Manifesto
m.
How we make product decisions.
Six principles. Every decision passes through them.
When trade-offs land mid-sprint, and they always do, these are the rules we hold the line on. Founder included.
- 01
Ship to learn, not to launch.
Every week ends with something testable, even if it's ugly. Demos beat decks.
- 01.1 A launch is a bet. A test is a question. We ask first.
- 01.2 Ugly and real teaches you more than polished and hypothetical.
- 01.3 If a week ends with nothing to put in front of a user, the week was theory.
- 02
Scope is the product.
What we cut matters more than what we keep. “Yes, but later” is a feature.
- 02.1 Every feature you add is one you have to defend, maintain, and explain.
- 02.2 The fastest way to a clear answer is a smaller question.
- 02.3 Cutting is a decision, not a delay. We cut on purpose.
- 03
AI where it earns its place.
Agentic patterns only when they remove real work. No AI theatre.
- 03.1 We use agents to delete tasks, not to decorate the product.
- 03.2 If a human does it better, a human does it.
- 03.3 The point is the outcome, not the demo.
- 04
One owner per decision.
No committee design. The founder owns the vision, we own delivery.
- 04.1 Shared ownership is no ownership. Every call has a name attached.
- 04.2 Disagree, decide, commit. Then move.
- 04.3 Speed comes from clear authority, not more meetings.
- 05
Real data from day one.
No mocked screens past week two. Real inputs, real outputs, real edge cases.
- 05.1 Mock data hides the problems that kill products.
- 05.2 Edge cases are where the truth lives.
- 05.3 If it only works on the happy path, it doesn't work.
- 06
Boring tech, sharp judgment.
Default to proven stacks. Save the novelty for where the product actually differentiates.
- 06.1 Novelty is a budget. Spend it where customers notice.
- 06.2 Proven tools fail in known ways, and known is cheaper.
- 06.3 The interesting part should be the product, not the plumbing.
Therefore
t.
We end every week with something real. We cut before we add. We use AI where it removes work and nowhere else. We give every decision an owner. We build on data, not mockups. And we keep the novelty for the product, not the stack.
Six principles. Every decision passes through them. Founder included.