Would we get better design if we paid for it differently?

Here's a question worth sitting with for a moment.

If you commissioned a structural engineer to design a bridge, and paid them by the hour, what would happen if the bridge turned out to be unbuildable? You'd pay for it to be redesigned. By the hour again. And if the redesigned version still had coordination problems with the drainage engineer? More hours. Nobody loses money but the client.

That's not a cynical take on the profession. It's just what happens when the commercial model has no mechanism for connecting payment to value.

Nine out of ten major infrastructure projects go over budget. The academic evidence for this is substantial and consistent, most thoroughly assembled by Professor Bent Flyvbjerg*  when at Oxford, whose database now covers more than 16,000 projects in 136 countries. Rail projects average 44.7 per cent cost overrun. Not occasionally. On average. And when researchers look for the root cause, failures in design sit at the top of the list**, ahead of material price variation, ahead of inadequate planning, ahead of scope change. 

The problem isn't that designers are bad at their jobs. The problem is that the payment model rewards the wrong things.

Where the value actually lives

Design value isn't evenly spread across a project. It's concentrated at the front, when almost nothing is committed and a single decision about how an asset is configured can shape its whole-life cost for decades. The concept design stage, where you're choosing what to build and why, anchored to the client's whole-life cost model and the operational life the asset needs to deliver, is where the leverage is highest. Change costs are lowest. The design space is open. Get it right here and everything downstream is cleaner, faster, more buildable.

Then there's the integrated design stage, where structural, civil, mechanical, environmental and drainage disciplines work through how this thing will actually go together, how it will be maintained, how it can be built without, say, closing a road for six months. Get that right and your working drawings are a mechanical translation of decisions already made. Miss it and your working drawings become a rolling negotiation between disciplines that haven't properly spoken to each other.

Under time and materials contracting, the bulk of the fee goes to the working drawings. That's where most of the hours are. But that's the stage of least strategic leverage. The concept and integration work that determines whether those drawings will be buildable, coordinated, and fit for a 60-year asset life? That attracts a fraction of the budget, absorbed into early stage fees alongside desk studies and traffic surveys.

A different way of thinking about it

What if the design fee were distributed in proportion to the value each stage creates?

The paper we've been developing at BPA, 'From Time to Value', proposes exactly that. Forty per cent of the design fee at concept design. Thirty per cent at integrated design. Twenty per cent at coordinated design. Ten per cent at working drawings and specification, with that share declining further as automation tools make production progressively less expensive.

The payment trigger at each stage is a two-condition test. First, the client must have provided complete information at the level that stage requires. You can't properly price a whole-life cost comparison if the client hasn't told you what the asset needs to deliver across its operational life. Second, the designer must certify, by name, through an integrated quality management system, that the output is complete, coordinated, and fit for purpose at that stage. If the certification is genuine, the payment is earned and the designer is protected. If it's not genuine, that's a different kind of problem, and the Fraud Act 2006 and Economic Crime and Corporate Transparency Act 2023 both have something to say about it.

The working drawings stage isn't diminished in this model. It just earns its proper share rather than carrying a disproportionate weight it never deserved. And as automation develops, model-based drawing production, AI-assisted coordination tools, generative specification systems, the cost of this stage falls. Token pricing at working drawings falls with it. The client benefits. The designer's margin holds because the efficiency gain comes from better tools, not from cutting corners. And the commercial centre of gravity of the project sits where it should: at the concept and integration stages, where the human made decisions that actually determine whether a project succeeds are being made.

The question worth asking

Would infrastructure projects produce better, more complete, more buildable design information if the fee were proportionate to where value is actually created?

I think the honest answer is yes. Not because designers don't care about quality under the current model. Most of them care a great deal. But because a payment model that rewards strategic thinking at the front of the process, and holds designers accountable for the completeness of what they certify, creates conditions where that care has commercial traction behind it.

That's what I'm building out. If it's a question you're thinking about too, I'd like to hear where you've seen the current model serve the project well, and where it hasn't.

Dr Martin Perks, FRICS, MICW, MAC | Black Pear Advisory Ltd | +44 7771 865271 |

*Flyvbjerg, B. (2014). "What You Should Know About Megaprojects and Why: An Overview." Project Management Journal, 45(2), 6–19.

**Cantu-Silva, A., et al. (2020). "Factors Influencing the Cost Performance of Infrastructure Projects." Applied Sciences, Vol. 10, No. 16, 5519.