Epic fails

One in eight IT projects succeeds. What do we learn from the other seven?

Project management Remember this comment about a particular project somewhere in the Middle East? “There are known knowns; there are things we know that we know. There are known unknowns; that is to say, there are things that we now know we don’t  know. But there are also unknown unknowns; there are things we do not know we don’t know.”

You guessed it: the words were uttered by the infamous former US Secretary of Defence, Donald Rumsfeld in 2002.

There’s a truth in there, but which one? Perhaps it’s that if you come out with stuff like that your P45 is never far away. Rumsfeld eventually picked up the US equivalent in 2006, shortly after eight retired generals and admirals called for him to go citing his “abysmal planning and lack of strategy” in Iraq.

Lack of planning and strategy is certainly important but the causes of project failure are many and varied. They could be external factors, management, technical problems, poor communication, lack of focus, insufficient stakeholder commitment, as well as the aforementioned lack of planning and strategy.

Nor is failure clearcut. It might range from cancellation, late completion and being seriously over budget to a whole grey area of success tinged by a degree of failure.

BCS, The Chartered Institute for IT (formerly known as the British Computer Society) researched IT project failure in 2008. It found that only one in eight projects can be considered truly successful, counting out those that do not meet time, cost and quality requirements criteria.

The research concluded: “Despite such failures, huge sums continue to be invested in information systems projects and written off. For example the cost of project failure across the European Union was €142bn in 2004.”

But all failures provide indicators to keep today’s project managers from stepping into the abyss.

One of the best known UK government IT failures was the 1999 Passport Agency project, which had a huge impact on the public.

A new computer system at regional offices created a logjam which kept applicants waiting up to six weeks for their passports, instead of about ten days. At the worst point in July over half a million passports were being processed and thousands of people had to cancel or rearrange their travel plans.

Post-crisis analysis revealed problems with the new system in the Liverpool and Newport offices: training had been inadequate and data entry took longer than anticipated. In addition the problems in Liverpool were not solved before the system was introduced in Newport, and extra staff were not hired early on because budgets were not available.

The total cost of failure was estimated at £21m.

The project ticked most of the boxes used to track failures: planning, management, training, contingency funding and strategy. Communication was poor too. The lack of a good PR plan gave the media a field day and for a long time the public had no idea why passports were taking so long to return.

Government IT failures are sadly all-too frequent but at least they offer a degree of transparency when the dust settles.

In the private sector it is generally agreed that IT failures are just as frequent, although often less costly, but they are shoved under the carpet and not open to scrutiny. However, when the result is obvious to customers it is less easy to put a lid on it.

This was the case for British Gas (owned by Centrica) in 2006/7. A huge number of customers complained to the company and the watchdog EnergyWatch about erroneous bills.

Centrica claimed the problems had been caused by a faulty SAP-based billing system named Jupiter, installed by Accenture at a cost of £300m. Centrica initially claimed there were “millions of errors” in the coding and filed for £182m in damages. We are still waiting to learn the outcome of this debacle as the case will be heard in court only next year.

In such cases online forums, including The Register’s, buzz with comment. Here are two comments made in August this year.

Anonymous Coward: “Judging by my experience with British Gas, I would guess that they: did not specify the requirement for the billing system properly; did not manage the implementation properly; then did not go about getting it rectified properly. After complaining to the regulator about their billing practices and getting compensation, I then switched to NPower.”

“The recurring theme is bad relationships"

Murray Hynd: “British Gas seems to need a little help in the following areas: understanding teeny weeny clauses hidden in the back page of the contract; being able to say no to significant changes from arrogant or incompetent stakeholders; avoidance of sharp objects. Can’t blame Accenture at all, being bean counters, sales and marketing, and lawyers first, corporate entertainment (for themselves) second, and a competent technology provider dead last.”

Accenture recently told The Register: “Centrica signed off on the design of the system and conducted extensive testing before formally accepting and using it. Centrica operated the system for over two years before starting the court process and it is still using the Jupiter system today.”

When the case is finally heard in court no doubt documentation of all aspects of the project, and particularly milestones, will be crucial, proving that planning, reporting and customer sign-off are a vital part of project management, especially when projects fail.

Finding the real cause of failure can be something of a forensic activity. Sometimes the problem is cultural. For example Bob Walker, Microsoft Project technical sales specialist, says: “We have found that many government clients don’t like the traffic light warning system. They think ‘red’ is a sign of failure, so you find them using other pastel colours. To us it’s not a sign of failure, it’s a flag that a decision has to be made.”

Roy Staughton, managing director of consultancy Shape International, is convinced that human factors are the most common cause of failure. “The recurring theme is bad relationships. You even have to look closely at the contracts to see if performance indicators for the quality of the relationships are built in. The models for project management often ignore the importance of soft skills,” he says.

And lastly, scope creep and change control are the familiar demons that stalk many projects.

Franco Valente, business manager at Prestwick Centre, NATS, the flight control centre which was recently awarded Project of the Year by the Association for Project Management, says: “Projects often fail when people don’t start with a clear understanding of what they need, so it is a real challenge to manage scope creep and change control effectively. Without a clear understanding among team and stakeholders about the ‘what’, ‘why’, ‘when’ and ‘how’ timescale, budgets and outputs are often wrongly identified.”

The more we look at failures, the more we learn that the fundamentals of project management are often ignored for one reason or another. The challenge for the profession is not so much spelling out those fundamentals but understanding the reasons why success remains so elusive. ®

Sponsored: Minds Mastering Machines - Call for papers now open

Biting the hand that feeds IT © 1998–2018