The book How Big things get Done by Bent Flyvbjerg and Dan Gardner is an excellent book in how to deliver large projects and, even more importantly, how projects can go wrong in spectacular fashion with lots of interesting and eye-watering examples, writes Fergus Wheatley. "The book has nine chapters; the children’s hospital project could easily be the cautionary tale in six of these chapters". 

"We are about to spend €110bn on infrastructure in the next five years, it is vital that this money is spent wisely. This is a once-in-a-generation sum of money and represents an investment of €52,000 per person living in the republic. The state is borrowing the equivalent of a small mortgage of €208,000 for every family of four living in Ireland. 

"I am hoping that the lessons within this book will resonate with those responsible and all stakeholders involved in building out this infrastructure." 

Examples and how they apply to NCH project

Lesson 1: Don’t believe numbers given by vested interests

“As simple as it should be to judge projects, it is anything but. In every big project there are blizzards of numbers generated at different stages by different parties. Finding the right ones – those that are valid and reliable – takes skill and work. Even trained scholars get it wrong.”

During a hearing of the Dáil committee that was looking at the children’s hospital overspend, Stephen Donnelly repeatedly asked what the final figure was, he was then met with fudge and obfuscation. 

According to the authors, this is quite normal, there is a section in Chapter 1 called 'Honest Numbers', but the introductory quote really nails it.

Lesson 2: Project over-runs are the norm

Of 16,000 projects from 20 different fields in 136 countries that were studied, only 0.5% came in "on-time, on budget and delivered the benefits that the project set out to achieve". 

The mean cost over-run of a big building project is 62%, but this number still hides the real danger, large projects have what is known as a "fat-tailed" risk. The ordinary bell curve standard deviation for cost over-runs goes way out to the right, this means that a cost over-run of 1,000% can easily happen.

The actual distribution curve for cost over-runs is not a standard bell curve, but looks more like this…

Lesson 3: Beware of the Commitment Fallacy (Ch 2)

"Dig a big hole, and make it so expensive that it is impossible to not continue the project". 

Strategic misrepresentation is the tendency to deliberately and systematically distort or misstate information for strategic purposes. 

If your objective is to get a project approved, superficial planning is quite useful because it glosses over big challenges, which keeps the estimated cost and time down which, in turn, wins contracts and gets projects approved. 

But challenges ignored during the planning will eventually boomerang back as delays and cost over-runs but, by then, the project is too far along to turn back.

Getting to the point of no return is the real goal of strategic misrepresentation, it is politics resulting in failure by design.  

In the case of the children’s hospital, it was even worse. The rush to get shovels in the ground and dig a big expensive hole was so acute that this was being carried out before detailed planning had even started.  

As Flyvbjerg and Gardner note: "Projects don’t always go wrong, they often START wrong"... "on project after project, rushed superficial planning is followed by a quick start that makes everybody happy because shovels are in the ground. But inevitably the project crashes into problems that were overlooked or not seriously analysed…. People run around trying to fix things… more things break…" followed by even more (expensive) running around.

Lesson 4: Start with the problem you’re trying to solve (Ch 3)

"Projects are often started by jumping straight to a solution, even a specific technology. That’s the wrong place to begin. You want to start by asking questions and considering alternatives. At the outset always assume that there is more to learn. Start with the most basic question of all: why?"

I strongly suspect that the 'why' of the children’s hospital is easy: hubris. “The largest children’s hospital in the world, built in a complex city-centre location with world-class architecture and the most advanced digital record system in Europe”. This was always going to be a legacy for some ego-driven individuals.

The whys of treating sick children should have been central to this decision. A greenfield site, near transport nodes, using a proven design (eg, copy a hospital already built somewhere else) would have been a much better starting point.

As the book notes: "Steve Jobs once notably remarked 'you’ve got to start with the customer experience and work backwards'. Jeff Bezos gets his project teams to write the final press release and product FAQs before a line of code is written." 

Lesson 5: Copy the iterative and testing process formalised by Pixar (Ch 4)

Chapter 4 discusses methods and strategies of getting large projects to perform. It develops on the iterative process used by Pixar in planning movies. 

Unlike traditional planning, which often relies on rigid, linear processes, Pixar embraces continuous refinement, collaboration, and adaptation – allowing creativity and problem-solving to evolve organically. Mock-ups, storyboards are all tested rigorously with focus groups before work starts on the project proper. At this stage, the story (design) is completely nailed, so expensive rework is avoided.

Frank Gehry also used this approach when planning the Guggenheim in Bilbao; detailed models were built for all aspects of the project, so that stakeholders could offer insights at planning stage rather than during the building.  

The children’s hospital, in contrast, had many late design changes. While on the outside looking in can individually seem quite trivial, in reality, it messes up contractor sequencing and scheduling, leading to knock-on delays and costs in unrelated areas. 

Lesson 6: Prioritise experience on the whole team (Ch 5)

The lesson from this chapter is: "if delivering a project on time and within budget is your priority, then don’t be the first, biggest, custom or bespoke". 

Repeating what has gone before with a master builder who has done it previously drastically cuts the risk of a project going wrong. Simply copying another hospital anywhere in the world would have bypassed most of the expensive and disruptive design changes that plagued the children's hospital project.

Experience is mistakenly under-rated (underweighted on e-tenders) and de facto can be sidelined during the early excitement that results from the initial conception of a large project (normally with other people’s money).  

