An AI strategy your board can approve.And then hold to account.
A board wants to know where the money goes, what comes back, what could go wrong and who answers for it. This work produces that answer.
The request usually arrives like this.
- “The chair has asked for our AI strategy at the next board meeting.”
- “We have a long list of use cases. We don't have a strategy.”
- “I can describe what we're doing. I can't say what it's worth.”
- “The non-executives want to know how we're governing this.”
A strategy is a short list of choices
Most documents called an AI strategy are inventories. They describe what is possible and what is under way, and they leave the choosing to later.
A board needs the opposite. It needs the few changes that would move the business, the organisational change each depends on, the risks, and a plan for the first ninety days with people’s names on it. It also needs to see how firm each number is, because a hypothesis and a validated case deserve different amounts of money.
What the board needs to decide.
A strategy is a set of choices. These are the ones a board paper has to make explicit.
- 01
Which few opportunities to fund
Ranked by expected value, with revenue, cost savings, released capacity and customer outcomes kept separate.
- 02
What the organisation must change
The operating-model consequences of those choices. Roles, teams, workflows and decision rights, and the boundaries between people, AI assistance and automation.
- 03
How it is governed
An AI governance and risk view, covering accountability, regulation and the assumptions the case depends on.
- 04
What happens in the first 90 days
A mobilisation plan with named owners and decision gates, so approval leads to action.
Why board AI strategies fail.
A list where a choice should be
Thirty use cases with no ranking is a way of avoiding the decision.
Benefits that blur together
Time saved is presented as money saved. The numbers do not hold when finance examines them.
No confidence levels
Hypotheses and validated cases look the same on the page, so the board cannot tell how much weight each will bear.
Approval without mobilisation
The paper is noted. Nothing has an owner or a date.
This is the Decide stage, done properly.
Stage 01
Decide
The Workshop aligns leaders, the Diagnostic sizes one or two areas, and the Deep Dive covers the whole business and produces the board pack.
Stage 02
Design
If the board approves, the operating-model outline becomes a detailed design.
The Deep Dive is built for this.
Two weeks, a fixed fee of £12,500, and a pack you can put in front of the board. It contains prioritised opportunities with value hypotheses and assumptions, an operating-model outline, a 90-day mobilisation plan with owners and decision gates, an AI governance and risk view, and a board recommendation. If leaders are not yet aligned, start with the one-day Workshop.
Engagement 03
AI Transformation Deep Dive
£12,5002 weeks
- An operating-model outline: roles, teams, workflows and decision rights
- The boundaries between people, AI assistance and automation
- A board-ready 90-day mobilisation plan with owners and decision gates
- An AI governance and risk view
Also worth a look: AI Opportunity Workshop (£2,500, 1 day).
Questions we hear.From people in this spot.
How long does it take to produce a board-ready AI strategy?
The Deep Dive takes two weeks and covers the whole business or several functions. A narrower question, on one or two areas, fits the one-week Diagnostic.
What is in the board recommendation pack?
Prioritised opportunities with value hypotheses and assumptions, a readiness, risk and dependency view, an operating-model outline, a 90-day mobilisation plan with owners and decision gates, and an AI governance and risk view.
Will you present to the board?
That can be agreed in scoping. Continuing board-level support is available through CEO, Board and Investor Advisory.
How do you put a number on AI value?
Each opportunity states the expected outcome, current performance where evidence exists, and the assumptions. Evidence gaps and confidence are recorded against every estimate. Time saved is treated as capacity, and becomes a cash saving only when spending is reduced or avoided.
What if the honest answer is that we should do very little?
Then that is the recommendation. We sell no software and take no payment by results, so there is no reason to recommend more than the evidence supports.
Related reading
TeamCEOs, founders & boards
Investment decisions, value cases, a board-ready plan and independent challenge.
Use caseAI governance
Accountability, Consumer Duty, UK GDPR and the EU AI Act, built into the design.
Use caseAI operating model
Workflows, roles and decision rights redesigned around AI, with the people boundary made explicit.
Start with a conversation.Thirty minutes, free.
Tell us what you are trying to decide. If we can help, we will say how and what it would cost. If we can't, we will say that too.