Independent review

Digital and AI investments are usually approved on evidence prepared by people who gain from the approval. The supplier wants the contract and the delivery team wants the programme. Their evidence tends to show what each of them needs it to show, without anyone being dishonest. A review tests that evidence on the owner's behalf.

A decision that affects several developments has a different problem. Each development reports against its own budget and programme. What the decision means for the others appears in nobody's report. It usually comes to light when it is too late to change. The review asks whether the portfolio still lands within its range after the decision. If it does not, the review sets out what can still be moved.

When it applies

A review fits a decision point of one of two kinds.

A digital or AI commitment

• A supplier's proposal for a digital or AI system, before you accept it.
• A digital twin or data platform, before you fund it or fund it again.
• A digital asset, before you accept handover into operation.

A decision that affects the portfolio

• A change of plan at group level, such as a pause or a re-phasing, before anyone has worked out what it releases or delays elsewhere.
• A portfolio forecast to miss its range, where the owner needs to know which developments can still make up the difference.
• A business case approved some time ago whose assumptions have since moved, such as a utility connection or a handover date next door.
• A committee choosing which phases to fund within a limit, where each case was made on its own and none has been compared with the others.

What is reviewed

The review works from material that already exists. That is usually the business case, the supplier's proposal, the information requirements and any assurance evidence. A small number of interviews are added, agreed and named before work starts.

For a decision that affects several developments, the material is the programmes and contracts each development already holds. Nobody is asked for new reporting. Much of the value comes from setting those documents side by side. A business case that assumes a March handover, next to a programme that now shows September, is the kind of gap a review is there to find.

The decision is then tested against five questions.

• What decision does this evidence actually support.
• What would have to be true for the claim to hold.
• What breaks it.
• What does it cost to keep true after year one.
• What else it commits, and the date after which that cannot be changed.

What you receive

You receive one written opinion, short enough to put in front of a board and specific enough to act on. It sets out what the evidence supports and what it does not. It states the conditions under which the investment can be defended. It lists the risks that matter, ranked by severity. A one-page summary comes with it for your own paper. I present the opinion once to the sponsor or the board.

Where the decision affects several developments, the opinion also sets out what it commits elsewhere and by when. It shows which options are still open. It is about the decision and does not apportion blame between teams.

Every opinion carries a reliance statement on its front page. It limits the opinion to the decision reviewed, the material examined and the date of issue. An opinion, like a digital twin, is worth relying on only within stated limits.

How it runs

Scope and end date are agreed before work starts. A review typically takes 3 to 4 weeks from kickoff. The fee is fixed after a first call and does not change. There is no time-and-materials billing.

The independence rule

On any review I take no delivery work and no supplier-side work. I have no stake in the outcome. Any interest is declared before the engagement starts. If a conflict cannot be cleanly separated, I decline the engagement.

Out of scope

A review does not include implementation or delivery resource. It does not run a supplier selection, though a completed selection can be reviewed. It carries no certification or regulatory sign-off. The supplier's technical work is examined and is not redone.

Common questions

Is this an audit?

An audit tests compliance against a standard. A review gives an opinion on whether a decision is supported by its evidence. That evidence is usually a familiar set of documents, such as a business case, a supplier's proposal, a pilot's results and a set of benefits claims. The review tests whether those documents answer the questions the decision turns on. It carries no certification and can run alongside formal assurance.

Can a decision that has already been made be reviewed?

A decision already made can still be reviewed, such as a completed supplier selection or a platform that has been funded. The question changes from whether to commit to what must now be true for the investment to pay back. The method is the same.

Can you review a procurement before it goes to market?

This is one of the most useful moments to look. By the time bids arrive, most of the outcome has been set by what the tender asked for and how it will be scored. A review at this point examines the technical scope and the evaluation criteria. It asks whether the contract should be bundled or decoupled. It also asks whether what is being bought fits the architecture and data it will depend on. Fixing a tender costs far less than managing a wrong award.

Does this need new reporting from my teams?

No team is asked for new reporting. The review works from the material each development already holds. Nothing has to leave your organisation. The review can run on your own systems under your own confidentiality terms if that is simpler.

You spent years on the supplier side. How is the review independent?

For 17 years I helped build the evidence that sellers put in front of infrastructure owners. I know what that material is written to show and which questions it is not built to answer. Independence is written into each engagement. There is no delivery work behind the opinion. Any interest is declared before work starts.

Do you take supplier-side work?

I take no supplier-side work on anything under review. If a conflict cannot be cleanly separated, I decline the engagement.

What does it cost?

The fee is fixed and proposed after a first call. There is no published rate because scope is agreed for each decision. Once the fee is agreed it does not change. There is no time-and-materials element.

Why one person and not a firm?

A firm that reviews a proposal is often in a position to sell what follows, such as the implementation or the programme office. I have seen that from the inside. A review here has no delivery team behind it and no follow-on work it is priced to win. The opinion is the whole of what you buy.

Beyond the review

Continuing advice can follow a review where an owner wants it. The other ways to engage are described under Work. If a review is not the right shape for you, say so in your first message.

Start here

Tell me the decision you are facing and when it has to be made.