How can AI contribute to strategy development?

AI can provide another perspective on a decision, organise the evidence you supply and suggest alternatives. Its usefulness depends on the quality of the brief and your ability to assess the response. A persuasive recommendation is a starting point for investigation, not proof that the strategy will work.

Begin with a specific choice. “Should we enter this customer segment with our current team?” gives the discussion a clearer purpose than “Write our growth strategy.” Define the outcome, the time available and the constraints that cannot be ignored.

Give the system a balanced decision brief

Include the business model, customers, current performance, capacity, commercial constraints and options already considered. Separate established facts from estimates and unknowns. Use information that your organisation has approved for the chosen system, and share only what the task needs.

Avoid presenting a preferred option as the answer before asking for feedback. Describe each option against the same criteria. Ask the system to identify missing information before recommending a course of action, and require it to distinguish supplied evidence from its own assumptions.

Challenge the plan before polishing it

Ask for the strongest case against the preferred option. Which assumption carries the most weight? What could make the plan fail even if the team executes well? What evidence would cause the recommendation to change?

You can also ask for a credible alternative and the conditions under which it would be better. The aim is to reveal trade-offs. Producing ten variations of the same idea does little if every variation depends on the same untested customer demand.

A practical example: entering a new market

Suppose a business is considering a new service for an adjacent customer group. Give AI the proposed offer, delivery requirements, available capacity and evidence from customer conversations. Ask it to compare a full launch, a limited pilot and postponing the investment.

It might flag an assumption that existing customers will buy the new service at a particular price. The next step is to test that assumption with customers and delivery colleagues. Asking the model to debate the price again does not replace that work. Treat this as an illustrative decision process, not a forecast of results.

Turn the discussion into an evidence plan

Create a short record of each consequential assumption, the evidence needed, an owner and a date for checking it. Review source documents and calculations directly. If a recommendation depends on market information, establish its source and freshness before using it.

Agree what would count as a reason to proceed, adapt or stop. Record the decision and the uncertainty that remains. This makes later review more useful than judging the strategy only by how confident the original presentation sounded.

How CSM would approach the work

We would start with a real decision and the people accountable for it. Together, we would define a useful brief, the approved evidence and the points where human judgement matters. AI can then help prepare comparisons and challenges within that process.

Assess the contribution by whether it exposes a material gap, improves the options or reduces preparation without increasing review effort disproportionately. Keep the decision owner involved where priorities or commitments change. A better strategy conversation should lead to a clearer next action and a more testable plan.