to consider. Kabbage, forinstance, was able to find a unique use case for datafrom UPS. As a partner, UPS provides transactionalshipping data to inform Kabbage’s health assessment ofthe companies it finances—determining those companies’access to capital. 1753ACCENTURE TECHNOLOGY VISION 2014

TREND 3: DATA SUPPLY CHAIN54Just as companies are finding new reasons for ingestingexternal data sources, there have to be companies outthere to provide them. It’s not just about selling data; it’salso about strengthening partnerships and developingecosystems around data to monetize it. Companies nowhave the opportunity to think outside of the box for newways to realize and take advantage of the true value intheir data.Adding one data supplychain, and then anotherEvery day, as the amount of data grows, so too doesthe seemingly impossible task of realizing the dataservices platform and, ultimately, the data supplychain. Unfortunately, there are no shortcuts. The pathis incredibly difficult and exceptionally long—in fact,it never quite ends. But it is also one of the mostrewarding journeys that companies can make in theirtransformation to become truly data driven.Progress becomes possible when the transformationprocess is viewed as a matter of small steps rather thanone giant leap. So, we expect that leading organizationswill start by establishing a data services platform,followed by implementing a single data supply chain for aspecific outcome. Once that’s done, they will incorporateanother—and another. It may not be easy, but it’s doable.The implications of enabling the data supply chain arehuge. In the first stage, organizations now have theopportunity to ingest new sources of data. In the second,the manipulation of data through new methods of datadiscovery adds significant value. The future of this phaseis to mask its complexity and truly embed this valuethrough cognitive computing technologies. And finally,businesses can now look externally to realize value fromdata in new ways.It’s time for business leaders to start thinking about theentirety of the data supply chain—as an end-to-endprocess that is outcome driven and fit to strategy.Today, data should be more than just a premier IT asset;it should be a premier business asset. Let’s start treatingit that way.#techvision2014

TREND 3: DATA SUPPLY CHAINYour 100-day planIn 100 days, begin to develop a comprehensive strategyaround laying the foundation for your data supply chain.• Start to build an inventory of your data, beginning withyour most frequently accessed and time-relevant data—which will be given first access to your data platform andaccelerated on it.• Identify any manual, time-consuming data curationprocesses (e.g., tagging, cleansing) for potential replacementwith machine learning algorithms.• Identify data silos within your organization (e.g., HR,finance, engineering), along with corresponding data needsthat are currently unmet across the business.• Begin to simplify/federate access to trusted data. Create astrategy for standardizing data access via the data platform.Solutions may be hybrid, utilizing a combination of traditionalmiddleware and API management, or even a PaaS offering.• Prioritize your individual data supply chains to develop aroad map for implementing the data supply chain at scale.• While building your platform, start looking outside yourcompany for external data sources that can be incorporatedto complement existing data and help lead to morecomplete insights.This time next yearIn 365 days, begin the journey to ROI by building a supplychain that is designed to drive outcomes.• Pilot an initial data supply chain targeting a single businessfunction. Once proven, each subsequent data supply chainwill build upon this initial framework.• Accelerate data on your data platform to enable thepractice of data discovery across your organization.Uncover value in your data by quickly asking questions,failing fast, and iterating to solutions.• Empower your business users to be value creators in thedata supply chain by giving them data discovery tools andtraining them to deliver insights.• Reduce the manual effort of curating data by usingmachine learning algorithms to automate the repetitiveprocesses in your supply chain.• Find a specific use case that can be addressed by cognitivecomputingtechniques. Use this focused issue as a chanceto experiment with and learn the ways in which cognitivecomputing can be applied to your organization.• Investigate opportunities to monetize your data. Theseshould include developing new partnerships or ecosystemsaround data insights, in addition to directly selling them.ACCENTURE TECHNOLOGY VISION 201455

TREND 4: HARNESSING HYPERSCALE56#techvision2014

TREND 4: HARNESSING HYPERSCALETREND 4Harnessing hyperscale:Hardware is back(and never really went away)Eclipsed by more than a decade of innovation in software,the hardware world is again a hotbed of new development asdemand soars for bigger, faster, lower-cost data centers. Doesyour IT organization understand the new developments allowingcompanies to realize the benefits of “hyperscale” systems? In thisnew world, hardware matters more than ever in transformingenterprises into digital businesses with access to unlimitedcomputing power that can be turned on and off as needed.57ACCENTURE TECHNOLOGY VISION 2014

TREND 4: HARNESSING HYPERSCALE58Why now?Rising demand for scale: Across industries, demand forprocessing at scale is surging. Businesses need reliablehardware to support the immense amounts of data processedfor predictive analytics and real-time insights.Hardware and server architecture innovation surge: Fromadvances in storage to power consumption to processorsto server architecture, infrastructure innovations such asnonvolatile memory are paving the way for faster, cheaper,and bigger hyperscale systems.Open source: Facebook’s Open Compute Project isaccelerating the adoption of infrastructure innovations bysharing those breakthroughs freely. Founded in 2011, theOpen Compute Project has already grown to more than 60official members and thousands of participants. 1Who said that everything old is newagain? He may have had a point—especially where computing hardwareis concerned.Not so long ago, every announcement of a new,multicore microprocessor or personal computer metwith fanfare—lots of buzz about clock rates and cachememory capacity. But in recent years, the prevailingsentiment has been that hardware no longer matters—that x86 servers are nothing more than off-the-shelfcommodities and all the important advances now happenin software.Fast forward to today and looking to the future, it isbecoming clear that “hardware as an afterthought” is ablinkered view. In fact, it is now a harmful view becauseit will make it more difficult for enterprises to evolve intodigital businesses.Every industry will be touched by the technologiesbeing developed for the era of “hyperscale” computingsystems—the supersized, super-scalable, and resilient datacenters pioneered by heavily data-dependent companieslike Google and Facebook. Innovations in technologies#techvision2014

TREND 4: HARNESSING HYPERSCALEsuch as low-power CPUs, solid-state data storage, andin-memory computing will benefit the performance ofall enterprises’ servers and data centers, enabling theirnext generation of infrastructure to support the digitaltransformation of their business.Unilever, a consumer goods company; Pirelli, a tiremanufacturer; and the NBA, a professional basketballleague, are not companies that would typically beassociated with one another, let alone be consideredin the same context as Amazon and Google. And yet,similar to those technology giants, these and othertraditional companies like them are faced with challengesthat hyperscale computing can help to solve: immenseamounts of data that need to be processed at speed.However, these companies had neither the scale nor thecomputing hardware to solve these challenges in-house.So they looked to hyperscale hardware appliances to helpthem perform at the speed of their nonstop businesses.For example, Unilever, Pirelli, and the NBA are all usingSAP HANA to perform real-time analytics on hugedatasets to gain insight and competitive advantage. 2HANA gives them the capabilities of hyperscale thatAmazon and Google enjoy in a single appliance insteadof an entire data center. Others are using solutions fromOracle, IBM, and Teradata.These data challenges exist across every industry andfor companies of all sizes. This is why every CIO willagain be tasked with understanding the advantages andtrade-offs of hyperscale systems and big data appliances,as well as the opportunities and risks behind choosingwhich hardware will be used to inform and optimize theirdigital business.Inside hyperscaleBefore we examine the hardware innovation that is beingpropelled by hyperscale developments, it is important toexplain what the word hyperscale means. Accenture usesthe term to describe not just the physical infrastructureof giant and distributed systems that support the datacenters and provide companies such as Google, Amazon,and Facebook with the computing power to deal withvast volumes, variety, and velocity of data, but also theability to scale computing tasks to achieve performancethat is orders of magnitude better than the status quo.Hyperscale data centers consume storage, bandwidth,memory, and computing cycles on a scale unimaginableto most. They make pragmatic use of the latest hardwareinnovations, but never at the expense of scalability.ACCENTURE TECHNOLOGY VISION 201459

TREND 4: HARNESSING HYPERSCALE60What does this scale actually mean? In the world ofhyperscale data centers, there are very few players.Google is the largest with well over a million servers;Microsoft also has more than a million. Then comesAmazon with about half a million, and Facebook andYahoo with a couple hundred thousand each. 3 In termsof power consumption, Google’s global operationscontinuously draw 260 million megawatts of power—roughly a quarter of the energy generated by a nuclearpower plant. 4Not only are these companies operating on a scalethat defies most people’s imaginations, but theservices provided by these hyperscale data centers areexpected to be up 100 percent of the time—not just99.99999 percent of the time. This is the era of “always on.”As a case in point, note that companies such as Amazon,Facebook, and Google are not just responsible for theirown bottom lines—they are increasingly responsible fornumerous other businesses’ survivability as well. Today,hyperscale data centers are pivotal to an enormousecosystem of companies and organizations that rely onhyperscale platforms as a necessary component of theirbusiness models.But the demands on hyperscale computing do not stopwith the need to be always on. It is equally critical tobe “always optimized”—application-specific computingtechnologies are available to solve increasingly specializedtasks. For example, hyperscale data centers with theability to horizontally scale worldwide with commodityservers are appropriate for a website. But complexcalculations for applications such as 3D rendering, DNAsequencing, and cryptography are more effectively andefficiently handled by a system of servers with advancedgraphics processing units (GPUs)—specialized processorsoptimized for parallel processing of very large floatingpointnumber calculations. GPUs are so effective at thesetasks that application-specific integrated circuits (ASICs)are starting to become available that offer orders ofmagnitude more processing capability than server farmsfull of advanced GPUs with dramatically less energyconsumption and a smaller physical footprint.There are two main reasons why this hyperscale trend isimportant to the enterprise. First, hyperscale is drivingsignificant hardware innovation that will put businessesat a nexus where decisions on public or private cloud, as#techvision2014

TREND 4: HARNESSING HYPERSCALEwell as commodity versus specialized hardware, becomesignificant differentiators. Second, these hyperscalesystems are no longer limited to large Internet companies.Utilities, oil and gas, and automotive are only a fewexamples of industries that are quickly approaching thehyperscale level of interconnected devices, sensors, anddata centers. Utilities now have networks of integratedsensors and smart meters that rival those of communicationcompanies. Ford, GM, and Toyota are building vehicleswith hundreds of sensors, telematics, and real-timeconnectivity. As all of this information begins to becollected every day, every hour, and, in extreme cases,multiple times per second, the systems required to storeand analyze this data at speed will soon demandhyperscale solutions. Take gas turbines at power plants,for example. “We’re almost putting a data center on agas turbine,” GE said recently in reference to the hundredsof sensors the company is placing on those machines tocapture data. If those sensors, combined with robust dataanalysis, are able to improve efficiency by just 1 percent,that could save nearly $6 billion a year. 5“Not only are these companiesoperating on a scale that defiesmost people’s imaginations, but theservices providzed by these hyperscaledata centers are expected to be up100 percent of the time—not just99.99999 percent of the time. Thisis the era of ‘always on.’”61ACCENTURE TECHNOLOGY VISION 2014

