Contents · 7 min read
01 / 05 · Effort
What pilot counts and AI spending tell the board
In short
Pilot counts show that work has begun. AI spending shows what management has backed. Neither tells the board if AI makes the company stronger or weakens why customers choose it.
Take a vertical software company with ten years of customer interaction data. Its board wants to know whether AI will strengthen the firm. Management points to pilots and AI spending. The product team also ships some feature work faster.
Management points to its pilots. It also shows the money spent on AI tools. The product team now ships some feature work faster.
That work may be useful. Faster feature work is a real gain. The firm is right to track it.
But the board asked about the company’s place in the market. A project list cannot answer that. As AI strategy for boards makes clear, spend and pilot counts show effort rather than position.
For the software company, a long pilot list may hide the key risk. AI could help the firm make more of the very thing that is losing value.
The next step is to test what customers buy.
02 / 05 · Value
Does AI make customer value common or make you better at it?
In short
Start with the value customers buy from you. Then ask what AI does to it. AI may make that value easy for many firms to offer, or it may help your company build an edge that grows through use.
For the software company, run these six tests. Use facts from sales calls and product work. Do not turn them into one score.
Name what customers buy
Do they buy a long feature list? Or do they buy help with hard work in their field? Write the answer in one line.
Test the price
If rivals can match the offer with common AI tools, the old price has less support. If the product helps with rare domain work, the price has a stronger base.
Name the reason customers choose you
If the answer is feature count, AI may weaken it. If the answer is deep knowledge of their work, AI may help the firm use that knowledge at scale.
Check where customer knowledge lives
Does it stay in calls and staff notes? Or does each useful fact become part of how the product works?
Ask what grows with use
More users alone are not enough. Each customer should help the firm learn more about the work it serves.
Reopen the old belief
The firm once assumed that more features meant a better product. What proof would show that this belief no longer holds? A good AI board memo should show what the company can repeat and where its plan still breaks.
The answers show which side AI is helping. For this firm, that puts its feature set under the next test.
03 / 05 · Features
Generic features weaken the reason customers choose you
In short
AI helps software teams ship more features. It also helps new firms do the same. When similar features become common, feature count gives customers less reason to choose one firm over another.
For the software company, faster product work first looks like a clear win. Its team can move through the roadmap at greater speed. More requests can reach the product.
That gain still matters. Customers expect the product to keep up. The firm cannot stop its roadmap.
The turn comes when rivals gain the same speed. A new firm can start with newer systems. It can also ship common AI features fast. The old firm’s longer feature list then gives less cover.
For the software company, adding an AI label to each feature does not change this. Generic AI features are easy to match. Shipping them sooner may buy time, but it does not build an edge that lasts.
The same point appears in Engineers Went First: saving time is not an edge when every firm can do it. The key question is what the company builds with the time it saves.
The firm now needs an edge that grows as those features spread.
04 / 05 · Knowledge
Customer knowledge can become a capability that grows
In short
The firm already knows things a new entrant does not. The gain comes when that knowledge stops living in a few people and starts shaping the product. Each customer can then make the firm better for the next one.
For the software company, ten years of customer contact is raw material. Yet much of it may sit in call notes. Some may live only in the heads of product staff.
The firm can encode the reasoning behind its product. Why does this data model exist? What does a regulator care about? Where do users tend to fail?
Those answers go beyond a feature list. They show how work in the field gets done. They also help the product give better support in the next case.
Each new customer can add to that base. A support case can expose a missed rule. A product choice can show which part of the work matters most. The lesson should stay in the company after the person leaves.
For the software company, this changes what grows through use. The feature set may still grow. More important, the firm’s grasp of customer work grows with it.
Management must know enough about AI to judge that shift. It cannot do so from theory alone. An AI strategy workshop grounds the choice in real use and a ranked set of moves.
That leaves one job for the CEO: move the money toward it.
