“Titanic Lessons for IT Projects” - FVP PMP Luncheons - Noblis
“Titanic Lessons for IT Projects” - FVP PMP Luncheons - Noblis
“Titanic Lessons for IT Projects” - FVP PMP Luncheons - Noblis
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
“<strong>Lessons</strong> From the Past<br />
that Assist the Projects<br />
of Today to Shape the<br />
World of Tomorrow”<br />
www.lessons-from-history.com<br />
<strong>“Titanic</strong> <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> <strong>Projects”</strong><br />
<strong>IT</strong> Projects from Hell<br />
Authored by Mark Kozak-Holland, HP Services<br />
March 14th, 2007<br />
Presentation <strong>for</strong> <strong>Noblis</strong><br />
<strong>Lessons</strong>-from-History<br />
Strictly Proprietary & Confidential<br />
Copyright 2007 MultiMedia Publications
Objectives: Analyze Titanic’s construction project and voyage<br />
and provide a contrary view to popular held belief<br />
• It explains how we can take stock and use lessons learned<br />
to help understand key project issues.<br />
• It looks at key decisions made during the project that led to<br />
compromises and cutting corners in the project stages.<br />
• It questions why the ship’s captain was unable to prevent<br />
the disaster.<br />
• It makes a step by step comparison to today’s <strong>IT</strong> projects.<br />
• Please prepare questions <strong>for</strong> the end of the presentation.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 2
As you probably know the success rate of <strong>IT</strong> projects is stubbornly<br />
low as first shown by the “Chaos” reports from Standish Group<br />
• Hypothesis:<br />
– <strong>IT</strong> projects set seeds<br />
<strong>for</strong> future operational<br />
failures - take months<br />
to come to light.<br />
– Problems in on-line<br />
operation attributed<br />
back to poor<br />
decisions made in <strong>IT</strong><br />
project.<br />
Source: “Chaos, a recipe<br />
<strong>for</strong> success,” Standish<br />
Group, 2004. Part of a<br />
series 1994, 1996, 1998,<br />
2000, 2004 - 280,000<br />
projects evaluated<br />
Check www.lessons-from-history/Project Success or Failure/<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 3
Some notable project failures during the project implementation or<br />
into operation<br />
• January 2007, Sweden's largest Bank, Nordea, the biggest heist of customer accounts on record<br />
more than $1m was stolen.<br />
• The Hershey Foods ERP system implementation failure ($112m) led to massive distribution<br />
problems and 27% market share loss.<br />
• The FoxMeyer Drug ERP system implementation failure led to collapse of entire $5 bn company.<br />
• July 2005 HSBC admitted hardware failure caused a major systems crash that hit thousands of<br />
customers <strong>for</strong> ATM, credit/debit, online services and internet, and it was the worst in its history.<br />
• June 2004, RBC fell behind processing salary deposits thousands of Canadian workers as<br />
millions of transactions were affected by a computer glitch that caused payroll delays.<br />
• June 2004, an air traffic control computer failure saw massive air disruption across the UK. All<br />
flights from UK airports were grounded after a problem at the National Air Traffic Service.<br />
• August 2004, a computer crash prevented thousands of UK pensioners collecting benefits<br />
payments on the busiest day of year after the £500m Benefits Transfer system went down.<br />
• September 2004, hundreds of flights grounded <strong>for</strong> 3 hrs, Western US airports. A computer failure<br />
stopped pilot/traffic controller contact. In 5 instances planes passed very close to each other.<br />
• October 2004, computer failure at Waikato Hospital (NZ) left thousands of health workers out of<br />
pocket and <strong>for</strong>ced the manual processing of patient records.<br />
Check www.lessons-from-history/Project Success or Failure/<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 4
Notable project failures during projects. Considering evolution of<br />
PM in last 3 decades and increasing maturity of <strong>IT</strong> why do problems<br />
still occur and some <strong>IT</strong> projects fail catastrophically?<br />
• 2006 Department of Homeland Security scuttles its $229m Emerge2 program (new<br />
financial <strong>IT</strong> system).<br />
• 2005 US Justice Department stated $170m FBI Virtual Case File project a failure, after 5<br />
yrs & $104m. In a 18-month period, FBI gave contractor 400 requirements changes.<br />
• 2005 UK Inland Revenue gave $3.45 bn of tax overpayments because of software errors.<br />