TREND 4: HARNESSING HYPERSCALE62The innovation behindthe hyperscale pushThe rapid growth of hyperscale systems has sparkeda renaissance in hardware innovation from which allbusinesses stand to benefit. While the tech-savvy readercan jump into a bit more depth around the technologyin the “Driving innovation in hardware” sidebar in thischapter, everyone should understand that innovationsaround processors, storage, and specialized hardwareare proving to be opportunities for tangible benefits forthe business.Putting limits on design specifications, such as energyconsumption, has driven innovation. Hewlett-Packard(HP) drew inspiration from mobile processors for itsMoonshot servers, which operate with up to 89 percentless energy than traditional servers. 6 In a similar focus onenergy efficiency, Facebook reduced energy consumptionby 38 percent in its Prineville, Oregon, data centercompared with existing facilities by using technologiesdeveloped as part of the Open Compute Project. 7Storage advances are providing enterprises with freshways to access and manipulate data faster. Flash storagearrays are becoming the norm for e-commerce sitesand financial services firms with trading platforms. 8E*Trade uses EMC’s XtremIO storage array to maintainexceptional performance for customers while performingdata maintenance tasks in parallel with no performancedegradation, giving E*Trade the ability to scale operationsand support new features for users. 9And innovations in server architectures are offeringmore options to match specialized infrastructure tothe applications being run, helping CIOs identify theimportant requirements of each application, whetherthat be performance, scale, or cost. Illustrating theadvantages that specialized hardware can provide, SAPhas 28 customers that now run analytics jobs in HANA10,000 times faster than they did previously. Threeof these customers have experienced 100,000 timesimprovement. One of those three, Yodobashi, a Japaneseelectronics retailer, took a three-day batch process toanalyze loyalty customers down to two seconds, enablingreal-time couponing at the point of sale. 10#techvision2014

TREND 4: HARNESSING HYPERSCALEAlthough the details of these innovations may not needto be completely understood outside of the IT department,it is important to recognize their potential to transformbusiness processes and strategies. In doing so, it’sparamount for companies to understand the advantagesand trade-offs of these new tools as they look to procure,build, and use their next generation of IT systems.Adopting an open-sourceview of hardwareFacebook’s Open Compute Project is anotherdevelopment that is causing significant discussion withinCIO and other technology circles. The initiative involvesopenly sharing hardware innovations, following themodel associated with open-source software projects.Members participating in this exchange of technologyinnovations include Goldman Sachs and Fidelity, which,in conjunction with AMD and Intel, have been reviewingnew motherboard designs to incorporate in their owndata centers. 11 The payoff for enterprises can be significant.For instance, Facebook designed and built its own serversand software, claiming that it can build its data centersat one-fifth of the cost of a traditional data center. 12Asking hard questionsabout hardwareThe relationship between hardware innovation and risingdemand for cloud services is raising some new questionsfor IT leaders. First and foremost, over the next fiveyears, every IT department will consistently be facedwith the choice between leveraging external cloudsand building computing infrastructure on premise. Tomake matters more complex, on-premise choices arecapital intensive and increasingly specialized. There arecomputing appliances that give hyperscale capabilitiesfrom a bevy of manufacturers—SAP HANA, OracleExadata (now enhanced with Exalytics), IBM PureDataSystems (an evolution from Netezza), and Teradata areall top-tier options. Choosing the cloud does not removethis complexity either. Not all clouds are equal; there aremany varieties from which to choose. How CIOs makethis choice will depend significantly on the systems theyare looking to build and their demands.63ACCENTURE TECHNOLOGY VISION 2014

TREND 4: HARNESSING HYPERSCALE64One company may need highly resilient services thatfluctuate little during the workday, but it may demandmore of those services at the end of every month.Another may require real-time decisions on vastquantities of data.The array of decisions now required is reflected in thesheer number of choices to be made when configuringan instance on Amazon Web Services (AWS). Amazonprovides a wide range of configuration options to suitall sorts of business needs—ranging from “computeoptimized” (with a high ratio of CPU to memory andlowest cost per virtual CPU of all Amazon’s instancetypes) to “storage optimized” (suited to applications withspecific input/output and storage capacity requirements).So the decisions that IT leaders make about their datacenters and cloud services must consider the underlyinghardware. How will this provider’s use of low-powerCPUs in its data centers affect operational costs over thecontract term? Can our applications run on low-powerCPUs, or would GPUs be more efficient for computation?Does the code for our critical business insights need tobe rewritten to take advantage of these new technologies?To what extent can the data center scale up for apharmaceutical client, say, if the client starts puttingmost of its clinical trial simulations in the cloud? Do ourbusiness requirements mean that we are better off withflash, in-memory, or hard-disk storage? The choice ofhardware depends very much on what the specificapplication needs are, and what the usage patterns willlook like.On this digital journey, solutions must be tailored forindividual use cases. Every organization is going toface a set of heterogeneous requirements that willbe best served using the correct recipe of commodityversus specialized hardware and private versus publiccloud architecture. This means that for the foreseeablefuture, enterprise infrastructures will be, by necessity,hybrid solutions—weaving hyperscale cloud, on premise,specialized hardware, and an enterprise’s existing systemsinto a computing fabric that serves more parts of thebusiness in more demanding, reliable, and scalable waysthan before.#techvision2014

TREND 4: HARNESSING HYPERSCALEAdopting hyperscaleIn a dramatic shift of IT strategy, companies areleveraging hardware innovation on a piecemeal basis,with some companies going so far as to build their ownhyperscale systems as a competitive differentiator. GMis one such company that is embracing the hyperscaleconcept and building out its own mega data centers. Thisis not just another private cloud, according to GM: “ITisn’t here to manage daily operations. It’s a strategic toolto drive business forward.” 13With that in mind, the automaker is setting up its ownhyperscale environment starting with the first of twodata centers located in Michigan. Moreover, GM hascancelled its multibillion-dollar outsourcing agreements,believing that it needed to bring these strategiccapabilities in-house to be closer to the businessdecisions needing to be made in a rapidly changingindustry. 14GM’s goal is to accelerate new product introductions,especially when it comes to keeping up with the growthof on-board vehicle electronics. Imagine fleets of vehicleswith thousands of sensors sending gigabytes of databack to the data centers every second of every day fromevery corner of the world—all while maintaining highreliability levels in its data centers. Downtime at one ofits factories can cost GM $1 million per minute, givencurrent just-in-time inventory practices. 15And GM is not alone. By 2018, some 10 million BMWswill be connected, asking for, and receiving more than1 terabyte of data every day. 16 More large organizationsare interested in building their own high-efficiency datacenters. Gartner estimates “large data centers”—thosewith more than 500 racks—will account for 29 percentof all data hardware revenue by 2017, up from 22 percentin 2011. 17Even if CIOs do not have an appetite for building theirown infrastructure, companies still need to examinetheir path to becoming digital enterprises and determinewhether they have the infrastructure and skills neededto support it. There is no set path on how to grow tomeet the hyperscale challenges that businesses are goingto face. But forward-looking companies are seeing theadvantages of hyperscale technology.65ACCENTURE TECHNOLOGY VISION 2014

TREND 4: HARNESSING HYPERSCALEAt a minimum, they are seeing an opportunity to drivedown the operational cost of running their data centers.As such, high-performance companies are increasinglyrecognizing that hyperscale systems are a vital part ofbecoming a digital business. To get started, technologyleaders need to ask themselves, “What could our businessdo with unlimited compute power that can be turned onand off as needed?”66#techvision2014

TREND 4: HARNESSING HYPERSCALEYour 100-day planIn 100 days, make sure your organization is informed aboutthe available hyperscale technology options.• Ensure that your IT organization is aware of consortiumsand/or is testing the benefits of the latest hardwareinnovations. Identify those that are most important to yourbusiness.• Identify your data storage needs and the magnitude ofdevices producing data in your network (including sensors,smart meters, devices, and data centers). Forecast theirexpected usage based on one-year and three-year businessgrowth strategies.• Create a plan that allows key data assets to be portableacross architectures.• Explore participation in open-source communities such asthe Open Compute Project to leverage emerging hardwareinnovations.This time next yearIn 365 days, be prepared to have your IT road map includehyperscale technologies. Have the knowledge to make theright investments and act.• Update models of your digital business’s most demandingcompute processes to understand the advantages, tradeoffs,opportunities, and risks of hyperscale hardware choices.• Create a hyperscale task force. Have it include hyperscalehardware during infrastructure planning.• Build the reference architectures for hyperscale workloadsand evaluate the applications that are deployed to hyperscale.Maintain hybrid hyperscale deployment strategies thataccount for improvements in offerings across SaaS, PaaS,and hardware to support varying enterprise activities.• Investigate how edge-device computing can makeapplications more efficient and provide a vehicle formachine-to-machine interaction.67ACCENTURE TECHNOLOGY VISION 2014

TREND 4: HARNESSING HYPERSCALE68sidebarDriving innovation in hardwareInnovation in three corners of the hardwareworld is driving hyperscale:ProcessorsHistorically, complex instruction set computing (CISC)–based servers ran on high-performance x86 processors.This helped to rein in costs and maximize compatibilityfor mass-market applications. Over the last decade,however, the massive consumer demand for mobilephones has driven radical progress in the capabilities oflow-power processors—largely to the credit, and benefit,of ARM Holdings. Low-power processors give data centerdesigners an opportunity to address data center powercosts which, in many cases, have risen to be 30 percentor more of operational expenditures. Calculations showthat the cost of power to run a server over its lifetimewill very often eclipse the cost of the server itself. 18Innovation is not limited to power consumption. ARMlicenses its chip designs, creating a rich ecosystemof chip manufacturers and software developers ableto pursue niche opportunities. Benefiting from thisecosystem, HP recently announced that an upcomingmodule for its Moonshot server line will use a newversion of Calxeda’s ARM-based EnergyCore chips. HPclaims that this new module will offer nearly double theperformance and four times the amount of memory permodule compared with the previous generation of lowpowerprocessors. 19Complementing these low-power CPUs, the massivelyparallel architectures of graphics processing units (GPUs)are ideal for big data analytics. As a consequence,x86 servers are being replaced by servers with lowpowerARM processors, using GPUs as accelerators orcoprocessors. This allows massively parallel database(MapD) technology to realize tremendous speed gains bystoring the data in the onboard memory of GPUs insteadof CPUs, as is typical. Use of a single high-performanceGPU can make data processing as much as 70 timesfaster than a conventional CPU. 20 That is why 53 of thetop 500 supercomputers now use GPU accelerators orcoprocessors. 21#techvision2014

TREND 4: HARNESSING HYPERSCALEStorageOrganizations must now store more data and getmore insights more quickly from the data. With dataprocessing capacity radically improved, the bottleneckbecomes the ability to store and retrieve large volumes ofdata as rapidly as needed.Traditionally, data has been stored either in limitedvolatile memory, which provides the fastest access, or ondisk, which provides much more storage but is far slower.Companies could have their data fast or cheap, but notboth. Now, though, innovations in flash memory aremaking this trade-off unnecessary.Vendors such as Fusion-io and Violin Memory areproviding flash memory products that address hyperscaleissues such as the need for low-latency, low-powerconsumption; a smaller data center footprint; and ahigh degree of resilience. These companies are providingcase studies and performance numbers that indicate thepotential for application performance to increase by atleast 25 times, and sometimes by as much as 40 times. 22“Innovation at the server and datacenter levels is causing a bifurcationbetween the types of processingcapabilities being offered at hyperscale:those optimized for general-purpose,Web-scale computing and thoseoptimized for specialized tasks.”69ACCENTURE TECHNOLOGY VISION 2014

