The economics of IPTV - Roland Berger Strategy Consultants

The economics of IPTV – Making IPTV profitableThe economics of IPTV –Making IPTV profitableInfoCom Competence Center

2InfoCom Competence CenterIt's character that creates impact!> We support our clients in achieving a sustainable competitive edge that leads to superior growth,returns, and shareholder value> We develop creative, rule-breaking strategies and advise on their implementation based on ouroutstanding operational skills> We differentiate ourselves by harnessing our functional and industry competence and creativity,combining a global perspective with local sensitivity> We are an independent partnership – from a professional as well as business point of view –and give independent and objective advice> We attract and retain the best talent and work with open, non-hierarchical structures,striving to achieve the highest value for our clientsThe economics of IPTV – Making IPTV profitable> IPTV as a walled garden application (as opposed to WebTV in the open internet) has seen notablesubscriber growth in parts of Europe. This will continue in Asia and the Americas> The business models employed are highly diverse. Often they are specific to individual countries oreven particular companies. Business models range from simple budget IPTV solutions to premiumhigh-end offers in terms of content, features and supporting infrastructure> The middleware ecosystem enabling IPTV services is highly fragmented. However, only a handful ofvendors are currently able to position themselves as future market leaders in terms of capabilities,flexibility and the required speed of implementation/customizing> Companies' motivation for launching IPTV is varied. High on the list of motivating factors are preventingchurn, fending off competition, refreshing the company image, utilizing spare bandwidth capacity andcreating additional revenue streams> While the underlying reason for pursuing IPTV often has a strong strategic component, its stand alonemid-term and long-term financial performance is what will eventually make it a success or a failure.Making IPTV pay off remains a challenge however: IPTV is not suitable for every operator> Roland Berger has extensive experience in strategic, operational, content-related and financial issuesregarding IPTV. We can help you launch, optimize or re-launch your IPTV offer and help it along thepath to financial success

The economics of IPTV – Making IPTV profitable 31. IPTV – A true global success story?Today, in the closing months of 2008, Internet protocoltelevision or IPTV – TV content provided via a closed network,as opposed to "WebTV" which is offered over thepublic Internet (WebTV is the focus of another RolandBerger study published in August 2008: "WebTV insightsand perspectives") – can look back on a short but turbulenthistory. Its story has been one of rapid growth, at least incertain countries. For the end of 2007 experts were estimatingthat IPTV reached over ten million subscribersworldwide. This number is still relatively small comparedto a customer base of several hundreds of millions for alltelecommunications carriers worldwide. Yet the fact remainsthat experts are forecasting an increase of over 500% in theglobal IPTV subscriber base, reaching 60 million by 2010.Driving this increase will mainly be the expected rapid takeupof IPTV in countries like China. Here, SMG (ShanghaiMedia Group) launched the first IPTV services back in2005. By 2006-2007 growth rates were a stunning 900%.IPTV today is stronger in some regions of the world than inothers. Western Europe accounts for around 55% of subscribers,while Asia accounts for 35% and the Americas forjust 10%. Central and Eastern Europe (CEE) plays a minorrole. The reasons for the success of IPTV in Western Europeare varied. Often they include an initial situation of generalTV undersupply, stiff competition in the DSL and IPTVmarket (as in France, for example), high broadband penetration,few regulatory constraints and national or eveninternational telecommunications operators with correspondinglylarge investment budgets. Examining WesternEurope can reveal key drivers in the market and indicatebest practices for successful IPTV products everywhere.This study presents a detailed examination of the IPTVmarket landscape. It discusses our experience of how tomake an IPTV product successful – from a widely neglectedperspective. Current discussions on IPTV tend to concentrateon technology, gadgets, features, content and subscribernumbers. By contrast, we believe that generatingsustainable stand alone financial returns in the medium orlong term is what will ultimately make IPTV a success or afailure. For this reason we focus here on the key actionsthat companies need to take in order to make their IPTVbusiness profitable.1.1. The IPTV market in Europe – France as a rolemodel?The current European IPTV market covers the completespectrum of offers, from "basic" TV over IP only (mainlyfree of charge) to fully fledged premium pay TV offers withmaximum content and features. The market is also highlydynamic. Telecommunications incumbents, who are usuallythe first to introduce IPTV in their home markets, are constantlyfaced with the challenge of new players enteringthe IPTV arena. Moreover, the media business is unknownterritory for them. It is smaller in terms of revenue andmargins compared to the telecommunications market, andit is highly competitive. All in all, a tricky field to negotiate.

4InfoCom Competence CenterFigure 1: OVERVIEW OF MAJOR EUROPEAN IPTV PLAYERSQ4Q3Q2Q1Q2n.a.200420052006 n.a. 2007 n.a. 2005 n.a.Q2 -2004Q22006n.a.1)Q42008Q420062)Q42006Q42003n.a.Q22006Q12006Q42003Q32006n.a.Q42004Q42005LEGENDCompanyIPTV serviceLaunchdateQ22007Q22003Q12006Q32003n.a.Q32004Q420061) Expected launch 2) BT Vision provides VoD only; Freeview provides free digital TV via aerialSource: Industry press research; Company information; Roland Berger researchThe biggest European IPTV players in terms of subscribernumbers are found in France, where IPTV has enjoyed itsgreatest success – not only in European terms but worldwide.The French companies Iliad and France Telecom(Orange) top the league with around 1.55 and 1.15 millionsubscribers respectively, followed by Neuf Cegetel with0.75 million. Major IPTV players in other European countriesinclude the Spanish Telefonica (0.51 million subscribers),Belgium's Belgacom (0.35 million), Germany's DeutscheTelekom (0.25 million) and Italy's Fastweb (0.23 million).The reason France boasts three major players, each withmany more subscribers than any other European carrier,is simple. It also has significant implications for the basicdemand for IPTV services in a country. France has a lowfree-to-air (FTA) channel penetration and a comparativelyhigh penetration level of pay TV. Specifically, there are only6 FTA channels, which can be received via antenna. As60% of French TV households only have terrestrial TVaccess, the demand for more channels is high.Of course, the number of channels offered via cable orsatellite is significantly higher, at 80-90 channels. Butas only 40% of all French TV households can exploit thisoption (installing satellite dishes is often forbidden inhistoric cities) the core market for IPTV is much bigger.The IPTV channel portfolio thus represents a unique sellingpoint. As a consequence, Iliad, France Telecom and NeufCegetel have all introduced IPTV and today offer around80 to 100 channels. Stiff competition between the playershas led to higher than average spending on marketing. This,in turn, has created greater consumer awareness than inany other European country. Following a price war led dbyIliad, prices for 3 Play (i.e. telephony, Internet and TV) arenow around EUR 30 per month. In fact offers are promotedas 2 Play (i.e. telephony and Internet) with IPTV free ofcharge on top. Revenue is generated by providing premiumcontent with high prices via up-selling.