• April 2005 Australian inter-departmental warfare resulted in failure of $64m federal project.<br />
• 2005 British food retailer J Sainsbury wrote off $526m in automated supply-chain system.<br />
• IRS project on taxpayer compliance took decade to complete and cost $50 bn.<br />
• Oregon DMV conversion to new software took 8 years and public outcry killed the project.<br />
• State of Florida welfare system plagued with numerous errors & $260m in overpayments!<br />
• May 2005 major hybrid car manufacturer installed software fix on 20,000 vehicles. The<br />
automobile industry spends $2 to $3 bn per year fixing software problems.<br />
• July 2004 new welfare management system in Canada costing $200m unable to handle<br />
simple benefits rate increase. Contract never tested this in 6 weeks of acceptance tests.<br />
• 2004 Avis cancelled an ERP system after $54.5m is spent<br />
Check www.lessons-from-history/Project Success or Failure/<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 5
The costs of lost services are significant regardless of<br />
industry & relate back to decisions made in the project<br />
Outage Costs By Industry<br />
Industry Description By Minute By Hour<br />
Package shipping Like Fedex $467 $28,000<br />
Internet e-tailer (small) Like CDNow $750 $45,000<br />
Computer vendor Like Dell $1,000 $60,000<br />
Financial Institution Like Citibank (credit/debit) $1,300 $78,000<br />
Retailer sales catalog Like Sears $1,500 $90,000<br />
Internet e-tailer (large) Like Amazon $1,500 $90,000<br />
Financial institution Like Charles-Schwab (brokerage) $1,500 $90,000<br />
TV home shopping Like The Shopping Channel $1,550 $93,000<br />
Telco Like ATT $2,000 $120,000<br />
Pay-per-view Cable provider, like Rogers $2,500 $150,000<br />
Network vendor Like Cisco $2,783 $167,000<br />
Microchip manufacturer Like Intel $3,750 $225,000<br />
Airline Like Sabre $36,000 $2,160,000<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 6
As complexity has risen so has the risk. The internet brings<br />
further risks as it exasperates dependencies on on-line services<br />
• Integration of solution into a complex environment of<br />
applications and databases.<br />
• Interface to the Internet:<br />
– Puts backend operation on-line, exposes integration flaws to world.<br />
• An online mortgage can be approved in 20 minutes, from days or weeks.<br />
– Carries high expectations of 24x7 service delivery and on-line access.<br />
– Enables fickle, savvy customers to click over to competitive services.<br />
– Leave site in 4 seconds if no response<br />
– Increases speed that on-line services acquire mission critical status.<br />
– Expands B2B services to global 24x7, e.g., suppliers, resellers,<br />
partners.<br />
• Millions of lines of code<br />
• Integration with complex technologies<br />
• Massive user populations (to justify solution)<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 7
Not only do projects need to be successful but have to<br />
come in on time, e.g., like hot products<br />
• Xbox get it right - launch window has to be “bang on”<br />
• Some product lifecycles last no longer than 4-6 months and<br />
sometimes, 1 month<br />
• HP reduces laptop development cycle from 9 months to 4<br />
weeks to respond to changing technologies and product<br />
obsolescence.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 8
Bad <strong>IT</strong> service per<strong>for</strong>mance, <strong>IT</strong> project success rates, & <strong>IT</strong> investments<br />
are not just a CIO problem but <strong>for</strong> the whole c-level c<br />
executive<br />
• Corporate benchmarking study identified serious deficiencies<br />
in senior executive management skills with <strong>IT</strong> projects. Lack<br />
of PM skills cut benefits of <strong>IT</strong> projects by 25%.<br />
• “Project governance practices today focus on making<br />
commitments, not keeping them. Executives are involved in<br />
selecting and approving projects, but rarely delivering them.”<br />
49% experienced one project failure in past 12 months.<br />
• Source: KPMG's Global <strong>IT</strong> Project Management Survey, July 2005.<br />
• The CIO and entire c-level executive need to understand:<br />
– The relationship between <strong>IT</strong> projects & on-line operations<br />
– What can go wrong in a complex on-line operation? the cost and impact?<br />
– How it can be prevented?<br />
– ROI <strong>for</strong> creating a strong online operation<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 9
To better understand the relationship between on-line operations<br />
and <strong>IT</strong> projects imagine yourself back in 1912 in one of Titanic’s<br />