TREND 4: HARNESSING HYPERSCALE70That means that CIOs now have new options for solvingtheir scalability challenges. They can start asking differentquestions: Is tuning application performance a good useof expensive developers? Will future applications runmost efficiently on commodity hardware that scaleshorizontally? Does specialized hardware better serve thebusiness needs? What are the trade-offs? What is thetotal cost of ownership?ServersInnovation is also occurring at the server and data centerlevels. In fact, this innovation is causing a bifurcationbetween the types of processing capabilities beingoffered at hyperscale: those optimized for generalpurpose,Web-scale computing and those optimized forspecialized tasks.For general-purpose computing, data center designersnow have two options to build at hyperscale: lowpowersystems such as HP’s ARM-based Moonshot orcommodity x86 servers. In the most extreme cases ofthe latter, Google, Amazon, and Facebook are all sourcingcustom server designs from OEMs, optimizing for theirspecific operational objectives and computing andstorage needs. Companies looking to build or expandtheir own commodity x86 data center architectures mayfind benefit in joining the Open Compute Project. Forthose looking to radically improve operational costs,low-power systems make sense. While these low-powersystems can be configured to run standard Web orbig data applications such as Hadoop, they are notcompatible with software written for x86 architectures.However, the operational savings may be worth thedevelopment investment; HP claims that typical datacenter operational costs can be reduced by as much as77 percent with Moonshot. 23With specialized tasks such as analytics on largedatabases and computationally intense tasks such asweather forecasting and DNA sequencing, generalpurposehyperscale systems are insufficient to deliverresults at speed and scale. Database software companiessuch as SAP and Oracle have risen to the challenge bybuilding in-memory computing appliances that giveorders-of-magnitude performance improvements onbig data analytic challenges. Existing customers of SAP#techvision2014

TREND 4: HARNESSING HYPERSCALEand Oracle will find streamlined transition paths withHANA and Exalytics, respectively. These platforms andothers offer the benefits of hyperscale in a configurationoptimized for on-premise installations.Moving to a greater variety of complex data processingworkloads, however, calls for supercomputing solutions.Similar to the in-memory-computing appliances, theseindividual systems do not scale, but they do providecapabilities that go beyond what hyperscale can deliver.Amazon understands this, which is why its ElasticCompute Cloud (EC2) includes two of the world’s top 500supercomputing clusters (numbers 64 and 165). 2471ACCENTURE TECHNOLOGY VISION 2014

TREND 5: THE BUSINESS OF APPLICATIONS72#techvision2014

TREND 5: THE BUSINESS OF APPLICATIONSTREND 5The business of applications:Software as a core competency ina digital worldThe way we build software is changing. Mimicking the shift in theconsumer world, organizations are rapidly moving from enterpriseapplications to apps. Yes, there will always be big, complexenterprise software systems to support large organizations, and itwill still be necessary for IT developers to keep customizing thosesystems, providing updates, patches, and more. But now, as largeenterprises push for greater IT agility, there is a sharp shift towardsimpler, more modular, and more custom apps. The implicationsare significant for IT leaders and business leaders alike: they mustsoon decide not just who plays what application development rolein their new digital organizations but also how to transform thenature of application development itself.ACCENTURE TECHNOLOGY VISION 201473

TREND 5: THE BUSINESS OF APPLICATIONS74Why now?Digital transformation of enterprises: IT applications havebecome the primary driver for growth and differentiation forenterprises.Accelerated pace of IT change: The increasing push torapidly deploy new technology is increasing the pressure on ITto provide a faster way to develop and deploy the applicationsthat are driving corporate digital strategies.Maturation of application platform providers: PaaSplayers are offering ready-made data service platforms,with sets of services already connected and instant setsof app families available. Tibco, Apigee, and Salesforce arealready offering solutions that provide the foundation for acustomized enterprise app experience.Rising consumer and user expectations: Customers andemployees are looking for consumer-grade experienceseverywhere. They are pressing IT to give them, in the workplace,the kinds of low-cost, accessible, and often intelligent appsthey use every day on their own mobile devices.A large corporation that is as agile asits global customers need it to be?There’s an app for that.But there are big changes in how the application iscreated. Anyone with a smartphone can see the signs.In the consumer world, application development hasbecome the province of a new generation of developerseager to strike gold by creating the next Angry Birdsor Evernote app for the mobile phone—applicationsthat are smaller and simpler to use. That developmentapproach is now moving into the enterprise, creating newapplications for mobile, Web, and desktop platforms alike.The more quickly businesses can create and launch newapplications in today’s turbulent markets, the better theycan innovate, collaborate, improve customer experiences,and enrich personal interactions.The drivers of this change are not hard to find. Eager forrelief from some of their biggest pain points—especiallytheir systems’ lack of agility—business leaders have beenpushing for software that is far nimbler than the legacysystems they’ve relied on for decades. They have beenpressing IT to give them, in the workplace, the kinds of#techvision2014

TREND 5: THE BUSINESS OF APPLICATIONSlow-cost, accessible, and often intelligent apps they useevery day on their own mobile devices.IT is responding. Leading IT groups are building softwareplatforms and architectures that essentially separate thebig back-end services from the applications that usersinteract with. The outcomes are a win for both the IT sideand the business side: IT gets breathing room to buildsolid foundations for complex systems; the business getsplatforms that allow for an increased focus on customsoftware that enables rapid iterations to tailor solutionsto fast-changing market conditions.There’s another big consequence of the shift toward anenterprise app world: business functions are partneringwith technology organizations to assume joint ownershipof the new agile applications. The back-end services—from data centers to networks—still fall squarely under IT,but in more and more organizations, the lines are blurringas the business takes a more active role in many aspectsof front-end applications. These are not maverick activities:they are happening with the say-so of IT in thoseorganizations. In fact, savvy IT leaders are deliberatelypartnering with the business side not just to enable butto encourage them to take on some of these roles.In a sense, the building of these custom agile applicationsis becoming a hallmark of the new digital enterprise. Thequestions are now as much about resources as abouttechnology. How does IT help the business side to acquirethe necessary skills and development mind-set? Howdo IT leaders manage the balance between enablementof the business and building of a stronger technologybackbone to support a more agile enterprise? Thesequestions will be high on the agenda of every CIO in thenext few years.The push towardapplication ecosystemsIn the consumer world, applications have the luxury ofbeing relatively self-contained. Shazam, for example,excels at capturing a song heard playing in a café, butthe user doesn’t need it to manage her playlist; she hasanother app for that. In the enterprise world, however,the problems being solved are much more complex.They often involve multiple applications to run intricatebusiness processes that may span multiple time zones,several countries, and thousands of employees.75ACCENTURE TECHNOLOGY VISION 2014

TREND 5: THE BUSINESS OF APPLICATIONS76To attack these big enterprise-level problems, then,something more than a single, nimble app is needed.The way forward is to think in terms of the application’sability to connect with other applications. In fact, thepush is toward libraries or “ecosystems” of applicationsthat can, while still being individually simple and agile, bebolted together to tackle the most challenging problems.To do this successfully, businesses require more than justseparation of the front-end applications from the backendservices. They need software platforms to drive newdevelopment ecosystems.We see this trend most clearly in the movement towardenterprise app stores. Research company Gartner predictsthat by 2017, a quarter of all business enterprises willhave an app store for managing corporate-sanctionedapps on PCs and mobile devices. 1 The top performers arealready leading the way: Accenture’s High Performance ITresearch confirms that 54 percent of high-performanceIT groups have deployed a mobile-enterprise app store,compared with just 22 percent of other IT organizations. 2China Eastern is a good case in point. In developing itsown app store in 2012, the airline wanted to drive mobileapp adoption across the company; its rationale was thatfast-growing mobile phone use in China would help drivegrowth for the airline and also improve productivity andoperational efficiency, especially among the 50,000 ofits employees who use their mobiles for work. The airlinemakes good use of mobile apps in areas such as aircraftmaintenance, employee feedback services, and mobileoffice automation. China Eastern has also bought 2,500iPads for in-cabin services, helping it earn a reputation asan early adopter of the latest mobility technologies. 3But the real power of enterprise apps lies in how theyare amalgamated—combined and connected in ways thatcreate a customized system capable of handling largerbusiness tasks. To some extent, what’s old is new again:the rise of these application ecosystems is effectivelycashing in on the appeal of service-oriented architecture(SOA), the technology of a decade ago that promisedbusiness process owners the ability to rapidly iteratebusiness processes and user experiences by reusing andreorganizing small pieces of functionality—now known asapps—and by providing ways to string them together.#techvision2014

TREND 5: THE BUSINESS OF APPLICATIONSFor example, Japan Post Co. is one of Saleforce.com’s topfive customers and has been using apps on Force.com’splatform to streamline business planning and compete innew markets. They created 15 custom apps to providebusiness process visibility in selling services and productsfrom three companies as a retail outlet: Japan Post Bank,Japan Post Insurance, and Japan Post Service. With 24,000post offices nationwide, it efficiently serves 14 billionmail packages annually and more than 6 million insurancepolicies for upwards of 100 million clients while achievingflexibility and a reduced total cost of ownership. 4Software platform:middleware redux?The proliferation of app stores speaks to the consumptionand use of small, nimble apps. But the other importantstory is how IT systems are architected for this newapp world. The new direction is to separate applicationsfrom the back-end systems that support them. Imaginetaking the look, the flow, and the experience of afinancial reporting system and disconnecting it from thecomplex financial models and SEC regulation behind it.The resulting system would allow developers to rapidlyupdate how users interact with it; it would be easier,for instance, to get up-to-date cost projections withouthaving to touch the underlying financial models—traditionally a burdensome effort when such updatesare required.Software vendors and enterprises alike are alreadychanging how they architect their systems to enable thisseparation. In essence, middleware is being resurrected asthe “software platform”—a way to present data servicesthat can make it easier to find modular apps that willperform a particular business function, and to enablemodular apps to combine, like puzzle pieces, into “systems”in order to implement more complex business activities.Such systems are more flexible and provide a frameworkto create a custom, on-demand solution faster.Companies such as Mashery and Tibco Software—thelatter a longtime standard bearer in the middlewareworld—are pushing this data services space withtheir platforms to provide an integrated developmentenvironment to rapidly create, orchestrate, and integratemodular services and business applications. Helping USAToday open and manage its application-programming77ACCENTURE TECHNOLOGY VISION 2014

TREND 5: THE BUSINESS OF APPLICATIONS78interfaces (APIs), Mashery has expedited the mediacompany’s ability to negotiate and execute newpartnerships with distributors and device makers so thatdevices from iPads to Samsung refrigerators are nowconnected. 5 Similarly, Tibco’s Silver Fabric cloud platformis designed to give enterprises the ability to quicklyprovision their own app stores.Other companies such as Apigee are stepping forwardwith API management tools that help make it easier forlarge enterprises to extend their reach with mobile apps,create new products with partners and developers, andmore. Live Nation Entertainment uses Apigee’s EnterpriseAPI Platform to grant internal developers and trustedpartners and clients the ability to extend their ticketingfunctionality to mobile devices and provide additionalfunctions such as event calendaring and live eventapplications. As a result, Live Nation’s mobile transactionvolume has more than doubled in the last year. 6And the Platform-as-a-Service (PaaS) players areoffering ready-made data services platforms, with setsof services already connected and instant sets of appfamilies available. Salesforce.com’s AppExchange is auseful case in point. Using its platform as a base for howthese applications are bought, built, and woven together,Salesforce.com has created its own flexible ecosystemthat represents the next generation of enterprise-builtand -customized systems and solutions. Recognizingthat more than 40 percent of their transactions stemfrom API calls, the Salesforce1 Platform is expanding thisecosystem, opening up ten times more APIs and servicesfor developers to build upon. 7 And acknowledging theincreasing value of management required to exposeAPIs at scale, Microsoft recently acquired Apiphany tointegrate API management into Windows Azure.Importantly, IT leaders should not strive for the holy grailof software platforms; there is no such thing. It’s notpossible to use a single platform to handle every businessrequirement. And don’t assume that cloud will always bethe answer either. According to Accenture’s latest HighPerformance IT study, only 7 percent of applicationscurrently reside in the cloud, and survey respondentsexpect that figure to approach 33 percent by 2020. 8In practice, companies will need to adopt a hybrid mindset,with different platforms, local and in the cloud,for different sets of business needs. Abstraction of theback-end services will have to be a step-by-step process,starting with the highest-priority business needs. Perhaps#techvision2014

