The Shape of a Strategy
Profit is a ring. One circle is what you earn. The other is what it costs. Put the second inside the first and the ring between them is the business. Run that along a line -- time -- and every chart you have ever produced is a plane through the solid it sweeps.
A market is a segment. A cell is a coordinate inside it. The coarsest cell is the market alone -- precision 0, the floor every unqualified reference falls back to. Narrow it by who buys, then by who decides, and precision is simply how many narrowings you have made: also the depth in the tree, also how well you can explain where the money went. Attach to the most precise cell that applies, and only that one -- never to both a cell and its parent, since the parent receives it by rollup. The cells of a market sum to the market, nothing lost and nothing counted twice, which makes the cell tree an accounting partition rather than a tagging scheme: it can be filtered, sliced and rolled up without ever needing reconciliation. A cell holds no thesis and no commitment -- it records where a thing sits, not what you believe about it. Every reference to a market anywhere in the model is an array of cell ids, never a bare market, because a bare market would say which segment and nothing about how well you knew it.
A budget divides. A capability does not. Spend it here and it is not available there, so spend looks like a pie -- but build something once and it can serve four cells at no extra cost, which means reach exceeds spend. The ratio between total reach and total budget has a common name: it is what people mean when they say platform, and here it is measurable. Sum a company's reach as vectors across its cells and most of them cancel -- cells sit at opposing angles, so the resultant is routinely much shorter than the deployment behind it. What remains is where the company is genuinely pointed. Allocation sets direction; reuse sets reach -- two dials, moving independently. A company can be sharply focused while building nothing twice, or heavily reused and pointed nowhere at all, and those are different problems with different fixes. A single reading says where you are; the series says how fast you can move, and that is the number worth having when the market turns.
A build serving three cells lengthens three rays while costing one budget line -- the whole argument for building things once, in one line no pie chart has ever been able to show.
Market focus is not declared in a strategy document and then defended in meetings -- it is read off the order of the work. Reorder the queue and the arrow swings, by an amount you can see before committing to it.
The same records answer five questions most companies pay to have guessed at: where you actually make money, what you've built once that serves several, what the money says you're doing versus what you declared, what's reachable from what you already hold, and how fast you can change direction. Four of those come from one close. The fifth needs a series. Move a build up the queue and a ray gets longer -- which is where the next piece begins: the order, and what it costs to change one.