lifeboat being rescued by Carpathia. . You ask yourself:<br />
• How could such a disaster happen?<br />
• Why did the ship collide with ice?<br />
• Why did the ship sink it was supposed to invincible?<br />
• Why didn’t the ship slowdown through the ice field?<br />
• Was the ship carrying lookouts?<br />
• Who was in control of the ship?<br />
• Why were there so few lifeboat places?<br />
• Were there no other ships in the vicinity?<br />
• Why did it take so long <strong>for</strong> rescue?<br />
• Are these just the symptoms and not the causes?<br />
• We need to go back to 1909 and look at the project itself.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 10
In 1909 White Star was facing stiff businesses pressures no<br />
different to corporations today. Executives responded with a new<br />
business strategy and took advantage of emerging technology<br />
• Increase in competition and new entrants.<br />
• Aging technology infrastructure, ailing and inferior business<br />
services, leading to loss of market share and customers.<br />
• Invested in a technology infrastructure with 3 new super<br />
liners to sweep Atlantic.<br />
• Pushed emerging technology to limits.<br />
• Addressed all three passage classes, priority on first-class.<br />
• Improved services by quality of<br />
crossing and customer experience.<br />
• Ships built <strong>for</strong> space and com<strong>for</strong>t<br />
(capacity) and not speed.<br />
Mauretania<br />
Olympic<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
15% faster<br />
23% greater capacity<br />
Page 11
The strategy required major new technology investments but the<br />
business case was really solid. The business requirements<br />
specified a lavish ship with space that addressed 3 classes.<br />
• Project “profitability analysis” showed a breakeven in 2 years.<br />
• The construction project would take 6 years.<br />
• A staggering 75% of revenue driven by first-class passengers.<br />
• The price of:<br />
– 1 st class suite - $4,350, 2 nd class suite - $1,750, 3 rd class ticket - $30.<br />
• Titanic’s class segregation no different to how today’s<br />
organizations cater to customer segments.<br />
• By other ship standards third equivalent to second, and<br />
second to first.<br />
• Space allocated:<br />
– 60% <strong>for</strong> 905 first-class passengers.<br />
– 7% <strong>for</strong> 1134 third-class passengers.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 12
In comparison <strong>for</strong> an <strong>IT</strong> project today the project team can<br />
minimize risks with the following techniques:<br />
• Ensure due diligence in examining the business<br />
problem, articulating competitive services, defining<br />
potential costs, and assessing risk.<br />
• Determine by segments customer/target audience, value<br />
propositions, create profiles and scenarios <strong>for</strong> these.<br />
• Determine existing services to be integrated into the online<br />
operation, or data dependencies.<br />
• Establish desirable service level targets to guide the<br />
architect.<br />
• Carefully assess solutions driven by new emerging<br />
technology.<br />
B EST<br />
PRACTICES<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 13
And build a business case <strong>for</strong> the <strong>IT</strong> project that looks into the<br />
operational side and incorporates potential risks:<br />
• Determines the cost of unavailability, based on affected users.<br />
• Identifies and calculates hidden availability costs, especially<br />
people costs.<br />
• Considers branding, image and prestige.<br />
• Highlights the exposure and risk, (potential <strong>for</strong> “bankruptcy”).<br />
• Ensure investments required to move on-line out-weigh costs<br />
B EST<br />
PRACTICES<br />
Revenue generating on-line operation<br />
Cost saving on-line operation<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 14
You can compute your exposure to lost customers/profits by<br />
surveying your customers on the items below<br />
B EST<br />
PRACTICES<br />
# of customers<br />
who experience<br />
a problem<br />
Source: TARP Study<br />
X<br />
Sought<br />
Assistance<br />
50% X<br />
Did not seek<br />
assistance<br />
50%<br />
%<br />
Completely<br />
satisfied<br />
%<br />
Mollified<br />
%<br />
Dissatisfied<br />
% Will NOT<br />
Continue<br />
As Customer<br />
X<br />
X<br />
X<br />
X<br />
5%<br />
25%<br />
70%<br />
45%<br />
=<br />
=<br />
=<br />
=<br />
# Will NOT<br />
Continue<br />
As Customer<br />
$ Profit Lost<br />
due to lost<br />
customers<br />
X $ =..............<br />