The economics of IPTV – Making IPTV profitable 5Figure 2: OVERVIEW OF THE SUBSCRIBER BASE OF SELECTED EUROPEAN PLAYERS (Q1-Q2/2008)1,550,000 1)2) France1,150,000150 000 France750,000 1)510,000 2)350,000312,000FranceSpainBelgiumPortugal250,000230,000215,000GermanyItalyUK100,00065,00050,000ItalySwitzerlandAustria1) Direct customer relations, excluding indirect CanalSat subscribers 2) Q4 2007Source: Industry press research; Company information; Roland Berger researchThe story is different for example in Germany. Here, IPTVhas enjoyed a more moderate uptake in terms of subscribernumbers until now. Free TV penetration is high and pay TVplays a smaller role. German households by and large haveaccess to satellite or cable TV (terrestrial TV accounting foronly 6% of German TV households), which offer some 100free channels and a wide range of pay TV channels. In Germany,therefore, the channel portfolio does not representa unique selling point. Moreover, cable customers in apartmentblocks often have no say in the choice of cable TVprovider. These customers, like satellite customers, do nothave any "felt" TV costs. As a result the market for paidservices is mainly made up of cable and satellite PayTVcustomers, who can choose their offers independently andof those who are willing to pay for a second TV access.In many countries initial high prices for IPTV offers forcedcustomers to take IPTV bundled with broadband access andfixed-line access, presenting them with an all-or-nothingchoice.Some new IPTV services such as Video on Demand (VoD)and to some extent high definition television (HDTV) faceincreasing competition from existing infrastructure (satelliteor cable TV) and will not be unique to IPTV forever. And theVoD libraries are often small, offering little variety.Going forward, IPTV players in those countries where thenumber of channels is not the driving force will only havelong-term potential for innovation and revenue if they canprovide interactive and customized services and exclusivecontent for niche groups. They must clearly distinguishthemselves from traditional TV offers and access infrastructure.However, seeing is believing: Customers willonly change their consumption behavior once theyexperience the new products or services for themselves.Taking France as a role model for the rest of the worldwould be overstretching the point. While IPTV has beensuccessful here, the TV landscape is very specific.

6InfoCom Competence Center1.2 Business models in Europe – One size fits all?Before digging deeper into the reasons why telecommunicationscarriers deploy IPTV and how they can make aprofitable business out of it, it is worth looking at someof the structural details of the industry.The TV offer pyramid consists of three different categories:basic TV services, mid-range services and premium services.This categorization is based on the differences in theunderlying technology required and the resulting costs, aswell as the cost of content.The pricing of IPTV offers is a good example of this. Currentprice differences between budget, mid-range and premiumIPTV offers are by and large driven by the same two factors,the technological and the content costs. Over time, however,the technological differences will diminish. Pricingstrategy alone will drive the segmentation of the market,with offers ranging from the mass market right up to thehigh-end segment.Monthly prices for subscriptions – including DSL accessand telephony but excluding additional buy-through packages– currently range from EUR 30 for budget services toEUR 70 for premium services. The decision about whattype of IPTV to offer depends d largely l on a company's overallstrategic objectives, as well as the existing and plannednetwork infrastructure (DSL versus VDSL) necessary toenable premium features such as HDTV.Budget IPTV offers are comparable to analog or digitalmulti-channel TV (largely free TV) with respect to theircontent and features. Such offers can typically be foundin markets with a high level of terrestrial TV penetration.Revenue generation through additional buy-through packagesis not the top priority. Instead, companies aim toretain and extend their DSL subscriber base. Features aremainly confined to single SD stream TV, some rudimentaryVoD functionalities and maybe PVR. HDTV is usually notoffered, due to the lack of bandwidth. As a rule, a band-width of around 4 Mbps is sufficient i for enabling budgetIPTV offers.

The economics of IPTV – Making IPTV profitable7Mid-range IPTV offers usually compete with satellite payTV in terms of their content and features. Typically theyare found in strong cable markets. Revenue generation isdriven by a range of additional paid services, such as buy-through h packages and VDI VoD. In addition to budget IPTVfeatures, companies generally offer single HD stream andmultiple SD streams, PVR functionalities and HDD STBs.The necessary band-width to allow smooth IPTV operationis 8-12 Mbps. Mid-range offers dominate the EuropeanIPTV landscape at present in terms of the number ofplayers offering such services.Premium IPTV offers are currently rare in Europe. They arepositioned to be superior to satellite pay TV offers with respectto content and features. Premium IPTV offers includemultiple SD and HD streams, catch-up TV, PiP, fast channelchange, a wide variety of organized content (including VoD,buy-through packages) and STBs with PVR and HDDfunctionality, and much more besides.The necessary bandwidth to enable premium IPTV rangesfrom 16 to 25 Mbps, making a high-performing networkinfrastructure (ADSL2+ or VDSL) a prerequisite. Revenueis generated through a high-priced 3 Play offer and anextensive range of additional buy-through h packages. Inthe future, IPTV middleware will also enable comprehensiveadvertising capabilities, creating an additional revenuesource.Clearly there is no one-size-fits-all business model. Thedecision about what type of IPTV to offer – budget, midrangeor premium – largely depends on the company'soverall strategic objective as well as the existing andplanned network infrastructure. This, in turn, has implicationson the potential means of generating revenue.Figure 3: OVERVIEW OF BUSINESS MODELSBudget IPTV Mid-range IPTV Premium IPTV> Competes with basic multichannelTV, no "real" reasonto move from satellite TV> Features:– Single SD streamor only VoD– No HD, no PVR> Subscriber bandwidth of4 Mbps sufficient> Competes with satellite TV,similar offer, similar quality> Features:– Single HD stream,multiple SD streams– PVR, HDD STB> Subscriber bandwidth of8-12 Mbps> Superior to satellite TV offerin features and usability> Features:– Multiple HD streams, PiP– HDD STB, catch-up TV– Fast channel change> Subscriber bandwidth of16-25 MbpsTV free with BB –Target terrestrialhouseholdsTV in package(triple/quad-play) –Typical strategy onstrong cable marketsPremium TV withadditional features –Target high-end segmentSource: Roland Berger analysis

