Your AI use-case list has thirty items and no owner

Almost every AI strategy engagement produces a list. Thirty use cases, scored on value and feasibility, arranged on a two-by-two, with a roadmap underneath dividing them into now, next and later. It is thorough, it is defensible, and it is the reason nothing happens.

The list is not a strategy. It is a way of avoiding the decision a strategy is supposed to make.

Why the long list feels safe

It includes everybody. Each department contributed something, so nobody has to be told their idea lost. That is precisely the property that makes it useless — a document nobody disagreed with has not decided anything.

It defers the hard question. Ranking thirty items looks like prioritisation. It is not, because there is no capacity statement attached: nowhere does it say how many the organisation can actually do this quarter, which is the number that determines everything.

It survives scrutiny. A senior stakeholder cannot object to a comprehensive list. They can object very sharply to three specific commitments, which is exactly why the three-item version is the more useful document and the harder one to get signed.

It transfers no reasoning. A score in a table is not an argument. Six months later nobody remembers why item seventeen was rated a four, so the list cannot be revisited intelligently and is instead quietly replaced by a new list.

What a decision-shaped strategy contains

Three to five candidates, ordered. Not thirty filtered to a top tier — actually three, with the rest explicitly rejected and the reason recorded.

A named owner per candidate. Someone accountable for the outcome, not for the deliverable, with the authority to make calls. A candidate without an owner is a wish.

A number, with its assumptions visible. What it takes to build, what it costs to run, and what it saves or unlocks. The numbers will be wrong; the assumptions are the point, because those can be challenged and updated.

A statement of capacity. How many changes the organisation can absorb this quarter, given what is already running and who is on it. This single paragraph changes more plans than any scoring model.

A definition of success. What will be measurably different, written so that in six months nobody argues about whether it worked.

The rejections, with reasons. The most reread section in any strategy document we have written, because it stops the same ideas returning every quarter with a new sponsor.

The exclusions are the work

A useful strategy is mostly a record of what you decided not to do. That is uncomfortable to present, because the excluded ideas belong to people who are in the room.

But an organisation that has genuinely rejected twenty-five things has made a decision. One that has ranked thirty has produced a document. The difference shows up about four months later, when the first has two things running and the second has a steering committee.

How to tell which one you have

Take your current AI plan and ask three questions of it. Which items would we stop if the budget halved? Who is personally accountable for item one? What would we see in six months that would tell us this worked?

If those have crisp answers, the plan is a strategy. If they need a workshop to answer, the plan is a list, and the workshop is the piece of work that was skipped.

That is the shape of what we do in an AI strategy engagement: a short ordered list with owners and numbers, and a longer record of what was rejected and why — including, more often than clients expect, the finding that the problem is a process problem and the answer is not AI at all.