X $ =..............<br />
X $ =..............<br />
X $ =..............<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 15
In Titanic’s s architecture stage like in any <strong>IT</strong> project Architects<br />
faced many investment options<br />
Thomas<br />
Andrews<br />
• They created a luxury liner with priority to<br />
first-class service and accommodation<br />
priority on functional requirements (What).<br />
• Harland and Wolff were the most<br />
expensive craftsmen in Europe<br />
• Lavish attention and investments to<br />
passenger com<strong>for</strong>t implied an equivalency<br />
in safety and operations features or non<br />
functional requirements (How).<br />
• The designers had investment choices in<br />
safety features, from lifeboats to the new<br />
technologies like bulkheads, a double-skin<br />
hull, and electric doors.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 16
White Star invested in a ship-builder<br />
builder’s s model the modern<br />
equivalent of an <strong>IT</strong> pilot. They used it to analyze all exposures to<br />
the possibility of loss.<br />
• White Star used early modeling<br />
techniques to test the designs<br />
and identify vulnerabilities.<br />
• Shipbuilders relied heavily on<br />
flow analysis or “static testing” to<br />
review ship characteristics.<br />
• This was a sound strategy<br />
considering the limited testing<br />
options available, and it identified<br />
problems prior to construction.<br />
• Risks of crossing Atlantic well<br />
known with 400 years of travel.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 17
The Architects were well aware of the risks in crossing the<br />
Atlantic. The model tested worst case failure scenarios.<br />
• Running aground<br />
60 feet<br />
Waterline<br />
Tank top<br />
7 feet Hull bottom<br />
Double skin bottom<br />
• Collisions<br />
60 feet<br />
Front-end<br />
collision<br />
Side-on<br />
collision<br />
Crumple zone<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 18
In the architecture stage the project team can minimize risks with<br />
the following techniques<br />
B EST<br />
PRACTICES<br />
• Walkthrough the design, to catch problems early.<br />
• Walk along critical transaction paths end-to-end.<br />
• Complete “component impact analysis” and Architect to remove<br />
single failure points.<br />
• Build security zones <strong>for</strong> access.<br />
• Avoid under-investing in non-functional requirements.<br />
• Avoid one technology, lack of diversity increases susceptibility.<br />
• Avoid complexity, strive <strong>for</strong> simplicity, design <strong>for</strong> manageability,<br />
operability, scalability, per<strong>for</strong>mance, security, and ease of use.<br />
Check www.lessons-from-history/functional vs non-functional.html/<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 19
Titanic’s construction stage integrated many complex technologies<br />
and selected safety features to reduce risks<br />
• Many disparate technologies integrated and<br />
controlled from a single point, the bridge.<br />
• In selecting the “availability level” and “safety<br />
features” assumptions were made about nonfunctional<br />
requirements. This led to over<br />
confidence in ship safety as construction<br />
progressed.<br />
• Investment dollars were put into expensive<br />
safety features based on new technologies in<br />
preference to lifeboats.<br />
• A perception developed that Titanic was<br />
unsinkable.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 20
Decisions were made that compromised individual safety<br />
features and escalated the level of risk<br />
• No Titanic construction dollars were<br />
diverted from basic safety requirements to<br />
enhance the luxurious first-class.<br />
• BUT esthetic factors like spacious public<br />
areas compromised the bulkhead wall<br />
height, built 10 feet above the waterline.<br />
• The double skin not continued (narrowed<br />
interior) was only 7 feet deep and below<br />
the waterline.<br />
• 16 lifeboats were fitted rather than the<br />
recommended 48 (triple stacked) to<br />
provide an uninterrupted view <strong>for</strong> firstclass<br />
promenade decks<br />
60 feet<br />
15 Bulkheads/16 Compartments<br />
Bulkheads<br />
end10 feet<br />
above<br />
waterline<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 21
By the end of construction Titanic’s safety systems were severely<br />
compromised. But officers and crew believed they had the safest<br />
ship ever built and all risks had been mitigated<br />
• Safety regulations, <strong>for</strong> lifeboats, were hopelessly outdated by<br />
