|
 |
Gathering examples of IT project failures is not hard.
What is hard is to get a fair assessment of the issues
and reliable figures on the extent of the failure.
What is even harder is to get all the project protagonists agree on what caused the
failure and where the responsibilities lie (they can of course lie outside
the project team). Of course, every project insider or outsider has his
own version and those versions, almost inevitably, "cancel
out".
IT project failure has to be looked at with extreme humility.
Project managers are generally bright, experienced,
motivated and knowledgeable. When a project is deemed strategic the best
resources are generally pulled in. And sometimes - even quite frequently if we
look at the statistics - it fails, thereby demonstrating that good ingredients
are not sufficient to make up a good
mayonnaise.
To illustrate our point that IT project failure is not just
occasional a few
impressive examples are presented hereafter. The best documented
project failures are the ones involving "public money". The more
"public" the money is, the more likely it is that it will be published and, more importantly,
documented and quantified, mostly in Anglo-Saxon literature.
 | In February 1988, hardware problems caused the Bank of America to lose control of
several billion dollars of trust accounts. All the money was eventually found in
the system but all 255 people - i.e. the entire Trust Department - was fired as all the
depositors withdrew their trust money. |
| The first night where the US Fed's new system went life, it lost control over 28 billion
dollars in transfers and awarded interest to the wrong member banks.
Interestingly, the member banks only reported $24 billion dollars in errors the next day. |
US state DMVs (Department of Motor Vehicles) deserve a
special mention. Three of
them stand on our casualty list.
| In 1987, the California Department of Motor Vehicles (DMV) launched a major project to
revitalize their drivers license and registration application process. By 1993, after $45
million dollars had already been spent, the project was cancelled. |
| In 1993, the Oregon Department of Motor Vehicles embarked on what was to be a
five-year, $50 million project to automate its mostly manual and paper-based operation.
The project was justified on the ground that this automation would enable the DMV to
downsize its workforce by one-fifth and save over $7.5 million annually. Two years
later, the five-year project's completion date crept to 2001 (60 % slippage), and the
estimated total budget ballooned to $123 million (146 % slippage). Finally, in 1996, a
prototype was rolled out, but, according to one consultant brought in to examine the
project, the system was fired up on a Monday, and by midweek the DMV test office had no
longer lines than ever backed up around the block. The new system was a total failure.
Soon after, in response to public outcry, state officials killed the
project. In the
aftermath, the only serious downsizing was of the DMV officials who oversaw this disaster. |
| In March 1997 the state of Washington killed the biggest IT project in its history, the
License Application Mitigation Project (LAMP). aimed at automating the state's vehicle
registration and license renewal processes. The scrapped project began in the early
nineties and was supposed to be online in 1995. It planned to create a relational,
client-server system using IBM's MVS/CICS architecture. Initially budgeted at $ 16 M, the
project cost climbed to $ 41.8 M in 1992, $51 M in 1993 and was last estimated (March
1997) at $67.5 M of which $40 M had been spent without result. In 1993, LAMP faded when it
became clear that it was doomed to be a colossal money-waster and even if the plug had not
been pulled it would have been much too big and obsolete by the time it was finished.
LAMP was turned off in 1997, after legislators calculated that the
project ultimately would cost $4.2 million more annually to run than
the state's $800,000 per year incumbent system. |
Those three DMV projects were very similar. They do not require rocket science. There
exist much bigger and more complex IT applications than driver licenses and
registrations management.
But those three projects together squandered more than $ 100 million of public money.
| In early 1993, the London Stock Exchange abandoned the development of its
Taurus paperless share settlement system after more than 10 years development effort was
wasted. The Taurus project manager, Eliott Manley, estimates that, when the project was
abandoned, it had cost the City of London over £800 million (although the Financial Time
of Nov. 3, 1993 reports losses of "only" £400 M - this also points out
how hard it is to get accurate financial figures on IT project cost, especially failing
IT projects). Its original budget was slightly above £6 million. Taurus was 11
years late and 13,200 percent (yes, 132 times) over budget
without any viable solution in sight. |
| In the case of the computer-aided dispatch system for the London Ambulance Service
developed between 1987 and
1993, the problem was not so much one of excessive budgets or
project delays. The issue was rather the usability of the system that leads to its final
failure reported as follows: "...at 2AM on Wednesday, Nov 4, 1993, the system slows
down considerably and then locks up altogether. Rebooting does not solve the problem. The
automatic back-up system also fails to come on-line. A decision is made to revert to
purely manual methods". |
| In 1994, American Airlines settled their lawsuit with Budget Rent-A-Car, Marriott Corp.
and Hilton Hotels after the $165 million CONFIRM car rental and hotel reservation system
project was dumped. |
|
In the '80s, independent Blue Cross licensees (US Health Care Insurance) found themselves
in a major consolidation mode. Big ones were swallowing up little ones, and the newly
massive companies that resulted had to merge disparate claims-payment systems. One of the
biggest blue sharks in the water decided to develop a major new IT system that could grow
along with the company, which was looking to expand beyond health care. The company
brought in a major consultancy, conducted a study, made recommendations and signed an
acquisition contract for a new claims system. 75 IT specialists were put partially on the
project. For the $8 billion company it made sense
spending $5 million to $15 million over two to three years to complete a "business
critical" project. After plowing for one year the project was
eventually killed and the legacy systems were revamped. |
| Prudential Europe has terminated a $50 million
contract with Unisys following the collapse of its Unite project,
which aimed to deliver real-time processing of policies and pensions
over the Internet. The Prudential has referred to serious concerns
about Unisys' ability to deliver on its obligations. The "go
live" date had been delayed to the next calendar year without
commitment on a firm date. Prudential stated that this delay did not fit with
its business plans and sales targets. Unisys
argued that some specifications and requirements were not signed
off or fully defined by Prudential. |
| In 1991, the Passport Office kicked off a project to
automate their entire production process for Canadian travel
documents. The actual work that led to the development of a multi-year
project pompously labeled "Technology Enhancement Plan (TEP)", began
by the end of 1992. Initially,
the TEP project was expected to be completed in two years after the
fall of 1993 at a total cost of $15 million. After
several years of development, the
system is scheduled for implementation in 1999. The
total project cost, as approved by Treasury Board
Secretariat and the Department of Foreign Affairs and International
Trade, was
estimated at $32.5 million before the millennium bug
readiness review. |
Going over time and over budget is not a prerogative of IT
projects alone. It is common to all large projects.
| The channel tunnel was budgeted
at $7 billion, but it entered service in the second half
of 1994 with a
price tag of $13 billion. In 2002 it was still burdened by $9.3
billion in debt, supported by a mere $3.7 billion equity. |
| London – Canary & Wharf 1993 – owner
Olympia & York (the Reichman Brothers.) initial
cost of project $3 billion; final cost $5 billion,
resulting in the bankruptcy of the Reichman brothers. |
| Paris – EuroDisney 1989 – 1992. Initial
cost of the project $2.25 billion, final cost $4 billion. |
Should you want to share with us factual
documentation over failing IT projects do not hesitate to contact
us. This would be more than welcome. Back
to top
Ó IT
Cortex
|
 |
