Strategy.

The framework

1. 01

   Empty the bag before you rank anything.

   Rumelt's process starts wide: write down every gnarly challenge, worry and ambition facing the business — the churn you don't understand, the founder-dependence, the channel that's saturating, the competitor moving downmarket, the margin that thins each year. Do it without filtering, because the real crux is often the item you'd rather not write, and a list built to look composed will omit it. Take twenty minutes this morning and produce the honest version; a strategy chosen from a sanitised list is sanitised too.

2. 02

   Score by critical times addressable — the crux lives at the overlap.

   Two questions sort the bag. First: if this remains unsolved, does the next stage of the business actually happen? Second: with the resources, skills and position you genuinely have, can you crack it — not wish it away, crack it? A challenge that is critical but unaddressable is a condition, and you design around conditions rather than fling budget at them; a challenge that is addressable but minor is a distraction wearing the costume of progress, which is where most corporate energy quietly goes. The crux is the highest-stakes challenge you can actually break, and there is usually exactly one.

3. 03

   Diagnose until the problem cracks — reframing is the tool.

   A crux rarely yields to effort applied in its original framing; it yields to being restated. Rumelt's insight moments come from reframing — 'we have a sales problem' becomes 'we sell to whoever answers, so our pipeline is full of buyers we should never have pursued', and suddenly the challenge has a crackable face. Work the restatements: describe the crux as a customer would, as a competitor would, as an engineer would; ask what known, solved problem it resembles. You have diagnosed enough when the statement of the problem itself suggests the first move — until then, keep diagnosing rather than planning.

4. 04

   Concentrate — strategy is what you fund and what you refuse.

   A crux identified and then given a line item alongside nine other priorities has not been chosen; it has been filed. Coherent action means the calendar, the budget and the best people visibly shift toward the crux, and things that were defensible last quarter get stopped to pay for it — Rumelt's test for bad strategy is precisely that nothing real moves. So finish the exercise with subtraction: name what you will stop, starve or decline in order to concentrate, and put one owner on the crux with the authority to pull resource. If the strategy changed nothing about next week, it is a document, not a strategy.