8InfoCom Competence Center1.3 IPTV middleware in Europe –The leaders of the pack?Middleware constitutes the heart of IPTV deployment. Itis also one of the key cost drivers of IPTV, as we discussfurther below in more detail. Carriers planning to launchIPTV devote a significant amount of time to comparing andtesting different platforms in order to find the middlewarethat best suits their purposes. A wrong decision in the beginningcan not only cause delays in the launch but alsoconstrain future targeted development of the platform,ultimately leading to additional costs for the carrier.Choosing an IPTV platform is therefore an importantdecision, and it is worthwhile taking a brief look at themiddleware landscape as it appears in Europe today.Some of the major IPTV carriers in Europe in terms of subscribernumbers – Iliad, Neuf Cegetel, Telefonica – originallydeveloped their own middleware rather than relying onpurpose-built solutions. Today, Neuf Cgtl Cegetel still runs itsown middleware. Iliad and Telefonica, on the other hand,have sold their platform to Alcatel-Lucent, who manage iton behalf of the carriers and are in the process of mergingit with their own platform. In the future it is expected thatIPTV middleware will be exclusively purpose-built ratherthan homegrown. The purpose-built middleware market inEurope is currently dominated by companies such as Microsoft,Alcatel-Lucent, Nokia Siemens Networks and Thales-Thomson.Microsoft's Mediaroom middleware has managed toattract an impressive range of incumbent carriers in Europeincluding UK's British Telecom, Swisscom and DeutscheTelekom.Alcatel-Lucent was among the first firms to offer IPTVmiddleware commercially. Today the company is the marketleader in Europe in terms of subscribers and markets andmanages two different platforms: its own MiView platformand dMicrosoft's Mdi Mediaroom platform ltf (in cooperation withMicrosoft). Alcatel-Lucent was also one of the first middlewareproviders to start buying up IPTV carriers' platforms(those of Telefonica and Iliad, for example) and mergingthem with their own platform.Following its acquisition of Thales Broadcast & Media,Thales-Thomson gained ownership of the SmartVisionplatform. It successfully placed SmartVision with FranceTelecom in early 2000. However, this remains its onlylarge-scale deployment to date. Nokia Siemens Networkshas deployed its MYRIO platform with Belgacom and KPN.More recently, Ericsson has also entered Europe as anend-to-end IPTV integrator with licensed IPTV middlewarefrom Kasenna, thanks to a deal with Greek OTE 1) .Clearly the middleware market in Europe remains fragmented.There are numerous additional smaller playersand various homegrown solutions on offer. However,Alcatel-Lucent has started the process of consolidation andthis is a trend that is very likely to continue. The differencesin features between different platforms that persist todaywill decrease in number within the next two to three years.The central selling point will become speed and degree ofcustomizability and the ability/willingness of vendors toadopt to the IPTV operators' platform requirements ratherthan sticking to their own roadmaps often driven only bytheir biggest clients. Currently only few of the big platformvendors seem to adhere to these standards, often not evendelivering their own roadmaps on time thus endangeringthe business models of the IPTV operators. A smaller, moreflexible middleware vendor can be the better choice forsome IPTV operators.1) In fact, Ericsson already has one small end-to-end IPTV solution with Vodafone Iceland.

The economics of IPTV – Making IPTV profitable 9Figure 4: OVERVIEW OF THE EUROPEAN MIDDLEWARE MARKET = 360,000 subs = 1,150,000 subs = 530,000 subs = 2,160,000 1) subsPROPRIETARY = 750,000 subsPROPRIETARY = 230,000 subs1) Excluding Arcor launchSource: Industry press research; Company information; Roland Berger research2. Reasons for launching IPTV –Defining successIPTV is currently en vogue for telecommunicationsoperators. However, the strategic rationale for launchingIPTV is not always obvious. Neither is it clear how successshould be defined.Telecommunication carriers are branching out into IPTV fora number of different reasons. The motivation can dependon the market, the player category (incumbent versuschallenger) and even the company.In our client work we have encountered six strategicarguments underlying original investment decisions.Often a combination of arguments is involved. Interestingly,financial returns usually rank lowest on the list: Most companieswith realistic business cases understand that theeconomic benefits of IPTV do not come easily – as wediscuss in more detail in the next section.

10InfoCom Competence CenterFigure 5: REASONS FOR LAUNCHING IPTV1 Prevent churn (mainly access)by value-added added bundling2 Extend revenue streams byentering content/media realm3 Attack competition from cable on 4 Refresh image by positioningits own territory by adjusting thethe company as an innovationproduct portfolioleader5Utilize bandwidth capacitygenerated by the rollout of highspeedbroadband6Generate additional profit forthe company2.1 Prevent churn – A magic wand?Preventing fixed-line access churn is the reason that companiescite often for launching IPTV. IPTV is a value-addedservice that provides a range of bundling opportunities,raising the "churn barrier" for customers. However, makingthe IPTV investment reliant only on positive churn effectsis a risky approach. Even though IPTV can without doubthave visible impact on churn reduction it is very hard tomeasure cause and effect after implementation. We havecome across several business cases based on this lineof argumentation.2.2 Extend revenue streams – The new frontier?2.3 Attack competition from cable – The ultimateweapon?Some operators enter the IPTV business in order to establisha presence in the area of content, media or TV. How-ever, they often neglect two factors in their argumentation.First, in comparison with the telecommunications market,the content, media and TV market that telecom companiescan actually address is often rather limited. It is highlycompetitive and uncharted territory. Second, the focus onrevenue growth in this segment is a double-edged sword.The margins that companies can generate in these marketsare often significantly below typical telecom returns andgradual margin dilution can be expected.IPTV is giving telecommunications companies a longawaitedopportunity to strike back at cable companiesand others menacing their position in the market. Forsome time, operators have been under attack from cableoperators as they have moved from basic TV service intotelecom services such as broadband and fixed voicetelephony. Typically both traditional telecom and CaTVcompanies are present on customer premises – in the caseof telecom companies for PSTN/ADSL, and in the case ofCaTV companies for TV content/broadband. While CaTVcompanies have been able to extend relatively easy intotelephony and broadband, telecommunications carriershave had limited opportunities for addressing the TVsegment, for example through satellite. In markets withstrong competition from CaTV, IPTV gives telecomcompanies a chance to "cut the cable" on their CaTVcompetitors and become the only ones present on thecustomer premises. However, to date few of them havelooked at other strategic alternatives to fulfill this objective,such as direct-to-home (DTH) satellite TV.