advances in ship-building technology.<br />
• Titanic was presented and sold at the highest safety level, but in<br />
terms of passenger safety, it was at the low to medium level.<br />
• The expensive construction ef<strong>for</strong>t was very misdirected<br />
incorporating the earlier mistakes, of the requirements and<br />
architecture stages.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 22
In the construction stage the project team the project team can<br />
minimize risks with the following techniques:<br />
B EST<br />
PRACTICES<br />
• Identify building blocks (components vs prefab) and<br />
solution alternatives (build versus buy).<br />
• Identify non-functional alternatives (safety features).<br />
• Build in cycles. Start small (prototypes), and scale up.<br />
• Tier the solution, scale independently, and create<br />
redundancy.<br />
• Review Government regulations that may impact.<br />
• Ensure business executives and sponsors are involved<br />
throughout the construction, through steering committees.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 23
The business pressures <strong>for</strong> Titanic to go live were enormous with<br />
the large investments tied up in its four-year construction.<br />
• Olympic was in dry dock <strong>for</strong> repair because<br />
of the collision with HMS Hawke, one sixth of<br />
original cost. This delayed Titanic’s maiden<br />
voyage by a month. Titanic's propeller shaft<br />
was used on the Olympic.<br />
<br />
<br />
<br />
• Extensive sea trials and testing were not<br />
considered critical partly because Olympic<br />
was established in service.<br />
• Olympic was a test bed or yardstick <strong>for</strong><br />
Titanic. It is debatable how well the<br />
experiences from Olympic were transferred.<br />
• Formalized change-management/ control<br />
theory not been established.<br />
• Too much faith in Olympic’s track record.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 24
In the planning stage the project team can minimize risks with<br />
the following techniques:<br />
• Review existing/previous projects with the PMO <strong>for</strong><br />
commonality.<br />
• Follow a change-management process, use risk assessments.<br />
• Plan level of testing, and select right tests, and acceptance<br />
criteria.<br />
• Assign operations services ownership and control of process.<br />
• Define alternatives to launch (withdrawal), and back-out plans.<br />
• Create a test environment that mirrors the live environment.<br />
• Prepare <strong>for</strong> increase in frequency of changes with the Internet.<br />
• Deploy in test environment first, run parallel to live<br />
environment.<br />
• Ensure testing is broad not just on functions.<br />
B EST<br />
PRACTICES<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 25
The business pressures and pressing economic needs pushed<br />
Titanic into service and resulted in limited testing completed<br />
• In reality Titanic’s testing consisted of the maiden voyage<br />
across the Atlantic fully loaded with passengers.<br />
• On leaving port, Titanic nearly collided with the steamer New York<br />
coming within four feet indicating the challenges in operating a very<br />
large ship.<br />
• Only one lifeboat drill was per<strong>for</strong>med. The crew was unprepared to<br />
handle a disaster and the launch of all lifeboats.<br />
• No time was spent in preparing<br />
the crew <strong>for</strong> the maiden voyage.<br />
The crew of 900 had 83<br />
mariners.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 26
Titanic’s captain and officers were well aware of “Iceberg<br />
Alley” and the associated risks.<br />
• A well known feature of the North<br />
Atlantic to experienced mariners.<br />
• Winter of 1911 unseasonably mild.<br />
• April worst month <strong>for</strong> icebergs.<br />
• Sailing path moved 10 miles south.<br />
• Fate of French liner Niagara known.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 27
At this point in the project it is common to implement Service Level L<br />
Agreements. Bruce Ismay went further and compromised these.<br />
• Bruce Ismay published a shipping<br />
announcement in the New York Times that<br />
Titanic would arrive a day early to published<br />
schedule. This was a new service level that<br />
proved to be fateful in pushing the ship to its<br />
operational limits.<br />
• The passenger list was a “who’s who” of<br />
public life, with 300 very famous people,<br />
collectively worth of over $500 million. This<br />
underlined public confidence in the ship.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 28