TREND 5: THE BUSINESS OF APPLICATIONSthe organization starts out by developing a platformfor its consumer data and then eventually creates aplatform for its ERP data. The more services that areavailable, the greater the functionality that the apps willbe able to leverage and the more diverse and innovativethey can become.Leveraging separationto amplify impactOf course, nobody is suggesting that ERP, CRM, orlarge enterprise financial systems are about to becomeobsolete. Rather, it’s important to recognize that IT hasentered a new stage in which constant change is thenew normal—one in which IT not only supports businessapplications but also enables the business to play amore active role in those front-end apps. As businessusers start to embrace technology as a way to drivetheir business strategies—everything from tools foranalyzing consumer sentiment on social media sites topilots of new pricing models—IT must start opening upthe systems, tools, and processes to allow business usersto drive these initiatives forward themselves. This letsthe enterprise move from the paradigm of a backlogof IT requests from the business to one where IT isempowering the business to experiment, innovate, anddrive its strategies.Enabling the business is just the first phase; businessleaders will recognize that extending these servicesto business partners has the potential to acceleratebusiness imperatives. For example, AT&T opened itsphone activation APIs so that its partners could leveragethem fully—building apps on top of them, for instance.The APIs are now available as Web services, and the onboardingprocess can now be executed in a matter ofdays or even hours—resulting in a 1,500 percent increasein the usage of the APIs, not to mention a boost inphone activations. 9Other companies have taken this “ecosystem” conceptfurther. Positioning their software as an open platform,Facebook and Salesforce.com are among the companiesthat are using the developer community at large astheir app innovation engine. At last count, Facebookhad more than 10 million apps on its platform, wheremore than 25 percent of its annual revenue comes frommobile app install ads. 10 It’s not just tech companies thatare taking this approach; General Motors has openedACCENTURE TECHNOLOGY VISION 201479

TREND 5: THE BUSINESS OF APPLICATIONS80up its OnStar APIs to the public. And, based on aninnovative application developed on top its APIs, thecompany is now partnering with the start-up RelayRidesto enter the emerging ride-sharing market. In doing so,GM is threatening to disrupt both car rental and taxicompanies—business sectors well outside of automobilemanufacturing.This amplification effect is one of the most powerfuladvantages for the new approach to enterprise apps. Inthis way, the organization can be more innovative andreact more rapidly. It is not just about applications beingeasier to build—it’s about the numbers of apps that canbe built rapidly and concurrently.Adding intelligenceto applicationsAs more and more apps are driven by the business, theycontinue to evolve: they are becoming more intelligent.Apps will be better able to sense and respond to users’context, their history, and the world around them. TimeWarner, for example, has embedded intelligence in its settopboxes so that more-relevant content can be pushedto users based on their TV-viewing patterns. 11 In anotherexample, a mobile app called uChek allows users to spotpotential problems such as kidney and liver issues aswell as diabetes using a picture of a chemical test strip. 12And Amazon has announced the release of an analyticsservice for iOS, Android, and Kindle Fire developers thatwill allow any developers to embed more intelligence intotheir apps.Far-sighted business users are also eager to make theirapplications more useful and more contact-aware—whichis why there’s a growing push to embed analytics inbusiness applications. Analytics-rich applications canenable users to answer many more of their own dataquestions. They can effectively become their own datascientists. Tableau Software’s simple data visualizationcapability gives customers like Unilever the opportunityto answer complex business questions very simply, thusrekindling people’s interest in asking data questions. Thishas created a resurgence in analytics and discovery ofsimple solutions for previously complicated situations,such as determining the fastest-growing products fora very specific demographic. 13 When users can extractintelligent insights from their data in short order, they#techvision2014

TREND 5: THE BUSINESS OF APPLICATIONScan take intelligent action much more quickly. At thesame time, the organization’s actual data scientists arethen freed up to dig into higher-level strategic questions.The competenciesneeded to support aworld of appsAs enterprises get more involved with the new businessof applications, corporate leaders—IT and businessexecutives alike—will need to re-examine the skillsand organization structures that must be in placeto support the new arrangement. Business will bemore aware of technology and its opportunities, andIT will need a better understanding of the strategicbusiness imperatives that the technology will drive. Astechnology and business strategies become inseparable,multidisciplinary teams will become the norm. Userexperience skill will become a necessity to drive theadoption of every new application. And as the datacollected from these applications is dispersed across theorganization, scarce resources, such as data scientists,will need to be built up through hiring and training.But it’s not just tech skills that will be in demand. Unlikemany of the packaged software solutions, the businessprocesses are not set in stone in a world of apps. Thisnot only provides a big opportunity to move beyond justtechnology innovation, it also enables innovation of theunderlying processes and adds emphasis to the roles ofthe business process orchestrator and program manager.Apps also have different sets of expectations interms of maintenance. Agility implies iteration. In atraditional software world, deployment is the end of thedevelopment cycle. In the app world, deployment is justthe first iteration. The expectation is that the applicationswill essentially be “forever beta”—always evolving.Crucial juncture forIT and businessOver the next few years, we expect to see some strikingexamples of business groups that are much moreengaged in the lifecycle of “their” own front-end apps.We anticipate something of a resurgence in customdevelopment as leading companies view it as their bestoption for pursuing the objectives of a digital business.81ACCENTURE TECHNOLOGY VISION 2014

TREND 5: THE BUSINESS OF APPLICATIONS82Your 100-day planIn 100 days, begin to develop a comprehensive strategy thatwill lay out the foundation for enterprise app development.• Appoint a digital champion to coordinate development ofyour app strategy across organizations in your enterprise.• Determine your ability to enable cloud and mobile appsagainst your existing SOA, API Management, and PaaSinvestments. Based on this evaluation, start preparing astrategy to separate your back-end services from frontendapps.• Begin the design and architecture of an enterprise appsstore for the distribution of mobile and desktop applications.• Start creating a list of enterprise-level apps to be developed.Work cross-functionally across business units to prioritizethe items on your list.• Start planning a pilot for your highest priority apps that willdeploy in conjunction with the enterprise app store. Aim tovalidate the readiness of your app production environment.• Review and begin to update your app process and appgovernance strategy. Prepare to shift toward a hybrid buyand build environment, where your IT buy for front-endfunctions will decrease.And we are confident that we will observe more andmore CIOs and IT leaders sitting down with their businesscolleagues to discuss how they can help facilitate thenew application development trend.For many, though—and particularly for IT—this is acrucial juncture. The immediate task is to fully grasp theimplications of the evolving enterprise app environment.The second task is to understand the many benefitsthat come from having the business side assume largerroles in driving new agile applications. The next task isto establish the boundaries of the new roles: the extentto which IT focuses increasingly on back-end systemdevelopment or partners with the business on the frontend—or both.And the last task? It is to dive into those roles withrelish—knowing that they will surely help the company tobecome a truly digital business.#techvision2014

TREND 5: THE BUSINESS OF APPLICATIONSThis time next yearIn 365 days, begin the process of implementing the tools andservices to enable the development and distribution of smaller,more agile applications.• Begin the process of abstracting your front-end functionsfrom your back-end services. Start with the consumerapplications and services that support the applicationsgoing through the most lifecycle iterations.• Update your app governance strategy to support an agilemethodology.• Re-evaluate your testing methodology to make sure youare not disrupting it when you shift toward a continuoustesting methodology.• Pilot the enterprise app store. Utilize the first set of highpriorityapps to test the functionality of both the app storeand the applications themselves.• Based on your pilot results, create a multi-year road mapfor partnering with the business to build and deliver theremainder of your high-priority apps. Take into accountthe expected rapid change of pace for IT. Start small butscale fast.• Identify potential partners to whom you can open up yourapplication platform and development tools.• Pilot the deployment of a partner-developed application ontop of your application platform.• Evaluate the possibility of opening up specific APIs, and/or pieces of the application platform more broadly to thedeveloper community.83ACCENTURE TECHNOLOGY VISION 2014

TREND 6: ARCHITECTING RESILIENCE84#techvision2014

TREND 6: ARCHITECTING RESILIENCETREND 6Architecting resilience:“Built to survive failure” becomes themantra of the nonstop businessIn the digital era, businesses must support wide-ranging demandsfor nonstop processes, services, and systems. This has particularresonance in the office of the CIO, where the need for “always-on”IT infrastructure, security, and resilient practices can mean thedifference between business as usual and erosion of brand value.The upshot: IT must adopt a new mindset to ensure that systemsare dynamic, accessible, and continuous—not just designed tospec but designed for resilience under failure and attack.85ACCENTURE TECHNOLOGY VISION 2014

TREND 6: ARCHITECTING RESILIENCE86#techvision2014Why now?Digital transformation of enterprises: Transforming to adigital business implicitly increases a company’s exposureto risk through IT failures. More business processes areinterconnected and automated, all of which become potentialpoints of failure. The average cost of data center downtime byminute has risen by 41 percent since 2010. 1Increased cyber threats: It’s not just about gaining access tosystems; cyber criminals are also trying to bring them down.Denial of service attacks are increasing in frequency and size.The number of attacks has increased by 58 percent in thelast year. 2Increased IT complexity: More systems are being integrated,and continuous improvement is becoming the IT norm.But constant change to increasingly complex systems isintroducing more risk than ever before.The expectation of “always on”: In a digital world, whetheryour system is under attack, hit by a storm, or just beingupdated, the expectation is that it always works.Netflix loves to fail.Not by delivering movies late, by overbilling customersor in any of the other ways that the video streamingcompany could fall short. Instead, its engineers tryto find fault with their own IT systems—deployingautomated testing tools that they refer to as a SimianArmy to deliberately wreak havoc in unpredictable butmonitored ways. 3Why? Because Netflix’s engineers know that what doesn’tkill their company makes it stronger. Netflix is not alone;these practices were pioneered at Amazon a decadeago and have seen adoption at the likes of Flickr, Yahoo,Facebook, Google, and Etsy.Those companies’ technology chiefs understandsomething that IT leaders everywhere must grasp:failure is a normal operating condition. It must beanticipated, accommodated, and designed into ITsystems. Practitioners of these “game day” strategies—when days are set aside months in advance to performinternal failure testing, with dozens of staff on hand torespond to incidents—regularly find latent defects intheir systems, log hundreds of bugs, and continue to testagainst the repaired defects in future game days.

