Consulting Desk · AI Opportunity Assessment
A scored portfolio
of what to attempt.
And what not to, this year.
Every candidate use case across the business, mapped to a decision or process it would actually change, and scored on business value, technical feasibility and the state of the data it would need. The bottom of the list is argued as carefully as the top.
You hold: the scored use-case portfolio, the three we would start with, sequenced and the stop list.
What you get, and why the bottom of the list matters
Knowing what not to attempt is the cheaper half of the output.
Most opportunity work produces a long list with the exciting things at the top and no argument underneath. That list gets a budget, then gets quietly abandoned eighteen months later. We score every candidate on three axes — the business value it would create, whether it is technically feasible now rather than in principle, and whether the data it needs actually exists in the form a model needs it — and we write the reasoning for the ones we set aside as carefully as for the ones we recommend.
Every use case has to map to a named decision or process it would change. If it cannot, it comes off the list. That single gate removes most of what makes AI portfolios expensive: capability projects with no owner and no consequence.
The output is a portfolio, not a pitch. It is written so that someone who was not in the interviews can pick it up, follow the scoring, and disagree with a specific number rather than with a feeling.
We talk to the people doing the work
Not only the people sponsoring the change. That access is a condition of the engagement, because the gap between how a process is described upstairs and how it actually runs is usually where the real opportunity — and the real obstacle — is hiding.
How it runs
Value × feasibility × data readiness.
Three axes, applied consistently, with the working shown. A score you cannot interrogate is worse than no score at all.
- EngagementAdvisory Sprint — 2–3 weeks
- MethodValue × feasibility × data readiness
- GateEvery use case maps to a named decision
- Access neededThe people doing the work, not only the sponsors
- Ends inA written position you can take to a board
Candidates, gathered widely
Interviews across functions rather than a workshop with whoever was free. Shadow candidates count — the spreadsheet someone already automated is evidence of demand.
Each one tied to a decision
What decision or process would change, who owns it, and what a better answer would be worth. No named decision, no place on the list.
Scored on three axes
Business value, technical feasibility today, and data readiness for what this specific case needs. Scores carry their reasoning, so you can argue with the number.
Sequenced by dependency
Some cases only make sense after another has run. The order matters more than the ranking, and the order is where most portfolios go wrong.
A shortlist, and a stop list
The three we would start with, and the ones we would not, each with the argument. Both halves are the deliverable.
The two lists that matter
What you hold, and what we will not do.
Both are in the engagement letter before you sign it. The second list is the one worth reading twice — it is where most disappointment in this market actually comes from.
What you hold at the end
- The scored use-case portfolio, with the reasoning behind every score
- The three we would start with, sequenced, and why in that order
- The stop list — what we would not attempt this year, and what would change that
- The decision or process each case is attached to, with a named owner
- A written position you can take to a board without translating it first
What we won’t do
- Reviews that cannot name the decision they inform
- Maturity scores with no evidence behind the number
- Reviews where we cannot talk to the people doing the work
- Findings written to survive the sponsor reading them
- Advisory priced as the front end of a build we have already assumed
What usually happens next
The top of the list gets tested before it gets funded.
An opportunity score is a hypothesis. A Feasibility Sprint turns the strongest candidate into a working proof of concept against an agreed evals harness in another two to three weeks — and can honestly come back recommending you do not build it.
See the Feasibility Sprint →Often paired with readiness
An opportunity list the organisation cannot support is a wish list. The AI Readiness Assessment scores the same candidates from the other direction — and the two together are the usual first engagement.
Before you commit
How the scoring works, and what happens if nothing scores.
How is a use case actually scored?
Value, feasibility and data readiness, each scored against the specific use case rather than against a maturity ladder. Every score carries its reasoning, so you can disagree with a number and see exactly what changes.
What if nothing scores well enough to do?
Then that is the finding, written down, and the engagement is complete. A stop list is as useful as a start list, and it is usually the cheaper half of the output.
Do we have to start with the top item?
No. What you hold is a sequence with the reasoning behind the order, not an instruction. Most clients test the top one or two against an evals harness before funding a build.
Bring us the decision, not the deliverable.
Thirty minutes with a founder. Tell us what you are trying to decide and we will tell you which service answers it — including when the answer is that you do not need us yet.
Fixed fee · Quoted in writing before we start · NDA available