From an AI Mandate to an AI Strategy

Most enterprises now have an AI mandate. Few have an AI strategy. The two get confused, and the confusion is expensive.

A mandate is a target someone hands down. Adopt this. Hit that number. Be ahead of the competition by the end of the year. A strategy is something else entirely: a model of where AI actually changes the work your organisation does, what it makes cheaper, what it makes more valuable, and what has to be true before any of it pays off.

This whitepaper is about the distance between the two, and how to cross it. It is built on two years of doing AI-enabled product work in production, not on a forecast. What follows is the short version.

A mandate is not a strategy

You can recognise a mandate without a strategy by its symptoms. There is a spend target: this much on tokens or tools per month. There is an adoption metric: this share of the team using AI by the next quarter. There are rollouts, licences, training sessions. What there is not is a shared picture of what the AI is for.

The spend target is the clearest tell. We do not measure marketing maturity by ad spend; we measure what the spend returns. Token spend with no output measure is the same category error. It gets tracked because it is the one number that is unambiguously countable, and it gets tracked precisely when nobody can yet say what the AI is doing inside the business. The metric is standing in for a missing model.

A strategy replaces the target with a map: the specific places in your value chain where AI compounds value, and the places it does not belong. That map is the thing worth having. Everything else is procurement.

What changes for people

This is the part most coverage skips, and it is where the real consequences live. When AI absorbs a large share of production work, cycle times drop. But the value of senior judgement rises, because the scarce skill is no longer producing the work; it is knowing whether the work is any good and what to do next. Experienced designers, engineers and product leads become more important, not less.

The uncomfortable corollary is that the work which used to train juniors is the work that compresses first. The rungs people climbed to build judgement are the rungs AI removes. Organisations that do not think about this end up with a gap in their own bench. A serious strategy has a view on how judgement gets built when the apprenticeship work is gone.

Governance becomes a first-class problem

The moment agents touch real systems, accountability stops being a footnote and becomes a design decision. Who is responsible when an agent makes a change that breaks something? What can an agent do on its own, and what must it escalate? How is data that should never reach a model provider kept away from one? What is the rollback path, and who owns it? These are not questions to answer after deployment. They are the conditions that make deployment safe. A strategy treats governance as part of the build, not as a policy written afterwards to cover it.

The operating model, in brief

The shape of the work changes too. Context moves to the front: getting the problem, the constraints and the definition of good written down clearly becomes the phase that determines everything downstream. Delivery, which used to be the slow expensive part, speeds up. We treat this at length in the full whitepaper, including the frameworks we use and the failure modes we have hit.

The gap between a mandate and a strategy is not technical. It is the work of deciding what AI is for in your specific business, and what has to be true to get there.

The full whitepaper, download here, goes through each of these points with frameworks and examples. If you want to know where your organisation actually stands before you begin, book a readiness conversation.