TREND 6: ARCHITECTING RESILIENCEThis continuous improvement strategy involves morethan just ensuring that systems have high availability,a condition that still allows for downtime, howeverminimal. Today, the idea is no longer about designingfor “five nines” (99.999 percent) uptime; it’s aboutsupporting the nonstop business—literally 24 hours a day,365 days a year. There can be no exceptions: if systemsare to be as nonstop as businesses need them to be,they can no longer be designed just to specification orengineered to handle only particular incidents. They mustbe designed to work under failure and under attack.The rationale is simple. As organizations migrate towarddigital, every aspect of their business is becomingincreasingly interconnected and automated. In nativelydigital businesses, the digital channel may be theonly channel. In this context, resilience—the abilityof IT systems to maintain wholly acceptable levels ofoperational performance during planned and unplanneddisturbances—is of growing importance. True resilience iswhat will help organizations mitigate risks to revenue andbrand reputation caused by service disruptions. It’s timeto architect resilience into all dimensions of the nonstopenterprise, including applications, business processes,infrastructure, and security.More vulnerable inmore ways than everAs businesses go digital, they are far more susceptible todisruption—vulnerable because IT systems are constantlyevolving to do things they were never designed for,because update cycles keep shrinking, and because theintensity and frequency of sophisticated cyber attacksare increasing. Add the impact of natural disasters—seemingly more frequent and more severe than before—and it’s easy to sympathize with the challenges beingfaced by brand managers and risk officers of nonstopbusinesses. In an always-on world, business leaders haveto expect and accommodate the risks posed by internaland external disruptions.The economic risks associated with businessdiscontinuities can grow incredibly high, incredibly fast.This is especially true for digital companies that rely onInternet-based business models. Take Google’s fiveminuteoutage in mid-August 2013 as an example; it’sreported to have cost the company $545,000 in revenue. 487ACCENTURE TECHNOLOGY VISION 2014

TREND 6: ARCHITECTING RESILIENCE88Not all outages are so costly; a 2013 Ponemon Institutestudy found that the average cost of data centerdowntime across industries is approximately $7,000 perminute in losses. 5 The cost of disruption varies by industryand the scale of the compromised infrastructure.Arguably, the vulnerability that CIOs feel most acutely isfrom cyber attacks. As transformations to digital multiply,so do the associated risks from cyber criminals. Theseattacks are increasingly substantial, sinister, and sustained.In 2013, for instance, charges were brought against agroup of five hackers based in Russia and Ukraine forstealing more than 160 million credit card numbers overthe past eight years. In that same period, they alsocompromised more than 300,000 accounts from a singlebanking group. 6One of the myriad vulnerabilities highlighted by this group’scrimes is the increasing sophistication of brute-forcepassword attacks. Contemporary password cyphers drawfrom a dictionary with billions of passphrases, route themthrough rule engines, and use massively parallel graphicsprocessingunits to test trillions of passwords against asingle login credential. 7 In short, passwords—even thosestored under cryptographic hashes—are vulnerable.Organizations that understand this insist on multifactorauthentication policies.These days, cyber criminals are highly sophisticated andstrategic in their approaches—and rarely brought tojustice. Three of the five hackers in the aforementionedexample are still at large. Individuals are not the onlyoffenders: organized crime, nation states, and sometimesunscrupulous competitors are also guilty of cyber crimes.Cyber threats are not just about gaining access tosystems. In the case of distributed denial of service(DDoS) attacks, it’s also about shutting down ordisabling services—or at least causing enough secondarydiscomfort to damage a company’s brand. Securitycompany Prolexic reports that in the third quarter of2013, its clients experienced a 58 percent increase in thetotal number of DDoS attacks compared with the yearearlierquarter. 8More advanced threats are not aimed at entire systems;they target specific products and services that maybe beyond the protection of a conventional securityperimeter and may include physical assets. The “black#techvision2014

TREND 6: ARCHITECTING RESILIENCEhats” now have ready access to many helpful tools:for example, the Shodan search engine—labeled the“Google for hackers”—quite easily finds infrastructurecomponents that can be probed quickly for insecureauthentication and authorization. 9 Today, a botnet thatcan do millions of dollars of damage within minutes canbe rented for $7 per hour. 10A surprisingly large proportion of companies concedethat they are unprepared for the scope, severity, andsophistication of today’s attacks. Nearly 45 percent ofCIOs surveyed in Accenture’s 2013 High Performance ITResearch admit that they have been underinvesting incyber security. 11 Many feel overwhelmed about whereto begin; their chances of catching up seem dauntingand expensive.“Arguably, the vulnerability that CIOsfeel most acutely is from cyberattacks. These attacks are increasinglysubstantial, sinister, and sustained.”89ACCENTURE TECHNOLOGY VISION 2014

TREND 6: ARCHITECTING RESILIENCE90Engineering to bea nonstop business,even under attackThe more professional and prolific cyber attacks become,the greater the role that cyber security plays in businesscontinuity. CIOs must use a business-driven strategy tomanaging risk across the enterprise by understandingwhich assets are critical and then prioritizing resilienceand active defense measures accordingly. Theseinvestments should be proportional to the downside riskin the event of a disruption.The time to start architecting for resilience is right now—not when customers expect it or when losses in tradesecrets, revenue or brand value have reached painfullevels. After the necessary discussions about risk withthe organization’s most senior executives, IT leaders mustbegin to map out the threat models specific to theirbusinesses. With this information in hand, they can usebusiness process economics to identify the services mostcritical to the organization’s strategic direction and thusthose most in need of resilience. This might mean givingdifferent tiers of service to different users.After that, it’s necessary to look for investments thatprovide security “bang for the buck,” leveraging existinginvestments and going beyond compliance. Once thesesteps are complete, organizations can start to look atadvanced detection and external threat intelligencecapabilities to better orient their investments toward theareas most in need. This process will provide the CIO withan immense amount of data necessary to move from acompliance-focused stance to one that is more threatcentricand tied to strategic risk. Resilience is far ahead ofcompliance and best practices.Security experts must also architect for a diversityof economic conditions, business risk factors, anda multitude of entry points—including their ownsecurity fabric. Can their own control systems trustthe information they’re receiving? Is their whitelisting (identifying known entities that are trusted)really working? Has their end-point protection beendeactivated by trojan malware?Ensuring trust among all components of a system—through attestation—is the next security frontier. Oneof the best examples of exploits that could have beenmitigated through proper attestation was the targeted#techvision2014

TREND 6: ARCHITECTING RESILIENCEremote attack of Iran’s centrifuge control systemsat a uranium enrichment facility. The trojan malwaredeployed against Iran’s nuclear refining capacity causedcentrifuges to spin beyond their designed operatingparameters while reporting normal operating conditionsback to the control systems. 12In response to this new class of attack, companies asdiverse as HP and Siege Technologies are innovatingattestation solutions at the hypervisor level, while otherssuch as Mocana are concentrating on the machine andembedded device level. 13 Putting it another way, theformer are focusing on ways to verify and trust theoperating conditions of systems while the latter aresecuring end points so that they’re less likely to fall preyto an attack.Once an organization has the technical solutions inplace (DDoS appliances, highly skilled security personnel,applications and infrastructure designed to detect earlywarning signs, security analytics feeding into proactivequarantining, and automated traffic swings and sinkholing),the most effective response is coordinationamong peers. This practice has been adopted bythe financial services community as a response to arepeating pattern of prolonged, serial attacks againstits members. The victims later in the attack chain learnfrom earlier victims, share architecture recommendationsand IP reputation scoring, and provide for continuity inrelationships with law enforcement. This has proven tobe an effective countermeasure and is being mimicked inother industries and by regulatory bodies as a result ofthe successes in the financial services sector.Technologies toimprove resilienceCyber attacks aside, businesses that are striving tobecome digital are racing to keep up with always-onexpectations. It is no longer acceptable to simply postnotices about planned downtime. There is less andless tolerance for service interruptions in any form.Whether systems are brand-new or state-of-the-artdigital systems from the likes of Google and Facebook,or conventional legacy systems, there are many toolsavailable to help systems administrators provide alwaysondelivery of digital services.91ACCENTURE TECHNOLOGY VISION 2014

TREND 6: ARCHITECTING RESILIENCE92To a large extent, CIOs already understand that annualrelease cycles are a thing of the past. Facebook andYammer are among the leading organizations showingthe way forward: they answer the call to be alwayson by deploying updates in staged releases andusing quantifiable metrics and statistical modeling tomeasure their effectiveness. Only if the features reachpredetermined performance metrics are they rolled outto a broader spectrum of users. 14Technology companies are not the only ones movingin this direction; high performers in IT are beginning toembrace agile development practices and are adoptingrelated methodologies for operations—that’s six timesthe rate at which other IT departments do it, accordingto Accenture’s latest High Performance IT study. 15 Thechallenge of transitioning to agile at scale is beingmet by a suite of operational tactics and technologies,including DevOps, performance monitoring and failuretracing, workload management, and software-definednetworking (SDN). Combined, these practices andtechnologies pave the road to resilience by making itpossible to build always-on software and hardwaresystems.DevOps is the business-driven integration of softwaredevelopment and IT operations. DevOps tools such asChef and Puppet allow for highly automated deploymentof entire systems from version control. This enablesthe rapid deployment of new or extended systemsthroughout the compute fabric of the enterprise withoutdisrupting the nonstop business.The agile practice of automated unit testing hastransitioned to operations as well, where newlycommitted code automatically goes through thousandsof test cases before being deployed; once deployed, bestpractice calls for it being deployed on a “canary” serverfirst. If there are any issues, the canary discovers themand stops the cascading of flawed code or configurationto the rest of the production environment. Amazon,Facebook, and Google all use Chef to manage thecontinuous integration of new hardware and software ontheir cloud infrastructures—while staying always on. 16Performance monitoring and failure tracing tools suchas Nagios and New Relic provide data center managerswith real-time insights so that they can inspect andtroubleshoot their systems, from source code to hardwarecomponents.#techvision2014

TREND 6: ARCHITECTING RESILIENCEAnd workload management tools help to makeapplications more portable across heterogeneousinfrastructure—a factor that is increasingly importantwith cloud-first infrastructure strategies. With toolssuch as Akka and Docker, developers can now go beyondagile and leverage their cloud infrastructure investmentsto build more distributed and concurrent applicationsand services, adding resilience to the organization whiledecreasing deployment timelines. Gilt, the flash sales site,uses Akka to build a concurrent, distributed, and faulttolerantevent-driven application that handles the dailyburst in traffic when flash sales go live. 17Traditional content delivery networks (from vendorssuch as Akamai, CDNetworks, CloudFlare, Cisco, andF5) are providing businesses with integrated workloadmanagement technologies that allow them to stayagile all the way to their consumer-facing activities. Inmany cases, these CDNs also give businesses access toinnovation they may not have otherwise. For instance,CloudFlare’s proprietary technology was used to reducethe severity of the DDoS attacks on Eurovision’s annualSong Contest that reaches 170 million viewers. Bymoving to CloudFlare after the site experienced cripplingDDoS attacks during the semifinal round, servicedisruptions were eliminated—something that Eurovisioncould not have done on its own. 18For enterprises using private cloud solutions such asOpenStack, CloudStack, and Eucalyptus, SDN enablesseamless bursting to public cloud infrastructure whenbusiness demands on compute capacity overwhelminternal capabilities. SDN is also invaluable for helpingmanage the transition to agility at scale. When datacenters (or clouds) fail, SDN-enabled organizations caninstantly transfer operations to other online assets, oftenin automated ways and without meaningful serviceinterruptions. SDN showed its ability to contributeto resilience during Super Storm Sandy in late 2012.CurrenEx used a Vello SDN solution to dynamicallyreconfigure routes, service providers, and hybrid cloudinfrastructure. As a result, the company was the onlycurrency exchange in New York City that was able tomaintain connectivity throughout the storm and theensuing cleanup. 1993ACCENTURE TECHNOLOGY VISION 2014

TREND 6: ARCHITECTING RESILIENCE94These types of services make IT systems better able towithstand failure, notifying administrators of dysfunction,increasing portability, and providing self-healingcapabilities—features that circumvent the deficiencies ofthe highly available, state-of-the-art systems of just afew years ago. Those earlier systems were about hardware;now they’re about instances and processes. Rather thantrying to design resilience into every component, it isnow best to take a systemic approach where the servicedelivery architecture should be able to survive the lossof any component—including that of entire data centers.And when components or data centers do fail in aresilient architecture, it’s no longer a disaster recoveryevent; it is a high-availability event.A mindset for resiliencein the digital businessResilience is the new high ground for CIOs who take theirstrategic business roles seriously.That does not simply mean putting in place the rightcyber security structures and deploying best-of-breedhighly available systems. It calls for a wholesale shiftin mind-set to the idea of 100 percent uptime. It is amindset rooted firmly in the context of business riskand a deep understanding of the constant threatsof business disruptions—from hurricanes, hackers, orinternal upgrades—and the risks that those threats poseto maintaining operational continuity and brand value.Above all, the resilience mindset is categorically notabout compliance. Compliance means complacency; inan always-on world, it is not enough to simply check theSarbanes-Oxley boxes to confirm that this or that riskmanagement process is being followed. To be clear, leadersdon’t follow compliance frameworks; they set them.It’s important to know that many of the tools andmethods to engineer for resilience—to design foralways-on operation—are available and improving allthe time. It is not necessary to wait for the maturationor proliferation of a particular technology. As noted,agile development methodologies are already in use, and#techvision2014

