The Contract Bought Every Cable. It Didn’t Buy a Working System. Peter Cross on Where Projects Are Really Lost.

Leader in Spotlight

From the Diary of

Peter Cross

Buy the components and you keep all the risk. Buy the outcome and your supplier carries it with you.

“A strategy is fine but what makes the difference is a controllable plan.”

Peter Cross on the cover of Executives Diary

There is a contract for an automated ticketing system in a European capital city that specifies the last metre of cable and the last connecting clip, for every station, every tram and every bus. Nowhere does it require anyone to deliver a working ticketing system. It was worth more than 20 million euros, and it collapsed.

Peter Cross read it as project management adviser to the bank funding the scheme. He has read contracts like it from the other two seats as well, as the supplier delivering the work and as the client carrying what arrives. Most people know one seat. Knowing all three tells you where the risk was really placed, and when.

The scale of the problem is not in doubt. In July 2026 the National Infrastructure and Service Transformation Authority reported on 189 of the UK government’s largest projects, carrying a combined whole life cost of £924.2 billion. Twenty-nine were rated green. One hundred and nine sat at amber, meaning significant issues still need management attention, and 34 were red.

Cross has spent his working life being called in after that point. His view is that the loss happened much earlier, and that it is usually written down.

Peter Cross at a desk with his books Management Consultancy: How to Win and Project Management Bootcamp
“So I wrote the type of books that I wanted to learn from”

The Thing People Resist Until It Has Already Gone Wrong

Ask Cross what he tells clients that they push back on hardest, and the answer arrives without hedging. “That they need to contract for the solution that they want and not its component parts,” he says.

That they need to contract for the solution that they want and not its component parts.

On what clients resist hardest

The distinction sounds administrative. It decides who carries the risk. Buy a list of parts and every gap between the parts and a working system belongs to you, which means the supplier can deliver everything it promised while you receive nothing you needed.

As the Project Management adviser to the funding bank, I reviewed the contract which was worth over 20 million euros. It specified down to the last meter of cable and connecting clip what was required, for every station, every tram and every bus. Yet the supplier had no stake in providing an operating system.

Peter Cross, project turnaround specialist

Change control was a nightmare, he says, the client had to take all of the risk, and when the contract collapsed it had already paid for large amounts of unusable hardware.

None of this is secret knowledge. The Cabinet Office told departments to set clear and appropriate outcome-based specifications when it issued the Construction Playbook in December 2020. The guidance exists. The parts lists keep getting signed.

The Room Where He Watched It Work

The first time Cross saw the principle earn its keep, he was working with the UK Treasury and an NHS hospital on a lease option for a patient information system. Three providers, three funding mechanisms, and a decision about how to pay.

I negotiated a five-year contract with the selected provider in which there was a minimum acceptable 90% service level each month, and bonus payments for overperformance. For the first three months of operation the provider failed to meet the minimum service level, so the hospital paid nothing, but because the provider had a major stake in achieving the client’s objective it put in major additional effort to meet the targets.

Peter Cross, project turnaround specialist

Read that for the mechanism rather than the story. Beneath the floor the invoice is zero, above it the supplier earns more, and its revenue moves with the hospital’s outcome every month without anyone opening a dispute.

Three months of nothing is a powerful memo.

That instinct was formed across the wider NHS work, and he quotes a developer from those years who put the stakes plainly: in most systems, getting it wrong costs money, and in health systems it costs lives. Work like that, he says, “teaches the need for certainty in your work and the ability for it to stand up to scrutiny.”

Stop the Work Before You Spend the Rest of the Money

A UK government department had let a development and operations contract for a major Internet of Things statistical data system. Cross arrived as an adviser and then became the director responsible for delivery. He found a breakdown in communications, spending more than one and a half million pounds ahead of delivered value, and a client threatening to sue for breach of contract.

The first move was to stop.

With support from the client’s executive sponsor we agreed that I should halt further development work, review, clarify and validate understanding of the client’s requirements. We also agreed to change the culture of separation in the project and establish very close liaison between the team members and the client’s statistical team. Over the next eighteen months the project was turned round in terms of client satisfaction and our financial performance was brought back on track.

Peter Cross, project turnaround specialist

The second move matters as much. The cure for the culture of separation was daily contact between his people and the client’s statisticians, a habit he credits to a former boss, John Kopij, who showed him that engaging with people was not communication but something far deeper. Liaison is not a status report. It is knowing what the other team is worried about before it becomes a change request.