The economics of IPTV – Making IPTV profitable 11Figure 6: EUROPEAN REVENUE FOR MEDIA, TELECOMMUNICATION SERVICES, AND CPE& MOBILE EQUIPMENT [EUR BILLION] AND EBITDA MARGINS [%] IN 2007EBITDA MARGINS37%35%18%15%13%TELCOMS'SHARE118 97 49 308 35PRINT TV/RADIO INTERNET SERVICE DEVICES 1)FILM/VIDEODELIVERY1) Includes CPE and mobilesSource: PWC; EITO; Thomson One Banker; Roland Berger researchREVENUES2.4 Refresh image – The wind of change?An argument that cannot be underestimated is the positiveimpact of a successful IPTV offer on a company's image.This is particularly true for incumbents. A sophisticated,high-quality IPTV offer can bring substantial benefits, replacingtraditional customer perceptions based on voiceservices and technology with a message of "we havechanged". However, bad product performance or poorcustomer service can easily undermine the benefits orhave the opposite effect.2.5 Utilize bandwidth – Content for the big pipe?Sad but true: In some cases, companies base their launchof IPTV partly on arguments of the type "because we can".Although industry experts expect growth in bandwidthdemand from customers, real bandwidth requirements ofabove 50 Mbps can only be argued for on the basis ofhigh-quality content such as HD movies.In many cases, large-scale investment in infrastructuresuch as VDSL has generated a need to populate the infrastructurewith corresponding bandwidth-hungry services.This is done in order to drive often unproven customerdemand at profitable price points.2.6 Generate additional profit – The cash machine?We have come across various arguments for launchingIPTV based on the financial return it would generate.Unfortunately in many cases this return has been over-estimated t and companies are left struggling to findadditional resources for urgently needed investments.Many IPTV business cases are now conveniently forgotten,as they reveal significant gaps to real performance interms of both revenues and spending.

12InfoCom Competence Center3. Making IPTV profitable –Money countsThe underlying reasons for launching IPTV are diverse andoften have a strong strategic component. However, webelieve that stand alone financial performance is whatultimately makes IPTV successful or not. Whether launchedto prevent churn, attack competition, refresh the companyimage, utilize spare bandwidth capacity or create additionalrevenue streams, to be truly successful IPTV operators haveto manage to identify key focus areas that maximize thestand alone financial potential as a product, at least midtolong-term.IPTV allows telecommunications operators to move into themedia market. The technology offers great opportunities butit is still just a delivery mode – an entry ticket into the highlycompetitive media business. Telecommunications operatorswho want to enter this area must develop media competenciesin order to compete. They need to negotiate withcompanies that have defined the media market, includingcontent providers and cable operators, or take these firmson head to head. IPTV is not something for every telecommunicationoperator. From designing a TV productto making money in the media market is quite a journey.Telecom operators require a whole new set of skills.Complex TV products have many options designed to encourageup-selling and buy-through of premium packages.TV products usually consist of a basic subscription andvarious buy-through options offered on top. Most of thebuy-through options are bundled together as an "all youcan eat" premium offer. Current trends involve the reductionof the basic subscription rate and increased flexibility ofbuy-through packages. Basic subscription is being reducedto local must-carry TV and free TV channels, to allow easieraccess to value-added buy-through packages. Pay TVbuy-through packages tend to be as flexible as possible,allowing choices that range from à la carte channels todiscounted premium bundles.Below we take a closer look at the key drivers of a profitablestand alone IPTV product, one consisting of a basicsubscription, various buy-through packages and a bundledpremium offer. In this approach we follow the businessmodels chosen e.g. by Iliad in France aiming to achievehigh penetration of basic services first and upselling later.We identify key areas for this IPTV product by modelingIPTV deployment in full, based on our experience withseveral IPTV launches. In order to allow the assessment oflong-term impact we predict possible future developmentscenarios by identifying the highest sensitivities on thethree-year NPV. To this end, we examined more than90 parameters as part of 10,000 Monte Carlo simulationsof typical IPTV deployments. The ranges discussed relateto the aggregated result of this exercise. Parameters wereevaluated dbased on their contribution i to the three-yearNPV variance, which shows the percentage of the forecastvariance due to each assumption (e.g. subscriptionrevenues account for 40-45% of NPV variance). This isbased on computing rank correlation coefficients betweenevery assumption and the three-year NPV. The sensitivitycharts rank the assumptions from most to least important,based on their impact on NPV (through both uncertaintyand model sensitivity).The main area of focus for achieving stand alone profits issubscription revenue. This is followed by content costs, CPEcosts, and platform (middleware) and network costs.