TREND 6: ARCHITECTING RESILIENCEYour 100-day planIn 100 days, consider where you can make the most impact inbuilding a more resilient company.• Shift conversations with senior executives about securityto conversations about mitigating business risks. Talk aboutthe benefits of designing for failure.• Map and prioritize security, operational, and failure scenariothreat models to existing and planned business operations.• Develop a strategy to handle elastic business demand for ITservices.• Reaffirm a force-ranking alignment of IT systems andtheir dependent components with business-driven KPIs forsuccess and downside revenue risk. Evaluate the top five forresilience.• Test your resilience by planning a “game day” exercise for IToperations.• Consider hiring an outside security firm to attack yourinfrastructure, monitor the events internally, and reconcilewith logs from the security firm to see where your defensesare deficient.• Perform a data security review. Determine from a businessrisk perspective where data can reside; consider using thepublic cloud as a disaster recovery solution.• If not already doing it, plan a pilot for software-definednetworks and a software-defined data center. Start smalland scale over time.• Create a governance model for auditing and testing theentire ecosystem of IT system and process dependencies—both internally and externally. Be sure it includes policiesfor managing capacity utilization and using hybridinfrastructure.95ACCENTURE TECHNOLOGY VISION 2014

TREND 6: ARCHITECTING RESILIENCE96This time next yearIn 365 days, be ready to embark on projects that will buildresilience and reduce the operational risks of your digitalbusiness.• During the budgeting process, look for security- andinfrastructure-related investments that maximize businessprocess resilience per dollar spent.• Publish a plan to transition IT operations to a DevOps-basedagile organization.• Mitigate business downtime risks by aiming to shiftcompute loads to public cloud infrastructure—either duringpeak times or while under attack.• Consider piloting automated root-cause analysis tools in thedata center.• Use results from game day exercises to create a prioritizedlist for operational upgrades.• Test your system against agile software outputs. Verify thatdeploying faulty code leads to safe environment fallbacks.• Create a security road map to build advanced detection andexternal-threat intelligence capabilities.they can be used to even greater advantage in buildingresilient operations and infrastructure. Even some of thehackers’ most useful tools, such as Shodan, can be usedby the security community as tools to actively defendinfrastructure.The CIOs who truly get the concept of resilience havebegun transitioning their organizations to an alwaysonstate. Knowing that it is neither simple nor cheapto provide real resilience, they are taking a pragmaticapproach, phasing in resilience over time as business riskand process economics dictate. And some are alreadythinking ahead to the time when their entire business isdigital, cloud-based and always on.#techvision2014

TREND 6: ARCHITECTING RESILIENCEsidebarA framework for a resilient futureHow can IT leaders start to design for failure? In MichaelMehaffy’s and Nikos A. Salingaros’ studies of resiliencein the natural world, they uncovered four key principlesthat can be adapted for IT. 20 Any truly resilient IT systemshould demonstrate the following:Interconnectedness. The evolution of networks, frompoint-to-point, to hub-and-spoke, and now to mesh,embodies the benefits that interconnectedness brings.When there are more connections at the edge andthroughout a network, aggregate decision-makingimproves, happens more quickly, and has a greatertolerance for the failure of any one node. Many of thesesame features appear as part of the sharing economy andexpanded workforce as well, which further underscoresthe disruptive power of interconnectedness.Diversity and redundancy. There should be no relianceon singular data sources; embracing redundancy, ITsystems should demonstrate diversity and be designedfor failure. The Hadoop Distributed File System is aprime example of these concepts in action; it has dataredundancy at the document, file, and system levels. Thisredundancy allows analytic jobs to be broken into smallerparts, distributed across the cluster, and run in parallelto achieve results in a highly scalable way. Similarly, highavailability is a primary benefit.Modular scalability. Modular systems can be replacedeasily and they enable rapid scalability. They find usesacross solution architectures and they work well in largeand small deployments. Furthermore, when modularsystems are also decentralized, each cluster of nodesbecomes less and less significant to the functioningof the whole and more independent of centralizedcontrol systems.Adaptation. Sensors that are able to make localizeddecisions based on quantified measurements, domainexperience, and collaboration with peer nodes can havea significant impact on the physical world around them.These decisions on the edge are informed by sharedknowledge and, over time, can gain decision makingcharacteristics akin to biological intelligence.97ACCENTURE TECHNOLOGY VISION 2014

CONCLUSION98#techvision2014

CONCLUSIONCONCLUSIONThe pace of technology changeis acceleratingTen years ago, Facebook was just getting started. Eight years ago,Twitter and Instagram didn’t exist. Six years ago, most mobilephones were incapable of much more than phone calls andmessaging. Five years ago, tablets were a remedy for headaches—not a means of accessing a universe of data.99Digital technologies have burst into our lives with explosive force,and the shock waves are traveling outward at phenomenal speed.Accenture has tracked these waves in its Technology Vision reportsover the last eight years, outlining the impacts and opportunitiesthat continue to lead businesses to become digital businesses.ACCENTURE TECHNOLOGY VISION 2014

CONCLUSION100This year’s report continues to expand our TechnologyVision of the effects this digital surge will have—butwith one difference: this year, we see more and morelarge organizations, outside of the IT industry, activelyriding the surge, even fueling it, instead of beingpressured by it. These organizations understand that itis still early in this initial big wave, but they see just howmuch opportunity it gives them. They are beginningto leverage their vast resources and capital not just toreact to technology disruption but rather to embraceand interweave digital technologies throughout theircompanies to drive the disruptions themselves—andto their advantage. This shift in strategies marks asignificant inflection point. The last decade of changeand opportunity was driven by digital start-ups. Thecoming decade will see the traditional companiesbecome the next set of digital giants. Big is the nextbig thing.So it’s appropriate for all business leaders to nowstart asking pointed questions. Where exactly is ourorganization on its digital journey? Making a start,perhaps by issuing tablets to the sales force?Are we experimenting, seeing what’s working andwhat isn’t? Still learning which business processes lendthemselves best to digitalization? Or, are we confidentlyplanning a large-scale digital transformation that willbe the most important initiative our company hasundertaken in at least a decade—a way to help us hurtlepast our competitors?All businesses, from the global 2000 to small businesses,are facing a pressing need to re-examine their capabilitiesin light of the next stages of their digital journeys. Thepressure applies to both the IT and the business sides.The technology imperative is absolute: it is now time forthe CIO’s organization to decide what role it plays in theemerging digital business—reinforcing the organization’stechnology backbone while equipping the business sidewith the knowledge, understanding, and partnershipswith IT to leverage it.For the business, it is now incumbent on the leaders todefine their company’s place in the digital world. Theymust redefine their relationships with their customers,partners, and the Internet community at large;#techvision2014

CONCLUSIONerase organizational silos around how organizationscollaborate and data is shared; re-examine the roles thattheir enterprise plays in their industry; and lower theboundaries barring entry to other industries as potentialareas for growth.The opportunity at hand for each C-suite executive isto be a digital disrupter—re-creating and redefiningtheir businesses to create competitive advantage. Thepotential for growth is limited only by the creativityof the enterprise itself. Now is the time to move fromdisrupted to disrupter. The opportunity is here; take it.“The last decade of change andopportunity was driven by digitalstart-ups. The coming decade will seethe traditional companies become thenext set of digital giants. Big is thenext big thing.”101ACCENTURE TECHNOLOGY VISION 2014

RESEARCH METHODOLOGY102RESEARCH METHODOLOGYAbout theTechnology VisionEvery year, the Technology Vision team at AccentureTechnology Labs, with the Accenture Researchorganization, pinpoints the emerging IT developmentsthat will have the greatest impact on companies,government agencies, and other organizations in theyears ahead.This year’s research effort began with the collectionof over 3,000 ideas and comments on technologytrends, disruptions, and hypotheses from the architects,engineers, and scientists across Accenture who see theimpact of technology changes every day in their workwith clients. Using Mindjet’s SpigitEngage, an enterpriseinnovation platform, the ideas and examples of thetrends in action were gathered during a trends spottingcontest. Over 850 participants actively engaged incontributing, voting and developing the vast collectionof ideas.These crowdsourced ideas were then screened againstinputs from several other sources, including the recentactivities of commercial R&D labs, the academicliterature, the flow of venture capital funding, andtrends highlighted by IT analysts, and key themes atindustry conferences. For perspective, we also reviewedmany of the ideas with an External Advisory Boardfeaturing diverse perspectives from startups, venturecapitalist firms, IT ecosystem players, and public serviceand enterprise technologists. In addition, we tappedAccenture’s High Performance IT research and thefindings from our annual CIO Forum.The ideas that surfaced during this process stood outfor their relevance to “real world” business challenges.Specifically, the Technology Vision team sought ideas thattranscend the technologies that already drive change—discrete categories such as social, mobile, analytics, cloud—and concentrated on the themes that will soon start toappear on the C-level agendas of most enterprises.#techvision2014

RESEARCH METHODOLOGYThe Technology Vision team then worked with expertsfrom throughout the company to add to, consolidate,filter, and prioritize the accumulated ideas and to testeach idea against the following criteria:• Certainty of transformational impact on enterprises• Velocity and scale of technology change• Impact beyond any one industry “silo”• Ability to be more than a “one for one” replacement foran existing solution• Practical actionability within the next 12 months103• Transcendence of any one vendor or discrete “product”technologyThese tests produced a handful of robust hypothesesthat were synthesized into the six overarching themespresented in this year’s report.ACCENTURE TECHNOLOGY VISION 2014

NOTES104NotesIntroduction1“Tesco Hudl Review: Can a Supermarket Chain Put Out a DecentTablet?” Engadget, October 27, 2013. http://www.engadget.com/2013/10/27/tesco-hudl-review/2“Tesco Profile,” PlanetRetail, November 6, 2013.3“Thomas Staggs at D11: Disney’s MagicBand In Action,”AllThingsD, May 29, 2013. http://allthingsd.com/video/?video_id=528B4FF7-C8CA-42B4-AF89-59DCE53C18F1Digital–physical blur1“Worldwide Internet of Things (IoT) 2013–2020 Forecast: Billionsof Things, Trillions of Dollars,” IDC, Doc #243661, October 2013.2“The Zettabyte Era—Trends and Analysis,” Cisco, May 29, 2013.http://www.cisco.com/en/US/solutions/collateral/ns341/ns525/ns537/ns705/ns827/VNI_Hyperconnectivity_WP.html3“Predicts 2014: Business Intelligence and Analytics Will RemainCIO’s Top Technology Priority,” Gartner, November 25, 2013.4“Self-Driving Cars More Jetsons Than Reality for GoogleDesigners,” Bloomberg News, February 6, 2013.http://www.bloomberg.com/news/2013-02-06/self-driving-carsmore-jetsons-than-reality-for-google-designers.html“Nissan Announces Unprecedented Autonomous DriveBenchmarks,” Nissan press release, August 27, 2013.http://nissannews.com/en-US/nissan/usa/releases/nissanannounces-unprecedented-autonomous-drive-benchmarks5Cisco website: http://www.cisco.com/web/about/ac79/innov/IoE.html;“Embracing the Internet of Everything to Capture Your Share of$14.4 Trillion,” Cisco, February 18, 2013. http://www.cisco.com/web/about/ac79/docs/innov/IoE_Economy.pdf6“Osakidetza Teki Case Study,” Osakidetza, 2012.http://www.youtube.com/watch?v=VkF1u7pO7zc7“M2M Technology Powers British Airways ‘Look Up’ Campaign,”M2M Evolution, November 25, 2013. http://www.m2mevolution.com/topics/m2mevolution/articles/361580-m2m-technologypowers-british-airways-look-up-campaign.htm8San Francisco Municipal Transportation Agency, SFpark website:http://sfpark.org/how-it-works/.9“Amscreen Introduces Digital Technology to Tesco UK PetrolStation Network,” Amscreen press release, October 29, 2013.http://www.amscreen.eu/en/news-1/amscreen-introduces-digitaltechnology-to-tesco-uk-petrol-station-network“Staples’ Omnichannel Stores Showcase the Future of Retail,”Staples press release, June 20, 2013. http://investor.staples.com/phoenix.zhtml?c=96244&p=irol-newsArticle&ID=1831576&highlight=10“Neiman Marcus Testing iPhone App,” Associated Press,March 1, 2012. http://finance.yahoo.com/news/neiman-marcustesting-iphone-app-225258946.html11“Robot Serves Up 360 Hamburgers Per Hour,” Singularity Hub,January 22, 2013. http://singularityhub.com/2013/01/22/robotserves-up-340-hamburgers-per-hour/Companies offering solutions in this energy intelligence spaceinclude Bidgely, Verlitics, Home Energy Analytics, LoadIQ, EnergyAware, and Navetas.#techvision2014

