Why megaprojects can fail, and how to start thinking about the fix

Read the second article of our four-article series on megaprojects by Connor Butler, Billy Glennon, Chauncey Bell and Charles Spinosa.

Shock, horror: megaproject goes over budget

In our first article, we looked at why megaprojects (capital projects costing more than $1billion) are so important to communities and civilisation. In this follow-up, we’ll explore why they so often fail to be delivered on time, on schedule and with all the promised benefits. Indeed, according to Bent Flyvbjerg, the Danish economic geographer who has tracked megaprojects over the past 90 years, only 0.8% meet their initial targets and expectations in full.

Our view is that this figure of 0.8% is indefensible. Megaprojects require a very different sort of leadership and way of thinking, but they can be delivered on time, on budget and with all the promised benefits. We know, because we’ve done it. So as well as looking at why megaprojects can fail, this article will also show you how to start thinking about the solution.

Our next two articles will reveal that solution in more detail, then explain why it works. And the insights that we share here can also be applied to business-critical capital projects of any scale.

When a megaproject blows its budget or its schedule, the first thing that the media, the public and sometimes those leading the project themselves often seek to do is to find someone or something to blame. The finger is often pointed at the estimators. Most theorists put it down to the overconfidence bias, the optimism bias, or the planning fallacy (where planners do not adequately assess risks.) There is also ‘strategic misrepresentation’ —otherwise known as lying—which is sometimes excused by the so-called ‘hiding hand’ thesis of Hirschman.¹ Hirschman’s theory is that, while we are over-optimistic about what we can do, we are also overly skeptical about how creative we can be in overcoming things that go wrong.²

Bent Flyvbjerg, the Dean of Megaprojects, debunks Hirschman, but Flyvbjerg’s solution is no better than Hirschman’s magical handwaving. According to Flyvbjerg, we should debias planners and estimators.³ Welcome to your summer re-education camp.

It’s not a camp we’ll be attending though. Having had the experience of delivering megaprojects and capital projects of significant scale on time, on budget and with all the promised benefits, we know that none of the above answers truly hit the mark. Megaprojects don’t go wrong because of a poor schedule or budget, or because of a lack of clarity or control. The fundamental issue that shows up to us again and again is that the systems through which these projects are delivered is not designed to deal with the associated complexity. And when you try to overcome that with more control (e.g. a more detailed schedule or tighter management), you simply end up stopping the flow of a project.

Imagine a factory where everybody has their own workspace, does their thing and then passes the widget along to the next person. Now imagine that the workspaces overlap and people start bumping into each other and sometimes they pick up the wrong tool, and they have to improvise. That’s what a megaproject is like. The conventional solution would be to increase control by bringing in more people to manage the factory workers and make sure that they bump into each other less. But that’s no solution at all. It stops the flow, stifles innovation and encourages budgets and schedules to overrun. In the simplest of terms, the true solution lies in changing the way that those workers view and interact with each other. That’s the solution that we will build on over these four articles.

Intel and BAA

Our favorite megaproject is one we worked on with Intel to build semiconductor fabrication facilities (fabs). About 10 years ago, Intel was facing a crisis. With chip technology developing according to Moore’s Law (the observation that the number of transistors in a dense integrated circuit doubles about every two years), chip manufacturing had to develop even faster. 10 years ago, upgrading a fab would cost about $1 billion (including the tools installed). Today, a complete new fab costs in the range of $7-10 billion. At the rate that fab costs were increasing, Intel had to contemplate leaving the chip-making business altogether, and had already given up one major contract.

However, using a version of our Integrated Project Delivery and Commitment-Based Management™, we helped Intel and its suppliers cut the cost of its fabs by 30%, and complete plants at the scheduled time, and to the specified safety standards.

A similar approach has been taken with several other megaprojects, such as BAA’s construction of Heathrow’s Terminal 5. This took place between two of the busiest runways in the world, near the busiest roadway in Europe, and was subject to intense environmental management conditions. Early modeling revealed that a truck needed to make a delivery every 10 seconds onto one of the highest security sites in the world. BAA, the owner of the project, took a long-term view of these conditions and invested in experts outside a traditional construction project team to develop a logistics strategy that could meet the unique challenges of the project. The project was an overwhelming success, as was the investment in logistics infrastructure. Logistic costs on Terminal 5 were estimated at 7% of the total cost of work versus industry benchmarks of 12%.⁴

Running a megaproject is like running a small town. You can’t control 50,000 people, create a schedule for what they’ll be doing next week or micro-manage their spending. Megaprojects require a different way of thinking and a different style of leadership. In the next article, we’ll reveal this big fix in detail.

If you are involved in a megaproject or smaller capital project, our thinking and insights can help you now.

Read the other short articles in this series or, for a complimentary copy of our full report How To Fix Megaprojects (And All Capital Projects That Matter), email Mark Crampton.

Listen to our Megaprojects podcast series.

Contact the authors:

Notes

1 B. Flyvbjerg (ed.) The Oxford Handbook of Megaproject Management. Oxford, UK: Oxford University Press, pp. 159 & 185.

2 Flyvbjerg, pp. 157-158.

3 Flyvbjerg, p. 185.

4 We recommend applying new tools such as Touchplan to integrate the delivery of construction on a scale impossible with traditional critical path approaches. This technology allows decision makers at the point of work to understand their impact on each other in real time and also allows Project Management Teams to monitor real progress, not guesses and updates in GANTT charts.