The economics of IPTV – Making IPTV profitable 13Figure 7: OVERVIEW OF KEY DRIVERS OF NPV VARIANCEContribution to three year NPV varianceKey driver35 1 Subscription revenue4 3-6%2 Content costs10-14%3 CPE costs12-16% 40-50% 14 Middleware and network costs215-20%Source: Roland Berger analysis5Other factors31 3.1 Subscription revenues – Capture the customer'sFor example this means that t the impact of a change inwalletpackage pricing (25-30%) in the analyzed business modelhas twice as high impact an on NPV variance as premiumpricing (12-15%).While modern IPTV as a product can generate variousrevenue streams – Video on Demand and advertisingrevenue, for instance – the main contributor to the threeyearNPV variance remains subscription to live TV programs.Our following discussions and analyses therefore focus onan IPTV product with a low basic subscription to get "intothe living room" to achieve high basic penetration from thestart and later upsell to more premium buy-through packages.Obviously, for products designed as "all you can eat"premium right from the start actual "basic" penetration ofthose packages has a much higher impact. For our chosenproduct design composed of basic subscription, buy-through packages and premium offers however, NPV ismainly driven by buy-through packages (both SD and HDchannels).3.1.1 Buy-through package penetration and pricingTogether more than 45% of NPV sensitivity to subscriptionrevenues comes from buy-through packages (Figure 8).IPTV interactivity and proper design of the packages arethe keys to increasing buy-through package penetrationand maximizing profits.IPTV interactivity empowers the product management toincrease buy-through package penetration through toolsnot available to traditional pay TV: Improved methodsof pushing content, enhanced by ease of ordering andsuperior customer profiling, can significantly increasebuy-through package penetration. Programs can be pushedthrough interactive on-screen advertising triggered bycertain user actions and on-screen promotional offers orprogram previews (e.g. opening up a channel for fiveminutes to specific customers). Ordering can be madeeasier by "one button push" subscription, for instance.Customer profiling provides companies with precise informationrelating to customer activities. It allows targetedadvertising, segmented content push and leveraging thecommunity effect (e.g. "users that ordered this channelalso ordered …"). Customer profiling opens the door forcorrelation and advanced cross-selling techniques, suchas those already used on the Internet. However, howcompanies use this data is subject to certain legalrestrictions.

14InfoCom Competence CenterProper basic content offer can increase customers'propensity to opt for value-adding buy-through packages.The basic package is usually presented as the minimumpackage the customer has to purchase in order to accessthe service. This ranges from must-carry TV to, in somemarkets, dozens of channels. It might be tempting toincrease the content of the basic package and thus justifyhigher basic pricing in an attempt to increase ARPU.However, to do so would create an entry barrier to thehigher value buy-through package for customers withnarrowly defined interests.Although the ideal composition of the basic package variesin accordance with viewers' interests, successful TV deploymentsgive viewers themselves the choice. Companies canoffer viewers a basic package consisting of must-carry andfree-to-air channels at a price perceived as being almostfree. In this way they create the lowest possible barriers tocustomers moving up to higher-value channels and makingbuy-through choices.The interests of content providers drive companies towardincluding as many channels as possible in the basic package.After all, every provider wants as large a distributionas possible. Additionally, providers with one successfulchannel may make its distribution conditional on theinclusion of their other channels in the basic package.This often expands the basic package and creates barriersto specialized or exclusive content. Viewers interested inmovies, for instance, have to purchase the basic packagejust to get access to the movie package. Should the basicpackage prove too costly, they may end up rejecting bothpackages.While exploring additional revenue streams such asadvertising and VoD, companies involved in IPTV shouldconstantly check that their live TV package structure reflectscustomer preferences. Allowing users to choose specializedor exclusive content t without t the entry barrier of an overloadedbasic package can significantly increase bothoverall IPTV and buy-through package penetration.3.1.2 Premium packagesIn the analyzed business model (see 3.1) premium packagesare designed as bundled offers for premium users.These packages usually offer top-of-the-range content,functionality and options. This means exclusive content,HD versions, subscription Video on Demand, access toVideo on Demand, catch-up TV, network PVR and so on.This is where IPTV really comes into its own.Many features of premium packages have major costimplications, however. More expensive content, t costlierplatform licenses, higher network bandwidth requirements,more expensive STBs and residential gateways significantlyincrease the cost of premium packages. This can outweighthe superior revenue potential of premium packages.Figure 8 shows how subscription revenue contributes to thethree-year NPV variance. Premium penetration accounts forjust 0-5%, the increase in revenues being almost completelyoffset by the costs incurred. Clearly, many businessmodels peg the premium pricing too low.

The economics of IPTV – Making IPTV profitable 15Figure 8: SENSITIVITY – HOW SUBSCRIPTION REVENUE CONTRIBUTES TO THREE-YEAR NPVVARIANCEPackage pricingPackage penetration20-25%25-30%High Definition pricingHigh Definition penetrationPremium pricing15-20%14-17%12-15%Basic pricing5-10%Premium penetration0-5%Source: Roland Berger analysis3.2 The ultimate offer – Content cost is keyAs we have seen in figure 7, content makes a 15%-20%contribution to the three-year NPV variance. This makes itthe key cost. Unlike traditional TV, IPTV can offer both linearand on-demand content.Linear content is similar to traditional TV content. Here,telecom companies face the same tough negotiations thattraditional TV broadcasters have been facing for years.Content providers push for high subscription fees, highminimum guarantees and channel insertion into the basicpackage to reach all viewers. Carriers can sometimesalleviate content providers' claims by demanding contentdistribution fees.Subscription fees are usually charged per subscriber.They are high for a small number of subscribers and decreaseas the customer base grows. Minimum guaranteeshinder companies with low subscriber numbers and raisea barrier by requiring payments for a minimum number ofsub-scribers. Subscription fees depend on inclusion in theoperator's basic package, the package received by allcustomers.Some content providers, especially those relying onadvertising for a significant part of their revenues, pushfor inclusion in the basic package so as to improve ratingpoints. Three options are generally available: inclusion inthe basic package, extra basic, and à la carte. Inclusion inthe basic package usually costs less per subscriber, but itforces the operator to distribute the channel to its entirecustomer base. Extra basic option offers the operator morefreedom at a higher cost per subscriber. À la carte givescomplete flexibility, usually at the highest price and withthe risk of leaving value on the table by giving up bundlingopportunities.IPTV operators can leverage their increasing customersubscriber base by negotiating lower content prices andsometimes even demanding content distribution fees.Content providers that rely heavily on rating points arethe most willing to pay content distribution fees.