NOTES12Interview with Darrell Smith, Director of Facilities and Energy atMicrosoft; “88 Acres: How Microsoft Quietly Built the City of theFuture,” Microsoft, 2013. http://www.microsoft.com/en-us/news/stories/88acres/88-acres-how-microsoft-quietly-built-the-city-ofthe-future-chapter-1.aspx13“Weather Services for Energy Load Forecasting,” SchneiderElectric. http://www.schneider-electric.com/products/ww/en/5100-software/5190-weather-decision-support-solutions/61688-weather-services-for-energy-load-forecasting/?BUSINESS=914“Digital Agenda: EU-funded Project Uses Robots, Not Humans,to Inspect Petrochemical Containers,” European Commission pressrelease, September 3, 2013. http://europa.eu/rapid/press-release_IP-13-810_en.htm15“Philips Collaborates with Accenture to Create First Proof ofConcept for Delivering Vital Patient Data via Google Glass,”Accenture, October 3, 2013. http://newsroom.accenture.com/news/philips-collaborates-with-accenture-to-create-first-proof-ofconcept-for-delivering-vital-patient-data-via-google-glass.htm16“Gartner Says Smartglasses Will Bring Innovation to WorkplaceEfficiency,” Gartner, November 6, 2013. http://www.gartner.com/newsroom/id/261841517“Brains for Planes: Etihad Taps Big Data to Keep Planes on Time,”GE Reports, June 18, 2013. http://www.gereports.com/brains-forplanes/18“Printing Out Barbies and Ford Cylinders,” Wall Street Journal,June 5, 2013. http://online.wsj.com/news/articles/SB1000142412788732337250457846956028212785219“Ducati: Motorcycle Maker Accelerates Engine Design with FDMPrototyping,” Stratasys, 2013. http://www.stratasys.com/resources/case-studies/automotive/ducati20“Printing Out Barbies and Ford Cylinders,” Wall Street Journal,June 5, 2013. http://online.wsj.com/news/articles/SB1000142412788732337250457846956028212785221“Automated Driving: Legislative and Regulatory Action,” The Centerfor Internet and Society, 2013. http://cyberlaw.stanford.edu/wiki/index.php/Automated_Driving:_Legislative_and_Regulatory_Action22“Ashton Calls for Military-Grade Drones in EU Airspace,” EUObserver, October 22, 2013. http://euobserver.com/defence/12185423“Digital Strategy Does Not Equal IT Strategy,” HBR Blog,November 19, 2012. http://blogs.hbr.org/2012/11/digital-strategydoes-not-equa/24“Smartphone Apps Could Be Life-Saving Link for First Responders,”Nexstar, November 13, 2013. http://www.illinoishomepage.net/story/smartphone-apps-could-be-life-saving-link-for-firstresponders/d/story/Lcd5zTg0uUawXEVYhX7Nlg“More Paramedics Tap Smartphones for Emergency Information”Capital News Services, April 19, 2013. http://news.jrn.msu.edu/capitalnewsservice/2013/04/19/more-paramedics-tapsmartphones-for-emergency-information/From workforce to crowdsource1“Rally Fighter,” Local Motors, 2013. http://localmotors.com/rallyfighter/wp-content/uploads/2013/01/RFBROCHURE2013.pdf2“BMW and Local Motors announce Winners of ‘Urban DrivingExperience Challenge,’” BMW Blog, November 6, 2012.http://www.bmwblog.com/2012/11/06/bmw-and-local-motorsannounce-winners-of-urban-driving-experience-challenge/3“Group Think: Online Collectives Solve Creative Challenges,”Compass, Autumn 2013. http://compassmag.3ds.com/Society/GROUP-THINKACCENTURE TECHNOLOGY VISION 2014105

NOTES4“@WalmartLabs Crowdsources Walmart’s Product Selection withNew ‘Get on the Shelf’ Contest,” TechCrunch, January 18, 2012.http://techcrunch.com/2012/01/18/walmartlabs-crowdsourcesretailers-product-selection-with-new-get-on-the-shelf-contest/Data supply chain1“Big Data Business Benefits Are Hampered by ‘Culture Clash’,”Gartner, September 12, 2013.5“How Lay’s Is Tapping Its Audience for Its Next Big Chip Idea,” FastCompany, August 1, 2012. http://www.fastcocreate.com/1681333/how-lays-is-tapping-its-audience-for-its-next-big-chip-idea“Lay’s Cheesy Garlic Bread Potato Chips Win ‘Do Us a Flavor’Contest,” ABC News, May 8, 2013. http://abcnews.go.com/blogs/lifestyle/2013/05/lays-cheesy-garlic-bread-potato-chips-win-dous-a-flavor-contest/6Kickstarter website: http://www.kickstarter.com/help/stats.2“IDC Digital Universe Study,” IDC sponsored by EMC, December2012.3“The Steadily Growing Database Market Is Increasing Enterprises’Choices,” Forrester Research, Inc., June 7, 2013.4“Journey Toward Analytics ROI,” Accenture, February 27, 2013.http://www.accenture.com/us-en/Pages/insight-analytics-actioninfographic.aspx106#techvision20147“Motor Industry and Crowdfunding,” Megafounder Blog,January 29, 2013. http://blog.megafounder.com/blog/nuevasformas-de-financiar-la-industria-del-motor/8InnoCentive website: http://www.innocentive.com/aboutinnocentive/facts-stats.9“Worldwide and U.S. Enterprise Server Shipments and InstalledBase Forecast, 1996–2016,” IDC, December 2012.“Android Pushes Past 80% Market Share While Windows PhoneShipments Leap 156.0% Year Over Year in the Third Quarter,According to IDC,” IDC, November 12, 2013.10“Forecast: Devices by Operating System and User Type,Worldwide, 2010-2017, 3Q13 Update,” Gartner, September 25, 2013.11“Pebble Smartwatch Tops $10 Million in Kickstarter Pledges, SellsAll 85,000 Watches,” TNW News, May 10, 2012. http://thenextweb.com/gadgets/2012/05/10/pebble-smartwatch-tops-10-million-inpledges-sells-all-85000-watches/12“DARPA Language Challenge,” Amazon Mechanical Turk casestudy. https://requestersandbox.mturk.com/case_studies/cs/darpa5“Big Data Business Benefits Are Hampered by ‘Culture Clash,’”Gartner, September 12, 2013.6“Journey Toward Analytics ROI,” Accenture, February 27, 2013.http://www.accenture.com/us-en/Pages/insight-analytics-actioninfographic.aspx7“Powering Capital Markets Divisions at Major Banks,” Palantir,2012. http://www.palantir.com/_ptwp_live_ect0/wp-content/uploads/2012/06/ImpactStudy_CapitalMarkets.pdf8“Walgreens Opens API for Mobile Prescription Scanning toDevelopers,” eWeek, February 13, 2013. http://www.eweek.com/developer/walgreens-opens-api-for-mobile-prescription-scanningto-developers/9“Facebook Comes up with Exabyte Data Centers for Cold Storage,”StorageServers.com, January 23, 2013. http://storageservers.wordpress.com/2013/01/23/facebook-comes-up-with-exabytedata-centers-for-cold-storage/10Collaborate.org.

NOTES11“Hawaii Uses Collaborate.org as its Data Sharing Backbone,”Government Technology, June 3, 2013. http://www.govtech.com/geospatial/Hawaii-Uses-Collaborate-org-as-its-Data-Sharing-Backbone.html“Collaborate.org: A Global Intelligence Dashboard that Visualizesthe World’s Data,” GeekWire, May 22, 2013. http://www.geekwire.com/2013/collaborateorg-global-intelligence-dashboard-wanted/12“Beiersdorf Leverages HANA and SAP Demand SignalManagement to ‘Tell Stories from Their Data That Marketers Loveto Read,’“ SAP, November 27, 2013. http://www.saphana.com/community/blogs/blog/2013/11/27/beiersdorf-leverages-hana-andsap-demand-signal-management-to-tell-stories-from-their-datathat-marketers-love-to-read13“Using the Power of Math to Solve Medical Mysteries,” Ayasdi,June 20, 2013. http://www.ayasdi.com/blog/topology/using-thepower-of-math-to-solve-medical-mysteries14“Texas Medical Center and Ayasdi to Create a World-ClassCenter for Complex Data Research and Innovation,” Ayasdi,November 13, 2013. http://www.ayasdi.com/connect/pr-ayasdiand-texas-medical-center-to-create-world-class-center-forcomplex-data-research-and-innovation.html15“Enterra’s Cognitive Reasoning Platform Brings Era of Big Data toConsumer Packaging Industry,” Enterra Solutions, October 28, 2013.http://www.enterrasolutions.com/media/Enterra_McCormick_Big-Data_10-29-13_FINAL.pdf16Google website: http://www.google.com/enterprise/mapsearth/products/mapsapi.html.17“UPS Partners with and Backs Online Merchant Lender Kabbage;Will Provide Data for Underwriting,” TechCrunch, February 15, 2012.http://techcrunch.com/2012/02/15/ups-partners-with-and-backsonline-merchant-lender-kabbage-will-provide-data-for-underwriting/Harnessing hyperscale1“Calxeda Adds Frank Frankovsky, Chairman of the Open ComputeFoundation, to its Board of Directors,” Calxeda press release,October 17, 2013. http://www.calxeda.com/news-item/calxedaadds-frank-frankovsky-chairman-of-the-open-computefoundation-to-its-board-of-directors/2“Unilever: SAP HANA,” Accenture, 2013. http://www.accenture.com/us-en/Pages/service-unilever-sap-hana.aspx“Pirelli Analyzes Tire Information with SAP HANA,” SAP customersuccess story, September 25, 2013. http://www.youtube.com/watch?v=v5kr9hxC8D8“The NBA Utilized SAP HANA to Build an Interactive, Real-TimeGame Experience,” SAP customer success stories, July 30, 2013.http://www.youtube.com/watch?v=gjXYEFYI36o3“Datacenter Disruption—Servers,” Evercore Partners,May 27, 2013.“Steve Ballmer: Worldwide Partner Conference 2013 Keynote,”Microsoft, July 8, 2013. http://www.microsoft.com/en-us/news/speeches/2013/07-08wpcballmer.aspx4“Google Details, and Defends, Its Use of Electricity,” New YorkTimes, September 8, 2011. http://www.nytimes.com/2011/09/09/technology/google-details-and-defends-its-use-of-electricity.html5“Putting Data Centers on Turbines to Save Billions of Dollars,”Gigaom, October 16, 2013. http://gigaom.com/2013/10/16/puttingdata-centers-on-turbines-to-save-billions-of-dollars/6HP website: http://www.hp.com/moonshot.7“Building Efficient Data Centers with the Open Compute Project,”Facebook, April 7, 2011. https://www.facebook.com/note.php?note_id=10150144039563920ACCENTURE TECHNOLOGY VISION 2014107