Agile Is Useful. It Is Not a Universal Answer.

Cross is candid about who disagrees with him. Project management, he says, contains “an almost theological split” between those who believe every project should run in an agile way using a proprietary methodology, and those who hold that given the diversity of projects, one approach does not fit all. He states the other side at full strength. The Agile Manifesto, drafted in the Wasatch mountains of Utah in 2001, freed developers from control and fixed real communication failures of its era. His objection is narrower. It was not then and is not now a panacea.

The same argument is now being had about artificial intelligence, and he has changed his position. He expected it to be a major step forward. What worries him is the rush. Writing on LinkedIn, he listed what the current wave is discarding from fifty years of business IT: a computer never fixed a broken system, a demonstration is not a test, documentation is essential and not a luxury.

The evidence is not on the side of the rush. RAND, interviewing 65 experienced data scientists and engineers in 2024, reported estimates that more than 80 per cent of AI projects fail, twice the rate of IT projects that do not involve AI. The most frequent root cause its interviewees named was leadership failing to communicate which problem the thing was supposed to solve.

Which is the ticketing contract again, in a newer typeface.

This is the centre of his position rather than a side argument. “Whilst every new technology presents new opportunities and risks it also repeats many of those of previous innovations,” he says. What follows is that organisations should build on what is already known about implementing technology well, instead of relearning it painfully from scratch every time.

Whilst every new technology presents new opportunities and risks it also repeats many of those of previous innovations.

On every new technology

That conviction is why the books exist. He wrote Project Management Bootcamp: A Step-by-Step Guide to turn this kind of experience into guidance somebody can act on, with the theory carried by anonymised cases. “So I wrote the type of books that I wanted to learn from,” he has written, “yes containing the theory but brought to life with real experiences from the coal face.”

Three Turnarounds and What They Cost

Peter Cross in a grey suit standing beside a carved wooden fireplace

At Jacobs he took on a failing national transportation programme for a European country and then delivered a second one for the same client. At Capita he inherited a major software development and hosting contract the client was moving to terminate, and turned it into a profitable long-term relationship. Earlier, at PNC Consultancy, he ran the turnaround of multiple customer-facing systems for London’s Business Link.

There was a price. The hardest stretch of his career, he says, was three years working away from home when his daughter was very young, and he missed a lot of her growing up. Turnaround work happens where the project is, not where the consultant lives.

Both books are in use at universities internationally. Given a list of those universities to promote, what he chose to feature was a video review recorded by Dr Mariatul Liza, one reader explaining what Project Management Bootcamp is for.

Where Peter Is Now

Cross works from the Greater Oxford area, advising organisations on projects in difficulty, speaking at conferences, and writing for people entering project management and consultancy from other fields. Asked what he would be doing if he had never gone into this field, he says he would probably still be a manager in a 2,000-seat concert venue.

Animated Executives Diary vision graphic for Peter Cross: contract for the solution you want, not its component parts, with client and supplier joined by outcome, stake and liaison

On what comes next, he is watchful rather than gloomy: “AI, I thought it would be a major step forwards in computer systems, but it turns out that we have to learn how to cage the bear.”

The Playbook

The Peter Cross Playbook: Five Rules for Getting What You Actually Bought

  • Contract for the outcome, not the parts. Specify the working result, so the gap between the components and the outcome belongs to the supplier as well as to you.
  • Give the supplier a stake. A service floor with no payment beneath it, and a bonus above it, moves the supplier’s money in step with your result.
  • Halt before you spend the rest. Stopping to review and validate the requirement costs less than another quarter of building the wrong thing.
  • End the culture of separation. Put your team in daily contact with the client’s specialists, so problems surface as conversations rather than change requests.
  • Do not relearn what is known. Every new technology repeats the risks of the last one, so start from what has already been proved about implementing it well.

Share post:

Subscribe

Who should be on our next cover? Nominate a leader or yourself for a free feature on Executives Diary.

More Leadership Stories
Related

Darryl Heffline Says the Best Turnarounds Begin Long Before the Numbers Change

/* ==================================================================== ed-feature-paid.css: Executives Diary Spotlight feature...

You Were Coached on What Not to Say. Nicki Scolaro Bland Thinks That Was the Problem.

/* ==================================================================== ed-feature-paid.css — Custom CSS...

More Interview Rounds Do Not Equal Less Risk. Corey Babel Explains What They Cost Instead.

Why another round rarely makes a senior hire safer, and what a founder can do instead.