16InfoCom Competence CenterThe major linear TV costs in a typical IPTV package are theHD channel costs, the buy-through package costs and thebasic content costs. Figure 9 shows the relative importanceof each of these components in how content influences thethree-year NPV variance.Overall HD channel costs are greater than all other linearTV costs. For example, this means that a change in HDchannel costs in the analyzed business model has animpact more than five times as high (55-60%) as theimpact of basic content cost (8-10% on three-year NPVvariance. This sensitivity will even increase as in manymarkets. HD will soon become a standard feature with nomore opportunity to ask a premium for it. This is due bothto growing demand for HD and high HD costs. In a competitiveTV market, companies can use HD as a differentiatingfactor to gain key competitive advantages. With the explosionin sales of HD-ready and full HD TV sets in recentyears and the relatively low current penetration of HDcontent, t the hunger for HD content t is growing. g As yet, mostexisting content is SD, and HD channels are considerablymore expensive.On-demand content is one of the distinctive features ofIPTV. It has rapidly evolved from an additional "nice-tohave"option to a requirement for IPTV deployment. Contentis usually stored on the provider network and providedto users on demand.Content providers commonly impose a number of technicalconstraints and apply specific business models andconditions.Technical constraints t relate to the network's technicalspecifications. The operator's network and TV platform mustmeet the provider's security requirements. They must providea closed network environment with certified conditionalaccess and digital rights management systems. Technicalspecifications are often subject to third-party certificationsor security audits.Content providers negotiate the business model andconditions with the provider. This includes the paymentmodel (revenue sharing, minimum guarantees, flat fees,fees per event), release window (delay and length) andsupport of on-demand content. Revenue sharing is themost common payment model, favored especially by majorstudios and for new movies. Revenue sharing can be ashigh as 70% for new releases and is usually coupled withminimum guarantees. Flat fees are found mainly for smallerstudios, older releases or subscription Video on Demand(themed programs associated with linear TV programs).Major studios usually reject an advertising-based businessmodel for recent releases, but both large and small studioswill accept such a model for older releases and speciallydesigned content.Figure 9: SENSITIVITY – HOW CONTENT CONTRIBUTES TO THREE-YEAR NPV VARIANCE- 55-60% HD channel costs- 12-17%- 12-16%- 8-10%VoD contentPackage costsBasic content costs-0-2%SVoD contentSource: Roland Berger analysis

The economics of IPTV – Making IPTV profitable 173.3 CPE policy – Direct correlation to bottom-lineFollowing content, the next most important influence on thethree-year NPV variance is the cost of the customer premiseequipment t(CPE) (CPE). Advanced d equipment must be installedon customer premises to ensure video functionality (set-topboxes) and manage the underlying broadband connection(residential gateways). Set-top boxes can have internalhard drives offering PVR functionality. The analyzed businessmodel assumes premium HDD STBs are not distributedto all customers but only to the premium segment(limited in addition by loop length constraints). The costsof CPEs are proportional to the number of subscribers andhave a direct impact on customer profitability.Typical CPE distribution policies are custody models(in which customers can use the CPE for the duration ofthe contract), rental agreements and sale. While sales ofCPE generate immediate revenues, rental agreements cangenerate higher revenues over time. In established marketswith high ARPUs, CPEs are provided free of charge forcustody or included in the subscription. In custody andrental models, operators must pay special attention to thelogistics arrangements for returning CPEs at the end ofthe contract.Although HDD STBs provide obvious additional functionality,this has in fact already existed on the market forseveral years and is partly replaceable by network-basedIPTV platform functionality (e.g. network PVR, catch-up TV).An aggressive market penetration ti strategy t based on thebasic service and relying on up-selling of different packageswould focus on placing as many STBs as possible on customerpremises. The companies Iliad and HK Telecom havesuccessfully captured market share using this approach.Typical IPTV budget deployments rely on simple STBs withno PVR capabilities. In this case the impact of simple STBcosts account for almost 90%: figure 10 shows the powerfulimpact of simple STBs on the three-year NPV variance.Companies often use STBs to drive users toward highervalue-added services. Simple STBs without HDD arecommon for basic services, while HDD STBs with PVRcapabilities and extended features are usually reservedfor top value-added packages. STB discount schemes alsofavor high-value packages, with discounts of up to 100%for top packages.Figure 10: SENSITIVITY – HOW CPE COSTS CONTRIBUTE TO THREE-YEAR NPV VARIANCE- 80-90%Simple STB costs- 5-10%Modem costs-2-5% HDD STB costsSource: Roland Berger analysis

