Official skill · Strategy

Strategy choice cascade

v1.0.0 · Updated 4 November 2026

Writes a strategy on one page as five linked choices, builds where to play and how to win as pairs, and tests the pair with a what-would-have-to-be-true list. Use it when A team needs a strategy on one page.

Works in Claude, ChatGPT, and coding agents. On Team and Business, the official skill in your pod updates itself when we improve it.

When to use it

• A team needs a strategy on one page
• Leaders are choosing between strategic options and keep arguing
• The strategy is a vision plus a list of initiatives

What you give it

• The scope (company, unit, product, function)
• Facts on customers, channels, competitors, and your strengths
• The options on the table
• Who decides and by when

What you get

The five choices on one page, a what-would-have-to-be-true table with tests and owners, and a choice or a dated test plan.

How it works

1. Set the scope and the aspiration Write the winning aspiration in one or two sentences: what winning means for this scope and for whom. Don't overinvest in it up front: treat it as a direction, and put the effort into the next two choices.
2. Generate where-to-play and how-to-win pairs Write two to four options, each as a pair: where to play (which customers, products, channels, geographies, and stages of the value chain) with the how to win that fits it. How to win must be a theory of advantage for those customers, through more value or lower cost than rivals, not a list of good things to do. Reject any where-to-play that has no how-to-win.
3. Write what would have to be true for each pair For each option, list the conditions that would make it a great choice, grouped by industry (segments, structure), customers (channels and end customers would value and pay for it), the organization's relative position (capabilities and cost it can build), and competitors (how rivals would react). Do not argue yet whether each condition is true.
4. Find the conditions to test first For each pair, mark the conditions the team doubts most. These decide the choice. For each one, agree in advance what test or data would settle it, and how to get that cheaply. Martin's point is to agree on the test before anyone collects data.
5. Choose the pair, or say what must be tested first If the evidence is in, choose the pair and state what it rules out. If not, recommend the tests and a date to decide; do not pick by preference.
6. Name the must-have capabilities For the chosen pair, list the few activities and skills that make how to win real. For each, apply Martin's reality check: can rivals not, or will they not, match it? A capability every competitor has is necessary but not a source of winning.
7. Name the enabling management systems For each capability, the systems that build and sustain it: measures, budgets, hiring and training, decision rights, regular reviews.
8. Check the cascade both ways Read down (does each choice follow from the one above?) and up (can the systems build the capabilities, and do the capabilities deliver how to win?). Revise the upper choices if the lower ones fail.
9. End with decisions and next actions The choice or the tests, what stops because of it, owners, dates, and the facts still to check.

Share it with your team

Add it to an AI Skill Pod and everyone on the team, contractors included, uses the same version from their own AI. Read the best ways to share AI skills with a team

Add to my team's pod

Changelog

• v1.0.0 · 4 November 2026 · First release.