Concept
Weekly learning loop
The weekly learning loop is a thirty-minute weekly meeting in which a founding team reviews last week's experiment and customer conversations, names one learning that changed something, and rewrites its one-page plan for the coming week.
A loop that is not scheduled is not a loop; it is an intention. The weekly cadence is chosen because it is the shortest period in which a real experiment can be run and a handful of customer conversations can be held, and the longest a founder can go without checking whether the riskiest assumption has changed. The meeting has three parts and takes thirty minutes with co-founders only. Results in: what the experiment showed, what the conversations produced in facts and commitments, where the one metric stands by cohort. The learning: one sentence, beginning with what we learned from a customer, that changed what we will do; if there is no such sentence, the week was activity rather than progress, and that is worth knowing. Plan out: the one-page plan for next week, rewritten rather than edited. In BookYol's Think Like a Founder Yol that page is called the founder operating system, and it has seven lines: the riskiest assumption, the experiment, the customer conversations, the one metric, the one channel, the date of the next pivot-or-persevere meeting, and the truth not yet told. Over months the loop does what a plan cannot: it compounds, because every week starts from what the last one proved.
A real-life example
A three-person team in Ankara held its loop on Monday mornings at a café before the office opened. In the eleventh week the learning sentence was: two customers said they would pay double for a version that worked offline. That single sentence changed the following month's roadmap, the pricing and the channel. The team later said the café meetings were the only management the company ever had, and the only management it needed.
How to use it
- 1Fix the meeting: same day, same thirty minutes, co-founders only, every week without exception.
- 2Force the learning sentence: what we learned from a customer this week that changed what we will do. An empty sentence is the most useful result the meeting can produce.
- 3Rewrite the one-page plan (assumption, experiment, conversations, metric, channel, pivot date, truth not yet told) rather than editing it, so that each week starts from what the last one proved.