In the testing stage the project team can minimize risks with<br />
the following techniques:<br />
B EST<br />
PRACTICES<br />
• Undertake business and technical risk assessments.<br />
• Ensure independent test teams with incentives to test<br />
objectively.<br />
• Establish the ability to stop an implementation if testing fails.<br />
• Ensure that major testing, once under way, can be halted.<br />
• Ensure the change-management process has various<br />
strategies <strong>for</strong> rapid promotion and implementation.<br />
• Avoid change-management process that lacks political<br />
support and “teeth” to be effective.<br />
• Avoid giving developers deployment rights to live<br />
environment.<br />
• Refine your service level objectives and agreements.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 29
The operating stage required the deployment of the ship into<br />
production and her maiden voyage<br />
• Titanic had a number of built-in feedback mechanisms that were<br />
discounted, fudged (ice bucket test), or just ignored.<br />
• Operators overloaded by commercial traffic (noise) did not pass<br />
the ice warnings (signal) along in a timely fashion.<br />
• Ice-warning in<strong>for</strong>mation eventually communicated through the<br />
hierarchy to Captain Smith wasn’t adequately acted on.<br />
• The captain very resistant to technology relied on “gut” feel and<br />
experience and undermined Marconigram in<strong>for</strong>mation.<br />
• The officers kept their binoculars<br />
and did not share them with the<br />
lookouts.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 30
During the operating stage Bruce Ismay was determined to prove<br />
Titanic was a superior ship to Olympic and meet his new SLO,<br />
dramatically increasing the risks<br />
• The chain of command on board was<br />
usurped as Bruce Ismay overrode<br />
Captain Smith and pushed the ship<br />
and crew continuously to its limits.<br />
• The captain succumbed to the<br />
director’s pressure to sail at full<br />
speed through the danger area.<br />
• The business pressure overrode<br />
the operations services mandate.<br />
• Stringent guidelines were broken<br />
putting everything in jeopardy.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 31
The collision was inevitable and Murdoch almost<br />
succeeded pulling off a brilliant maneuver.<br />
• All the feedback systems were compromised.<br />
• The ship reached its peak speed as three additional boilers<br />
were lit, more than at any other time in the journey,.<br />
• Cali<strong>for</strong>nian’s last radio message was ignored.<br />
• The lookouts warning came 37 seconds be<strong>for</strong>e the collision.<br />
• Officer Murdoch tried to dodge the iceberg and decelerate the<br />
ship through a “port-around” or “S turn.”<br />
Ice Shelf<br />
Iceberg<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 32
The ship grounded itself on the ice shelf in an ef<strong>for</strong>t to avoid a head<br />
on collision. Consistent testimonies of the collision describe it as<br />
innocuous.<br />
• “I heard this thump, then I could feel the boat quiver and could feel a sort<br />
of rumbling..”<br />
– Joseph Scarott Seaman<br />
• “... It was like a heavy vibration. It was not a violent shock.”<br />
– Walter Brice Able Bodied Seaman<br />
• “…I felt as though a heavy wave had struck our ship. She quivered under it<br />
somewhat.”<br />
– Major Arthur Peuchen First Class Passenger<br />
• “I was dreaming, and I woke up when I heard a slight crash. I paid no<br />
attention to it until the engines stop.”<br />
– C E Henry Stengel First Class Passenger<br />
• “We were thrown from the bench on which we were sitting. The shock was<br />
accompanied by a grinding noise….”<br />
– Edward Dorking Third Class Passenger<br />
• “It was like thunder, the roar of thunder…”<br />
– George Beauchamp Fireman<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 33
Unperturbed the bridge sends two assessment groups<br />
to survey the ship <strong>for</strong> damage.<br />
• No sharp jolt of ship slamming horizontally into immovable<br />
object.<br />
• No rebound effect from side swipe or glancing blow.<br />
• No on thrown about, breakfast cutlery in dining rooms<br />
barely rattled.<br />
• No deaths, no injuries, no broken bones. The ship quivered<br />
or rumbled <strong>for</strong> several seconds.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 34
Two assessment groups surveyed the ship <strong>for</strong> damage. Bruce<br />
Ismay made a fateful decision to prove Titanic could save herself.<br />
• The first group quickly return with an inaccurate report of<br />
