The Foundation Matrix
Which cell is the real problem? Every problem sits in one cell of a grid: Strategy, Creativity or Technology, crossed with Judgement, Craft or Distribution. Name the cell before prescribing anything, because the fix for one cell does nothing for another.
Three areas, three qualities, nine cells.
The rows are what gets done. The columns are how well it gets done. Each cell is a different kind of problem, with a different fix.
| Judgement | Craft | Distribution | |
|---|---|---|---|
| Strategy | Thesis validation, ICP definition, kill decisions, scope reduction | Rigorous models, scenarios, defensible positioning | Go-to-market, pricing logic, growth loops |
| Creativity | Taste: choosing one idea out of the hundred a machine can generate | Identity, design systems, motion, interface quality | Narrative, demo assets that travel, launch creative |
| Technology | Architecture calls, build versus buy, what not to build | Production-grade engineering, documentation, observability | Ship velocity, deployment, adoption instrumentation |
What the rows and columns mean
List the symptoms first. Then, for each one, ask which area it lives in and which quality failed. Stop when you can name one cell.
Strategy
Venture thesis, positioning, ICP, product strategy, revenue and pricing logic, go-to-market.
Creativity
Brand identity, visual and design systems, motion, product UI/UX, demo assets, narrative.
Technology
Product engineering, AI and agentic features inside products, production builds, documentation, observability.
Judgement
Knowing what's worth building, and being willing to say what isn't.
The thing works and people can find it, but the wrong people want it, or nobody decided what it was for.
Craft
The execution quality that makes a decision survive contact with users.
The decision was right, but what shipped breaks, confuses or looks unfinished.
Distribution
Getting it shipped and adopted.
The right thing exists and works, and the people it's for haven't heard of it or haven't started using it.
The build was fine.
Not client work. The pattern is common enough to show without one.
Illustrative example
A team has a working product and flat numbers, and asks for a rebuild. The product does what it says. It passes its tests, it's fast enough, and the people using it can finish what they came to do.
So the problem isn't Technology × Craft, and a rebuild won't touch it. The people signing up aren't the ones with the problem badly enough to pay. That's Strategy × Judgement: the right thing, built for the wrong buyer.
Naming the cell changes the next step. Instead of a rebuild, the work is to find the buyer who has the problem, and to decide what changes for them.
Where it goes wrong
- Misread
Naming several cells
Why If the diagnosis spans three cells, it isn't a diagnosis yet. Keep asking until one cell is left.
- Misread
Finding the cell you're equipped to fix
Why Engineers tend to find Technology problems; brand teams find Creativity ones. Name the cell before you look at who's available.
- Misread
Reading it as a capability list
Why The point isn't that one team covers all nine cells. It's that products fail at the seams between them, so someone has to hold the line across all three areas.
In the Journal: Teardown: our old website · An AI venture beside your family business · White-label AI for agencies
Published 24 September 2026. Graylemon original, from our brand foundation.