If you’re going to be the 'first, biggest, tallest, most advanced' (or other superlative) then you by definition end up marginalising experience and dramatically increase the risk of a project going wrong. Even experienced builders/project developers then have to learn as a project develops.    

To labour the point, the children’s hospital was going to be the "biggest" and the "most advanced" in the world, but also built on a complex cramped site with extensive excavation required, so no wonder it went sideways.

Lesson 7: Don’t assume your project is unique (Ch 6)

Flyvbjerg and Gardner challenge the common assumption that large-scale projects are fundamentally unique.

They argue that this belief often leads to unnecessary risks, as teams dismiss valuable lessons from similar past projects. Instead, the authors advocate for a mindset of 'modular thinking' – breaking down seemingly one-of-a-kind endeavours into familiar, repeatable components.

By recognising patterns, leveraging best practices, and learning from comparable cases, project leaders can reduce uncertainty and improve outcomes. The chapter underscores that no project is entirely original, and success comes from embracing – not ignoring – the wisdom of precedent.

Successful projects do not rely on heroic assumptions but on hard data from comparable experiences. The authors introduce the "reference class" concept as a powerful tool for improving project planning and avoiding costly over-runs.

Instead of treating a new project as entirely unique, planners should identify a "reference class" – a group of comparable past projects – to predict outcomes.

For example, if building a new subway line, analyse data from similar subway projects worldwide. These reference class projects will capture the experience of other projects, average cost and delay over-runs, and give valuable insights into the infamous "unknown unknowns" that previous projects have experienced. 

Most project failures stem from underestimating complexity and overconfidence. The reference class method forces planners to confront historical realities, reducing bias and improving accuracy. Flyvbjerg’s research shows that projects using this approach are far more likely to stay on budget and schedule.

Lesson 8: Don’t let 'Survivor Bias' trick you into thinking that sometimes ignorance can be your friend, and to then 'Just Build It' (Ch 7)

The book gives a myriad of examples, but the one that stood out most for me was the Sydney Opera House. 

The project began with a rough conceptual design from a young Danish architect Jørn Utzon, but the intricate details – particularly the iconic shell-shaped roofs – had yet to be fully engineered.

After winning the design competition, this crucial missing engineering, led to a continuous process of redesign and rework as engineers grappled with unprecedented architectural and structural challenges. Utzon was fired mid-project and left it in disgrace and despite living to be 90 years old, never built another iconic building. The building was eventually finished 10 years late and 1,450% over budget. 

Survivors Bias means we only hear about the successes and not the hundreds/thousands of failures, to use a pithy quote from the book, "the typical leap in the dark ends with a broken nose".  

Lesson 9: How to beat the odds and make a project successful (Ch 8)

Heathrow’s Terminal T5 faced a daunting challenge before it was built, a badly congested site, tunnels, roads, parking, train stations, baggage handling, air traffic control all had to come together and work. Using the 'reference class' method, the project was likely to come in more than £1bn over budget, which would sink BAA.

With its back against the wall, BAA used three strategies:

1. T5 was planned to extreme using detailed digital models (much like the Pixar methodology). These models were used to run simulations on everything from aircraft arrivals and departures, passenger interactions, baggage handling, traffic/parking access and egress. The models were constantly tweaked and the simulations rerun until a final design was ready. 
2. Advanced computer aided models and design also meant that building components could be built off site, shipped to site to be assembled, using a process called 'design for manufacture and assembly'. This dramatically reduced the work required on the congested site and made buildings'' components more accurate, so rework was vastly reduced.   
3. Finally, the BAA team spent a lot of time getting the building team right. BAA actively led and shared risks so that intercontractor disputes were nipped in the bud. Staff facilities were excellent, and workers felt empowered.  

All contractor staff were told time and time again that their first allegiance was to 'Team T5' and the resources were put in place to foster this. Clean showers, first class canteens and free replacement PPE available to all staff (including subcontractors) which was central in getting the 'team' mentality right. (Contrast this to the infighting and court cases surrounding the children’s hospital.)

Lesson 10: Design your project to be 'Copy/Paste/Assemble' (Ch 9)

The chapter encourages a fundamental shift in thinking. A "LEGO mindset" involves:

1. Instead of viewing each project as a unique challenge requiring a unique solution, the Lego mindset seeks out opportunities to apply repeatable, standardised processes and components. Eg, modular bathrooms are built in an off-site factory, brought to site, put in place and five connections plugged in, (hot/cold water, foul/grey waste and power).
2. From Bespoke Art to Repeatable (Copy/Paste) Engineering: it moves the focus from artistic, one-of-a-kind designs to pragmatic, efficient, and well-tested engineering. The authors note that while iconic, the Sydney Opera House's design was a bespoke creation that became a logistical nightmare. In contrast, modern skyscrapers often use a standardised steel frame and prefabricated facade modules, that leads to far more predictable outcomes.  

Conclusion

Lessons from How Big Things Get Done include: plan slow, act fast; think 'right to left' by starting with your goal; break big projects into 'Lego' components; use reference-class forecasting to improve predictions by avoiding assumptions of uniqueness; manage stakeholders effectively; create capable delivery teams; prioritise ruthlessly by saying 'no' to unimportant tasks; identify and mitigate 'unknown unknowns'; build in modularity to reduce risk; and prioritise speed by keeping project timelines short to minimise the impact of 'fat-tailed' risks.

Author: Fergus Wheatley BSc(eng) MIEI CMVP, CEA, CEM. http://www.smartpower.ie and http://www.linkedin.com/in/ferguswheatley