no major damage. They had descended only a few decks.<br />
• Ismay’s dilemma as there were several options available <strong>for</strong><br />
recovery.<br />
– Remain static and put out a distress call.<br />
– Restart the engines and limp back to Halifax.<br />
• Be<strong>for</strong>e second group returned with architect and carpenter<br />
the director took the decision to sail off the ice-shelf.<br />
• First wireless message sent to White Star office in New<br />
York.<br />
– “T<strong>IT</strong>ANIC PROCEEDING TO HALIFAX. PASSENGERS WILL PROBABLY LAND<br />
THERE WEDNESDAY; ALL SAFE. SM<strong>IT</strong>H”<br />
– All true at 11:53 pm.<br />
• The pumps could not keep up with the increased flooding.<br />
• The architect predicted the ship had 2 hours.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 35
In the operating stage the project team can minimize risks<br />
with the following techniques:<br />
• Ensure the business/operations services refine<br />
SLAs, and adhere these.<br />
• Structure support <strong>for</strong> one holistic client view of<br />
the service, avoid technology silos, assign<br />
operations sole responsibility.<br />
• Build problem-management processes based<br />
on the speed of recovery, & recovery clock.<br />
• Base proactive problem-avoidance around an<br />
early warning system.<br />
• Synthesize and route timely in<strong>for</strong>mation from<br />
feedback mechanisms to decision-makers.<br />
B EST<br />
PRACTICES<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 36
And take a comprehensive approach to organization, processes,<br />
and tools, a basis <strong>for</strong> continuous availability:<br />
• Monitor strategic components critical to environment<br />
availability.<br />
• Avoid claiming a project success too soon.<br />
• Monitor broadly across the whole environment after an<br />
implementation.<br />
• Investigate environmental anomalies quickly using<br />
cause-and-effect, delays may be costly.<br />
• Identify meaningful metrics in “User outage<br />
minutes” rather than percentages of availability.<br />
• Re-evaluate the initial business case based on<br />
returns and metrics.<br />
B EST<br />
PRACTICES<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 37
The officers and crew operated in a state of disbelief unable to<br />
per<strong>for</strong>m an effective recovery. Panic ensued amongst passengers.<br />
• The disaster assessment took 20 minutes, and 65<br />
minutes be<strong>for</strong>e the captain ordered lifeboats filled.<br />
• Poor communication impeded passengers & crew<br />
from reacting, possibly deliberate to avoid panic.<br />
• The hierarchical organizational structure and<br />
physical segregation controlled in<strong>for</strong>mation flow.<br />
• Many passengers got up and went back to bed.<br />
• The first life-boat left half full reluctance to get in.<br />
• The crew skeptical that anything was serious. Any<br />
recovery plan would have been poorly executed.<br />
• Launching 16 lifeboats took over 90 minutes. The<br />
last 2 Englehardts were floated off upside down.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 38
In the operating stage the project team can minimize risks with the<br />
following techniques:<br />
B EST<br />
PRACTICES<br />
• Ensure once disaster is declared it is enacted according to<br />
the plan and followed without hesitation.<br />
• Ensure disaster recovery plans are easily accessible in the<br />
organization.<br />
• Nominate one group “guardian” (operations services) of<br />
the disaster recovery plan.<br />
• Ensure staff adequately trained to follow disaster recovery<br />
plans.<br />
• Practice and rehearse disaster recovery plans regularly.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 39
Following Titanic’s s disaster, both the U.S. and British authorities<br />
conducted post-mortems. The U.S inquiry came close to uncovering<br />
the cover up.<br />
• US inquiry called 82 witnesses,<br />
specialists and technical experts.<br />
• Determined the ship had reached<br />
top speed through the ice-field with<br />
no attempt to slow down.<br />
• Forced Bruce Ismay to remain in<br />
U.S. & grilled him over role as<br />
director, and relationship with<br />
captain, officers and crew.<br />
• Recommended lifeboat space <strong>for</strong><br />
every person on all ships from U.S.<br />
ports; lifeboat drills; adequate<br />
manning of boats; and 24-hour<br />
operation of radiotelegraph<br />
equipment.<br />
• British inquiry saved White Star<br />
from bankruptcy to stop German<br />
competition.<br />
• Saw European war looming, needed<br />