NOTES1088“Build Your Business Case for All-Flash Storage Arrays,” ForresterResearch, Inc., August 21, 2013; updated September 4, 2013.9“XtremIO Customer Case Study: E*TRADE Financial,” EMC,November 14, 2013. http://www.youtube.com/watch?v=VsL1PzZA4SY10“SAP AG Presents at Deutsche Bank Access TechnologyConference,” SAP, September 12, 2013.11“Facebook Shatters the Computer Server Into Tiny Pieces,” Wired,January 16, 2013. http://www.wired.com/wiredenterprise/2013/01/facebook-server-pieces/12“How the Mega Data Center Is Changing the Hardware and DataCenter Markets,” Gigaom Research, March 18, 2013.13“What the Enterprise Can Learn from GM’s Mega Data Center,”Gigaom Research, September 25, 2013.14“GM’s $130M Data Center Takes After Facebook, Runs Hadoop,”Silicon Angle, May 15, 2013. http://siliconangle.com/blog/2013/05/15/gms-130m-data-center-takes-after-facebookruns-hadoop/15“What the Enterprise Can Learn from GM’s Mega Data Center,”Gigaom Research, September 25, 2013.16“BMW’s Connected Cars Force New Data Center,” Slashdot,September 13, 2012. http://slashdot.org/topic/datacenter/bmwsconnected-cars-force-new-data-center/17“Forecast: Data Centers, Worldwide, 2010-2017, 2Q13 Update,”Gartner, August 19, 2013.18“Disruptive Technologies in the Datacenter,” 451 Research, May 2013.19“Calxeda Updates Its ARM Server Platform, Targeting PrivateClouds,” Data Center Knowledge, October 29, 2013.http://www.datacenterknowledge.com/archives/2013/10/29/calxeda-updates-arm-server-platform-targeting-public-clouds/20“Graphics Chips Help Process Big Data Sets in Milliseconds,” MITTechnology Review, October 18, 2013. http://www.technologyreview.com/news/520021/graphics-chips-help-process-big-data-sets-inmilliseconds/21“Nvidia, AMD Bulk Up Their GPU Accelerators,” eWeek,November 18, 2013. http://www.eweek.com/servers/nvidia-amdbulk-up-their-gpu-accelerators.html22Fusion-io and Violin Memory customer case studies.23HP website: http://www.hp.com/moonshot.24“Amazon Puts 26,496 CPU Cores Together, Builds a Top 100Supercomputer,” Ars Technica, November 18, 2013.http://arstechnica.com/information-technology/2013/11/amazonbuilt-one-of-the-worlds-fastest-supercomputers-in-its-cloud/The business of applications1“Enterprise App Stores Can Increase the ROI of the App Portfolio,”Gartner, February 4, 2013.2“High Performers in IT: Defined by Digital,” Accenture, 2013.http://www.accenture.com/highperformanceit3“Case Study: China Eastern Airlines’ Corporate App Store BoostsEnterprise Mobility,” Forrester Research, Inc., July 12, 2013.#techvision2014

NOTES4“Salesforce.com, Inc. Earnings Call Transcript,” Salesforce.com,May 23, 2013.“Japan Post Network Builds 15 Custom Apps for 75,000 Users at24,000 Post Offices on Force.com,” Salesforce.com case study.http://www.salesforce.com/showcase/stories/japanpost.jsp5“How USA Today’s API Powers Partnerships,” Mashery blog,August 6, 2013. http://www.mashery.com/blog/how-usa-todaysapi-powers-partnerships6“Innovator Spotlight: Live Nation—Industrializing APIInfrastructure for Management and Growth,” Apigee, June 5, 2013.https://blog.apigee.com/detail/innovator_spotlight_live_nation_industrializing_api_infrastructure_for_management_and_growth7“The $10 Billion Playbook,” Salesforce.com, November 19, 2013.http://www.sfdcstatic.com/assets/pdf/investors/df13-ics-financialoverview.pdf“Salesforce.com Introduces Salesforce1,” Salesforce.com,November 19, 2013. http://www.salesforce.com/company/newspress/press-releases/2013/11/131119.jsp8“High Performers in IT: Defined by Digital,” Accenture, 2013.http://www.accenture.com/highperformanceit9“AT&T: Using APIs to Help Developers Build Great Apps Quickly,”Apigee case study. http://apigee.com/about/customers/att-usingapis-help-developers-build-great-apps-quickly10“Facebook Annual Report 2012,” Facebook, February 1, 2013.http://files.shareholder.com/downloads/AMDA-NJ5DZ/2832440778x0x658233/46826077-D2FD-4E84-9BBE-C3F844B547A0/FB_2012_10K.pdf“Twitter is Testing a Mobile Ad Product that Will Give Facebooka Huge Headache,” Business Insider, November 4, 2013.http://www.businessinsider.com/twitter-testing-mobile-app-installads-2013-11#ixzz2mvx8fHjo11“Time Warner Cable Customer Case Study,” Alteryx, 2013.http://www.alteryx.com/customer-case-studies-and-videotestimonials12Biosense Technologies’ uChek website: http://www.uchek.in.13“Data Tells Consumer Stories at Unilever,” Tableau Software casestudy. http://www.tableausoftware.com/learn/stories/data-tellsconsumer-stories-unileverArchitecting resilience1“2013 Cost of Data Center Outages,” Ponemon Institute sponsoredby Emerson, December 2013: http://www.emersonnetworkpower.com/documents/en-us/brands/liebert/documents/white%20papers/2013_emerson_data_center_cost_downtime_sl-24680.pdf2“Q3 2013 Saw Significant Changes in Attack Methodologies,”Prolexic, October 23, 2013. http://www.prolexic.com/knowledgecenter-ddos-attack-report-2013-q3.html3“Chaos Monkey Released into the Wild,” Netflix Tech Blog, July 30,2012. http://techblog.netflix.com/2011/07/netflix-simian-army.html4“5-Minute Outage Costs Google $545,000 in Revenue,”VentureBeat, August 16, 2013. http://venturebeat.com/2013/08/16/3-minute-outage-costs-google-545000-in-revenue/5“2013 Cost of Data Center Outages,” Ponemon Institute sponsoredby Emerson, December 2013. http://www.emersonnetworkpower.com/documents/en-us/brands/liebert/documents/whitepapers/2013_emerson_data_center_cost_downtime_sl-24680.pdf6“Russian Hackers Stole More Than 160 Million Credit Cards,” NPR,July 25, 2013. http://www.npr.org/2013/07/25/205548308/russianhackers-stole-more-than-160-million-credit-cardsACCENTURE TECHNOLOGY VISION 2014109

NOTES1107“How the Bible and YouTube are Fueling the Next Frontierof Password Cracking,” Ars Technica, October 8, 2013.http://arstechnica.com/security/2013/10/how-the-bible-andyoutube-are-fueling-the-next-frontier-of-password-cracking8“Q3 2013 Saw Significant Changes in Attack Methodologies,”Prolexic, October 23, 2013. http://www.prolexic.com/knowledgecenter-ddos-attack-report-2013-q3.html9“Shodan Search Exposes Insecure SCADA Systems,” ZDnet.com,November 2, 2010. http://www.zdnet.com/blog/security/shodansearch-exposes-insecure-scada-systems/761110“Cybersecurity Pioneer Barrett Lyon Unveils Defense.net toCombat the Newest DDoS Attacks on Large Enterprise CriticalInfrastructure,” Defense.net Press Release, August 6, 2013.http://defense.net/index.php/company/press-releases/cybersecurity-pioneer-barrett-lyon-unveils-defense-net-tocombat-the-newest-ddos-attacks-on-large-enterprise-criticalinfrastructure/index.html11“High Performers in IT: Defined by Digital,” Accenture, 2013.http://www.accenture.com/highperformanceit12“Stuxnet Was Work of U.S. and Israeli Experts, Officials Say,”Washington Post, June 1, 2012. http://articles.washingtonpost.com/2012-06-01/world/35459494_1_nuclear-program-stuxnetsenior-iranian-officials“Obama Order Sped Up Wave of Cyberattacks Against Iran,” NewYork Times, June 1, 2012. http://www.nytimes.com/2012/06/01/world/middleeast/obama-ordered-wave-of-cyberattacks-againstiran.html?_r=013“Security 2020… What’s Next?” HP Enterprise Services, August2013. http://downloads.hpmediasolutions.com.edgesuite.net/managed/36185-1-InnovationInsight0813_Security_2020.pdfSiege Technologies; Mocana website: https://mocana.com/14“How Do Facebook and Google Manage Software ReleasesWithout Causing Major Problems?” Forbes.com, August 12, 2013.http://www.forbes.com/sites/quora/2013/08/12/how-do-facebookand-google-manage-software-releases-without-causing-majorproblems/“Release Schedule,” Yammer, 2013. http://success.yammer.com/product/releases/15“High Performers in IT: Defined by Digital,” Accenture, 2013.http://www.accenture.com/highperformanceit16“Opscode Guts Chef Control Freak to Scale It to 10,000 Servers,”The Register, February 13, 2013. http://www.theregister.co.uk/2013/02/04/opscode_chef_11_control_freak/17“The Typesafe Platform at Gilt Groupe,” Typesafe case study, 2013.http://downloads.typesafe.com/website/casestudies/Gilt-Live-Case-Study-v1.3.pdf18“Eurovision Taken Down with DDoS, Brought Back Online byCloudFlare,” CloudFlare case study, 2013. http://www.cloudflare.com/case-studies-eurovision19Chad Parris, Managing Director at State Street Global Markets,eExchange group including Currenex.20“Toward Resilient Architectures 1: Biology Lessons,” Metropolis,March 22, 2013. http://www.metropolismag.com/Point-of-View/March-2013/Toward-Resilient-Architectures-1-Biology-Lessons/#techvision2014

About the AccentureTechnology LabsThe Technology Vision is published each year bythe Accenture Technology Labs, the research anddevelopment (R&D) organization within Accenture. Formore than 20 years, the Technology Labs have helpedAccenture and its clients convert technology innovationinto business results. Our R&D team explores newand emerging technologies to create a vision of howtechnology will shape the future and shape the nextwave of cutting edge business solutions.The Accenture Technology Labs offers seminars on theTechnology Vision, which provide a forum to discuss thetrends in greater depth and explore the implications foryour organization’s business.About AccentureAccenture is a global management consulting,technology services and outsourcing company, withapproximately 281,000 people serving clients in morethan 120 countries. Combining unparalleled experience,comprehensive capabilities across all industries andbusiness functions, and extensive research on the world’smost successful companies, Accenture collaborateswith clients to help them become high-performancebusinesses and governments. The company generatednet revenues of US$28.6 billion for the fiscal year endedAug. 31, 2013. Its home page is www.accenture.com.To learn more about the Accenture Technology Labs,or our seminars, contact a member of the TechnologyVision team.Copyright © 2014 AccentureAll rights reserved.Accenture, its logo, andHigh Performance Deliveredare trademarks of Accenture.