05 / 05 · Investment
Shift investment before competitors reposition
In short
Keep using AI to speed up the roadmap. But move the main bet from feature count to the knowledge that makes the product hard to copy. Do it before another firm builds that position first.
For the software company, this is a choice about funds and staff time. The current plan gives feature delivery first claim on both. The new plan must also build the firm’s shared domain knowledge.
The company can move in five steps:
The CEO names the fading edge
State which part of the old feature lead AI is making common. Put that view in the board paper.
Product leaders record domain reasoning
Capture why key product choices were made. Record what customers do in practice, not just what they ask for.
Customer teams feed the shared base
Turn useful calls and support cases into input for the product. Give one leader the job of checking that this happens.
Finance shifts part of the plan
Fund work that makes customer knowledge reusable. Repeated budget choices train a company to change or defend the old model, as The Margin That Blinds You shows.
The board asks for proof
Show what the firm can now do more than once. Show where the new way still fails. Then name the choice due next.
For the software company, speed is an asset here. A mid-market CEO can decide without waiting for a vast program. A small leadership team can change how it works, which is the case made in AI consulting for Nordic companies.
The board should see proof of a new edge. It should see customer knowledge staying in the firm. It should also see that this knowledge gets better through use.
Take that proof to the board before the next pilot list.
The next decision
The task now is to test whether AI is strengthening what customers buy or making the company’s current edge common, then decide which advantage deserves the next investment. GRAIL’s five-minute assessment applies that question to your company through a short set of questions about the business. The personalized report shows where AI could give the company an edge, giving the CEO a concrete basis for deciding what to examine next.
A short set of questions about your company. Your personalized report on where AI could give you an edge arrives by email.
Questions CEOs ask about this
How can I tell whether our company will be a winner or a loser in the age of AI?
Your company is an AI winner when AI strengthens what customers buy. If AI makes that value common, faster work only weakens your edge. Winning requires an advantage that grows with each customer. The CEO must decide which edge deserves the firm’s next investment.
Which parts of our value are becoming common?
Start with features that a new firm can build with common AI tools. Then mark any know-how that customers can now get without your product. Those parts can no longer carry the old claim to be different.
Does AI strengthen or weaken our current edge?
If your edge is feature count, AI is likely to weaken it. If your edge is stored domain knowledge that shapes the product, AI can help you use it more often and improve it through use.
What happens to our pricing and delivery model?
A common feature gives the old price less support. Base more of the offer on the hard customer work your product helps solve, then test if buyers value that difference.
Which capabilities will grow over the next three years?
Back the ones that gain from each customer. For the software company, those are shared domain reasoning and a deeper record of how customers work. A skill held by one person does not grow in the same way.
What would show that our current strategy is no longer enough?
Watch for rivals matching key features faster than your plan can defend them. Also watch for customer knowledge that keeps leaving with staff or dying in call notes. Either sign should force a new choice from the CEO and board.
Founder, GRAIL
Johan has spent fifteen years in management consulting and executive leadership, including as Chief Product Officer of a SaaS company. He founded GRAIL to help European leadership teams work out what AI changes at the core of their business. He is the author of The Execution Revolution and the host of the ThinkRoom podcast.
Related reading
Sources
- AI strategy for boards: why spending and pilot counts measure effort rather than market position. (accessed October 8, 2026)
- AI board memo template: what the board should see about repeatable work, weak points and choices due next. (accessed October 8, 2026)
- Engineers Went First: why saving time is not an edge when every company gets faster. (accessed October 8, 2026)
- AI strategy workshop: why a sound answer rests on real use and a ranked plan. (accessed October 8, 2026)
- The Margin That Blinds You: how repeated budget choices train a firm to change or defend its old model. (accessed October 8, 2026)
- AI consulting for Nordic companies: why a mid-market leadership team can move fast once the CEO decides. (accessed October 8, 2026)
Next scheduled review: February 4, 2027