large ships <strong>for</strong> troops and materials.<br />
• Condemned Captain Lord <strong>for</strong> not<br />
responding to Titanic’s flares.<br />
• Criticized British Board of Trade <strong>for</strong><br />
failing to update lifeboat regulations.<br />
• Today private lawsuits would have<br />
brought White Star.<br />
• Titanic unnerved western society’s<br />
faith in technology progress.<br />
• Olympic, served distinguished 24<br />
year career be<strong>for</strong>e being scraped.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 40
<strong>Lessons</strong> learned - what can you take from all this. Your <strong>IT</strong><br />
project is little different to Titanic’s project.<br />
• Historical analogy cuts away layers of jargon and complexity<br />
that shroud project’s today.<br />
• Roots of Titanic’s disaster in the project, compromises to<br />
safety features and elevation of expectations allowed<br />
business pressures to override operational procedures.<br />
• This lead to numerous violations of the “rules of good<br />
seamanship”.<br />
• The probability of failure was very high because of the<br />
inability to recognize the introduced risks.<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 41
Why do projects fail? Recap of Standish group<br />
Project Success Factors<br />
1. User Involvement<br />
2. Executive Management Support<br />
3. Clear Statement of Requirements<br />
4. Proper Planning<br />
5. Realistic Expectations<br />
6. Smaller Project Milestones<br />
7. Competent Staff<br />
8. Ownership<br />
9. Clear Vision & Objectives<br />
10. Hard-Working, Focused Staff<br />
11. Other<br />
Project Impaired Factors<br />
1. Incomplete Requirements<br />
2. Lack of User Involvement<br />
3. Lack of Resources<br />
4. Unrealistic Expectations<br />
5. Lack of Executive Support<br />
6. Changing Requirements &<br />
Specifications<br />
7. Lack of Planning<br />
8. Didn't Need It Any Longer<br />
9. Lack of <strong>IT</strong> Management<br />
10. Technology lliteracy<br />
11. Other<br />
•Why we have the 9 knowledge areas from PMBOK<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 42
Mitigate risk from the project outset through the application of best<br />
practices at each stage of the <strong>IT</strong> project<br />
• 100s of best practices listed by project stage:<br />
1. Project life cycle, deliverables and iteration<br />
2. Business case <strong>for</strong> an online operation<br />
3. Mission critical application dependencies<br />
4. Architectural models and frameworks<br />
5. Enterprise application integration and<br />
interdependency of data<br />
6. Organizational and process elements<br />
7. Change and problem management<br />
8. Use of metrics, service levels objectives and<br />
agreements<br />
9. Use of automation and Early Warning Systems<br />
10. Disaster recovery & business continuity plans<br />
B EST<br />
PRACTICES<br />
Project<br />
Managers<br />
Business<br />
Executives<br />
Implementation of one best practice can save thousands of dollars<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 43
Questions<br />
This presentation will be available on-line<br />
Mark is available to work with you and your organization<br />
(PMs and Executives), speak or run workshops<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 44
Credits and Sources<br />
• 1998 MER<strong>IT</strong> Project. Best Practices in Enterprise Management.<br />
• Bonsall, Thomas E. Great Shipwrecks of the 20th Century. New York: Gallery Books.<br />
• Bristow, Diana. Titanic: Sinking the Myths.<br />
• Brown, David. The Last Log of the Titanic. McGraw-Hill.<br />
• Davie, Michael. The Titanic: The Full Story of a Tragedy. The Bodleyhead Ltd.<br />
• Hyslop, Donald, Alastair Forsyth, and Sheila Jemima. Titanic Voices. New York: St.<br />
Martin’s Press, 1998.<br />
• Lord, Walter. A Night to Remember. New York: Holt, Rinehart, & Winston, 1955.<br />
• Lord, Walter. The Night Lives On. New York: Holt, Rinehart, & Winston, 1985.<br />
• Spignesi, Stephen. The Complete Titanic. Birch Lane Press Group, 1998.<br />
• Thompson, Harvey. Customer Value Management. McGraw-Hill, 2000.<br />
• Wade, Wyn Craig. The Titanic: End of a Dream. New York: Rawson, Wade, 1979.<br />
• Wels, Susan. Titanic: Legacy of the World’s Greatest Ocean Liner. Alexandria, VA:<br />
Time-Life Books, 2000.<br />
• Illustrations were used courtesy of the Ulster Folk & Transport Museum<br />
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 45
Titanic <strong>Lessons</strong> <strong>for</strong> <strong>IT</strong> Projects<br />
Strictly Proprietary & Confidential<br />
Copyright 2007<br />
Page 46