Weeks 1–2 · Narrow
One ICP, one channel, one hypothesis
No experiments yet. Write a single page that names the customer paragraph, the channel you can actually run, and the hypothesis you believe enough to be uncomfortable about.
- ICP paragraph (3 real names)
- Channel chosen by access
- Hypothesis: If X in Y, then Z at rate A
- Friday 1 review
- Friday 2 review
Weeks 3–4 · Run five
One obvious, three sensible, one weird
Five small bets in parallel, all inside your narrow. Every experiment uses the same template. Two-week minimum window before reading any signal. Spreadsheet row per experiment.
- Experiment 1 (obvious)
- Experiment 2 (variation)
- Experiment 3 (variation)
- Experiment 4 (variation)
- Experiment 5 (weird)
Weeks 5–6 · Double down or kill
Concentration, not coverage
One will be working, one clearly not, two ambiguous, one surprising. Pour everything into the winner. Kill the loser the day after the review. Investigate the surprise.
- Winner: double-down plan
- Loser: shutdown date
- Surprise: 1-week investigation
- Ambiguous A: keep on autopilot
- Ambiguous B: keep on autopilot
Weeks 7–8 · Turn into a loop
From event to system
Describe the winning motion as a sequence whose output becomes its own input. Find what you can take yourself out of. Decide the order of magnitude for inputs and outputs. Put the rhythm on the calendar.
- Loop written as steps
- What can be templated
- Inputs per cycle (order of magnitude)
- Outputs per cycle
- Weekly rhythm scheduled