Two experiments. Four weeks.
One decision a week.
Four weeks, two real growth experiments, a decision every Friday. For developer-tool and AI-product founders who don't have a growth lead yet. Two experiments with your real users and channels, decided every Friday, handed over on day 28.
By Daria Dovzhikova · Updated October 2026
Growth Pilot
Week one is a baseline of your current funnel and two experiment cards: one product, one audience, one message, one channel, a metric and a stop date. Weeks one to four, both experiments run with your real users and channels, and a written Friday review decides continue, change or stop. Day 28 is a readout with results and hours spent, and what to keep running. You keep everything.
Continuation
Runs what worked in the pilot, adds one new experiment a month, and ends each month with a readout. Starts only after a pilot, because the first four weeks are where the cards get written and the first results come in.
Built for
A team, a working product, no growth lead yet.
Seed to Series A developer-tool and AI-product companies. The buyer is the founder or CEO, with no senior growth person on the team and under $10,000 a month for marketing.
Solo founder? The pilot still runs, prepaid and fixed-end. Two cards, four weeks, a readout, and nothing after unless something worked.
The trigger is usually a product people use and a channel nobody owns: the launch went out, the stars came in, and the question of what to try next sits with the founder.
Not built for
Where this does not fit.
- //Pre-product teams. An experiment card needs real users and a real channel to run against.
- //Companies with a growth team already. They should write and run the cards themselves; the pilot is for the team that has nobody to do it.
- //Anyone looking for a full-time hire. The pilot ends on day 28 by design, and Continuation is month to month.
- //Teams that want a strategy document instead of tests. The first week covers the diagnosis, and the experiments run from week one to week four.
Deciding whether to hire a marketer at all? The Marketer Question is the honest read, and the free tools cover what a first marketer needs to know.
How the measurement works
A number in, a number out.
Every experiment card names its metric before it runs, and the Friday review reads that metric. Where a card is about AI search visibility, the method is public: the scorer is open source and runs free below.
Run the free visibility checkTwo experiments, four weeks.
Thirty minutes to see whether a pilot fits. If it does not, the call says so.