18InfoCom Competence CenterIPTV uses the broadband infrastructure as support, andthere is close cross-selling between broadband and IPTV.Many operators condition provision of the IPTV service onthe broadband service. While IPTV might be the driver forbroadband dpenetration, ti modem costs are often not directlyattributed to IPTV and can be offset by the broadbandservice. Their influence on IPTV variance is thus reduced.Additional costs occur when IPTV subscribers do not opt forbroadband connection (if this is allowed by the operator)or when the modems already in use for broadband do nothave the necessary functionality to support IPTV (e.g.IGMP).A successful approach will involve operators deploying IPTVcompatiblemodems for all broadband users, cross-sellingIPTV and broadband, pushing for lower STB prices and finetuningthe logistics for returning CPEs at the end of therental period. The residential gateways required by IPTV aremore expensive than the minimum modems required forxDSL. However, deploying residential lgateways for broadbandusers provides an entry point for additional andsubstitute services such as IPTV and VoIP. Technologically,IPTV and broadband are prime candidates for cross-selling.Companies must target their marketing efforts carefully.The latest compression algorithms – H264, for example –are highly complex and place major performanceconstraints on current STBs. This drives prices upwards.As technology evolves and matures, however, compatibleSTBs will become available from various sources and priceswill begin to fall. Given the direct impact of STBs on thebottom line, operators should strive for excellence in theirsourcing activities.3.4 Middleware and network – The underestimatedt dcash burnerFor a traditional telecommunications broadband network tostart providing IPTV services, considerable investments areneeded both in the platform itself and in upgrading theexisting network infrastructure. Different platforms placedifferent demands on infrastructure. Network-agnostic IPTVplatforms can be deployed quickly and easily, while variousarchitectural trade-offs are possible to improve performancefor network-aware platforms.The investments in IPTV infrastructure (platform and network)mainly relate to the network. The sheer bandwidthrequired by video services means that the biggest share ofinvestments goes into upgrading network transport capacity,down to the DSLAMs. The traditional DSL access and corenetworks used to provide best-effort broadband accessrequire upgrading in order to provide the guaranteedbandwidth needed for IPTV.The two types of services – live TV and content on demand –place different constraints on the various components ofnetwork infrastructure. Live TV network load can dependboth on the number of live TV channels and on the numberof subscribers. Content on demand, on the other hand,depends solely on the number of subscribers viewing thecontent at the same time.Live TV is usually carried either solely as multicast traffic oras a combination of multicast and unicast traffic. Multicasttraffic increases load in proportion to the number of live TVchannels (multicast streams). Unicast traffic increases networkload in proportion with the number of users watchingcontent at the same time.In the case of multicast, each channel is streamed once forall users across the entire core, edge and access network.Viewers of a specific live TV channel join the same multicaststream as other users, thus not increasing the overall networkload. Adding a new TV channel increases the bandwidthrequirements on the entire multicast domain, however.To support full multicast distribution, DSLAMs must beIP-aware (rather than blind layer-2 DSLAMs) and be ableto perform channel replication (understand IGMP requestsand replicate the right TV channels to individual users onthe local l loop). The multicast t approach has the advantageof reduced bandwidth consumption and faster channelzapping.In combined multicast/unicast transmission, channelreplication is performed at a higher level inside the network,transforming the multicast stream into a unicast streamearlier on in the process. The live TV channel is carried asa combination of multicast and unicast traffic – multicastfrom the headend down to the edge router and unicastfrom the edge router to the user through the DSLAM.

The economics of IPTV – Making IPTV profitable 19Adding new users increases the load on the access network(from the edge router to the user), while adding new TVchannels increases the load on the core and edge networks(from the headend to the edge router). This approachallows for easier network management and simplerDSLAMs, thus reducing DSLAM costs.Content on demand (CoD) is generally different for eachuser and thus requires separate unicast streams. Thismeans that each new user increases the network load.Adding more content does not automatically increasenetwork requirements. However, the number of usersconcurrent directly impacts on network load. Core networkload can be managed by installing content edge serversand moving some of the most popular content closer tothe user consuming it.Network upgrade sizing must therefore take into account anumber of factors. These include the platform architecture(centralized versus distributed), ib t d) the number of SD/HDchannels, the CoD concurrency and the HD/SD bitrates.Depending on current bandwidth availability on the differentnetwork segments, architectural trade-offs can considerablyreduce the costs. Figure 11 shows how much network upgradeinvestments impact on the three-year NPV variancefor a typical telecom network providing up to 20 Mbps DSL.The investments only include bandwidth upgrades from corenetwork up to the DSLAM – not the expansion of DSLcoverage or local loop upgrades/shortening.The content and CPE costs discussed above are mainlydependent on external providers – STB manufacturers,content studios and so on. Infrastructure costs, by contrast,are mainly dependent on the operator's current infrastruc-ture. IPTV is an excellent means of filling available capacity,but it can prove a resource drain for under-dimensionednetworks. Companies deciding whether to invest in thenetwork must weigh the high bandwidth requirements ofvideo services and their cost against possible alternativeuses of bandwidth, if such exist. Pure bandwidth networkupgrades are non-IPTV specific and can be reused.As consumer usage patterns change, so does demand onnetwork capacity. Operators offering IPTV can make varioustrade-offs between bandwidth and the installed equipmentbase. As content on demand becomes more popular,bandwidth requirements on core end edge network can bediminished by replicating content servers from the core tothe edge servers. Access network bandwidth (down to theDSLAM excluding ldiglocal ll loop) can be saved dby minimizing ii iigthe live TV requirements through multicast transmissionmode and more intelligent DSLAMs.Figure 11: SENSITIVITY – HOW INFRASTRUCTURE COSTS CONTRIBUTE TO THREE-YEARNPV VARIANCE- 70-75%Network- 10-15%Middleware- 5-10%Head-end upgrade- 5-10%IT integrationSource: Roland Berger analysis

20InfoCom Competence CenterThe middleware provides the entire IPTV functionality andmust be closely aligned with the business model. Investingin middleware has less of an effect on the three-year NPVvariance than upgrading the network, but it is entirely IPTV-specific and thus harbors certain ti risks ik should ldIPTV notprovide the desired results.The headend provides a live video signal to the entire network.Operators that already provide live video signals viasome other technology (such as DTH) can upgrade theircurrent headend for IPTV functionality. Headend acquisitionor upgrade occurs at the start of the project for a givennumber of channels; it then increases step-by-step asmore channels are added.IT is essential for linking the different parts of the projecttogether. It enables the operator to provide a seamlessservice to customers, including CRM, ordering and billing.Most IT costs are incurred at the start of projects, due to theitg integration ti work required. dLt Later IT costs stem from servicemaintenance and change requests.While IT integration's importance looks relatively low inour analysis compared to the "war stories" of many IPTVmiddleware integrations, the results are based on an averageintegration success without a major failure. In additionsome IT integration costs are part of the middleware implementation(e.g. consulting services) item or are covered byinternal resources where costs of integration are often noteasy to separate or assign proper price tags. Obviously ITintegration will play a more important role in terms of NPVimpact e.g. for incumbents with very complex and failureprone OSS/BSS systems or when IT integration is notmanaged properly and fails to deliver expected results.Other factors such as defining the correct processes,providing service and installation, and developing selfinstallpacks, call centers and help desks are essential forthe proper development of an IPTV service. While development,if properly carried out, accounts for just 3-6% ofoverall impact, failing to do so can have a significantadverse effect on financial results.4. Conclusions – Eight lessons to belearnedA wide range of business models, middleware platformsand technical infrastructure options supporting IPTV finallyrepresents an opportunity for telecom companies to enterthe media market with some initial success in terms ofsubscriber numbers.1. Financial profit, i.e. making IPTV as a product pay offfinancially remains a challenge.2. IPTV is no "silver bullet" and not suitable for everyoperator. Whatever their strategic reasons for choosingto develop IPTV, companies must also pay closeattention to the areas of subscription, content, CPE,middleware and infrastructure in particular to build thefoundation for success and robust stand alone financialreturns at least mid- to long-term.3. For the business model of a basic entry product andlater upsell analyzed here subscription revenuesshould be actively pursued through content pushing.This requires superior customer segmentation, ease ofordering and smart content bouquet design. Contentcan be positioned through interactive advertising andonscreen promotional offers based on customer profiles.One button push subscription can provide immediaterewards for marketing activities. The mandatory basepackage should not represent a significant barrier tohigher value buy-through packages. Obviously for "allyou-can-eat"type of IPTV products overall package"basic" penetration is much more crucial.4. Content costs must be carefully negotiated with contentproviders – especially for differentiating content, such asHD channels and on-demand content. As the customerbase grows there will be opportunities for reducing thecost of content and even negotiating content distributionfees for rating-dependent content providers.

