Fixed scope
What the organisation can realistically do with AI, and what it cannot yet.
Organisational posture on AI and digital — before or beside a programme — including a prioritised use-case portfolio and what governance will support.
- 2–4 weeks · assessment + executive readout
- Scorecard · use-case map · 90-day roadmap
What prompts the call
- The board has asked what the AI strategy is, and the honest answer is eleven pilots, three of them duplicative, and no portfolio.
- Procurement is being pushed toward a substantial AI commitment and nobody can say whether the organisation could absorb it.
- Pilots keep succeeding and nothing reaches production, which is a readiness problem rather than a model problem.
- A programme is being scoped around AI and the data foundation underneath it has never been examined.
How it runs
Six dimensions — literacy, data, infrastructure, leadership, culture, governance. Workshops and interviews, then a readout to leadership. Vendor comparison included; vendor-agnostic.
Week 1–2
Six dimensions
Literacy, data, infrastructure, leadership, culture and governance — scored on evidence rather than on a survey. A self-assessment questionnaire records what people hope is true, which is a different thing.
Week 1–3
Workshops and interviews
The functions that would actually use it, leadership, and whoever owns the data. Where pilots already exist they get examined properly: what was learned, what it cost, and whether anything reached production.
Week 2–3
Use cases, collected then cut
A prioritised portfolio — value against feasibility, with feasibility scored against the readiness findings rather than against a vendor’s claim. Cutting is most of the value; a list of forty use cases is not a strategy.
Week 3–4
Readout
A working session with leadership, and a scorecard built to be uncomfortable in the places where it should be. Vendor comparison where a purchase is in question — I am not reselling anyone’s platform, which is the only reason the answer is worth anything.
What you get
On the table.
- A scorecard across the six dimensions, with the evidence behind every score.
- A prioritised use-case portfolio — value, feasibility, and what each one depends on.
- A 90-day roadmap: what to start, what to stop, and what to fix first.
- The governance that must exist before AI scales beyond pilots.
- Vendor comparison where a purchase is in question, vendor-agnostic throughout.
- An executive readout, and a version of the scorecard that can go to the board.
What I need from you.
- Access to whoever owns the data, which is frequently not whoever owns the AI initiative.
- Honest accounting of existing pilots, including the ones that quietly stopped.
- Two to three hours from each function that would use it, in a workshop rather than a survey.
- Leadership in the readout. The findings are decisions, and they do not survive being relayed second-hand.
When this is the wrong call
- You want implementation. This assesses and prioritises; building is separate work, usually with an implementation partner.
- A vendor has already been chosen and the assessment would be procurement cover for it.
- What you need is AI enablement training for teams. That sits with Applied Futures rather than here.
- There is no data function and no plan for one. The assessment will tell you that in week one — you can have the finding now, at no cost.
Questions
- Do you only recommend one AI vendor?
- No. I co-lead a Claude Code community in Copenhagen, which means I know Anthropic’s tooling well — but the advisory work is vendor-agnostic and frequently the answer is a combination, or an open model, or not buying anything yet.
- Do you do the technical implementation?
- Strategy, governance, vendor selection and team enablement. For heavy engineering I partner with implementation firms and stay on as the strategic advisor rather than pretending to be a delivery house.
- Can you introduce us to AI vendors?
- Yes, through the community network and industry relationships. I would rather do the assessment first, so that those conversations are specific enough to lead somewhere.
- Is this the same as the retired online assessment?
- No. The free tools that used to sit on this site were self-scored questionnaires. This is evidenced, involves your data and your people, and produces findings somebody can be held to.