The economics of IPTV – Making IPTV profitable 215. CPE should encourage IPTV penetration. Once establishedin the living room this can stimulate package buythroughby means of discount schemes and features.It can also be reused if loaned to the customer undera rental tlor custody td agreement. tAhi Achieving i excellence inCPE procurement will have a direct impact on thebottom line.6. Middleware platforms have to be carefully selectedbased on their revenue generating features and differentdemands on infrastructure as well as the ability andwillingness of the vendor to adhere to operators' requirementsand change requests in a timely manner.Network-agnostic IPTV platforms can be deployedquickly and easily, while various architectural trade-offsare possible to improve performance for network-awareplatforms.7. Infrastructure must be properly dimensioned andarchitectural t bandwidth equipment tradeoffs taken intoaccount. Core and edge bandwidth can be traded forCoD servers and access network bandwidth for moreintelligent DSLAMs. Network load is strongly influencedby product managers' ability to stimulate certain kindsof CoD.8. Telecom companies inherently have some of theskills required for running financially successful IPTVoperations. However, they also require skills that aremuch more specific to the media and TV market. Theytypically lack skills of this type. To ensure success inIPTV, they must effectively close the expertise gap.5. Roland Berger expertise in IPTV –How we can helpThanks to numerous international ti IPTV projects, theInfoCom Competence Center at Roland Berger StrategyConsultants has acquired substantial strategic, operationaland financial expertise in launching, optimizing and relaunchingIPTV portfolios. We offer our clients comprehensivesupport in the following areas:> IPTV strategy and business case: We have the relevantexperience to design actionable IPTV strategies. Thisexperience comes from in-depth market analyses of allmajor European countries, competitive assessments(including Web TV and DTH satellite offers and realisticmarket entry or re-launch business cases based onoperational benchmark data) and a deep technicalunderstanding of the underlying cost implications> IPTV optimization and operations: We have helpedoperators fine-tune their IPTV offers and increase marketsuccess. Our work has included process optimizationsand quality plus efficiency initiatives in technical andcustomer service, content portfolio and cost optimizations,and optimizations of customer touch points> Full-scale IPTV re-launch: We take a rigorous approachto re-launching IPTV offers that have not lived up toexpectations. Here, our work has included setting up andmanaging an IPTV re-launch program, redesigning go-tomarketplans, and implementing best-in-class technicaland customer service processes and operationsIf you wish to discuss any of the issues raised in thisdocument or have any other questions, please call us on+49 (0) 89 9230 8037.We look forward to hearing from you!

22InfoCom Competence CenterAuthorsDirk ReiterMember of the Executive Committeeand head of the Infocom Competence CenterE-mail: dirk_reiter@de.rolandberger.comKlaus-Ulrich FeilerPartnerE-mail: uli_feiler@de.rolandberger.comAlexander MoggPartnerE-mail: alexander_mogg@de.rolandberger.comTim BottkePrincipalE-mail: tim_bottke@de.rolandberger.comFelix IblherSenior Project ManagerE-mail: felix_iblher@de.rolandberger.comAdrian SarbuSenior ConsultantE-mail: adrian_sarbu@ro.rolandberger.comDennis MonzkaConsultantE-mail: dennis_monzka@de.rolandberger.comList of abbreviationsADSL:Asymmetric digital subscriber lineHDTV:High-definition televisionARPU: Average revenue per userIGMP: Internet group management protocolCaTV: Cable televisionIPTV: Internet protocol televisionCEE: Central and Eastern EuropeMbps: Megabits per secondCoD: Content on demandNPV: Net present valueCPE: Customer premises equipmentnPVR: Network private video recordingDSL: Digital subscriber linePiP: Picture in pictureDSLAM: Digital subscriber line access multiplexerPSTN: Public switched telephone networkDTH: Direct-to-home (satellite service)PVR: Private video recordingEBITDA: Earnings before interest, tax, depreciation and amortizationQOE: Quality of experienceEPG: Electronic program guideSD: Standard definitionFTA: Free-to-airSTB: Set-top boxH264: Advanced video codecUGC: User-generated contentHD: High-definitionVoD: Video on demandHDD: Hard disk driveVDSL: Very high speed digital subscriber line

Published by Roland Berger Strategy Consultants, September 2008InfoCom CC, Dirk Reiter, Alexander Mogg, Klaus-Ulrich FeilerMies-van-der-Rohe Str. 6, 80807 Munichwww.rolandberger.com, e-mail: Infocom@de.rolandberger.com© Roland Berger Strategy Consultants. All rights reservedThe economics of IPTV – Making IPTV profitable 23

InfoCom Competence CenterAmsterdamBahrainBarcelonaBeijingBerlinBrusselsBucharestBudapestCasablancaChicagoDetroitDüsseldorfFrankfurtHamburgHong KongIstanbulKyivLisbonLondonMadridMilanMoscowMunichNew YorkParisPragueRigaRomeSão PauloShanghaiStuttgartTokyoViennaWarsawZagrebZurich© Roland Berger Strategy Consultants, Germany09/2008, all rights reservedwww.rolandberger.com