--- title: "Using product data to retire a recurring idea: Issue 1 | Skene" description: "An anonymous founder trusts product data to confirm or kill a hypothesis, rarely to start one. What they miss is the why behind a drop-off, which no event records." url: https://www.skene.ai/resources/the-question/anonymous-founder doc_version: "c00117dbc7a2" last_updated: 2026-10-30 --- Resources / The Question / Issue 1 # Interview 01: the founder who uses product data to retire a recurring idea Anonymous founder. Founder and CEO of a B2B2C company of 50 to 100 people with several product lines, who chose to stay anonymous. Issue 1 of The Question, published by A Founder's Friend on 24 July 2026. By [Nicolas Dolenc](https://afoundersfriend.substack.com), author of A Founder's Friend. Skene adapted this page from his published issue. An idea keeps coming back. Leadership doubts it and no product data supports it, but it returns anyway. So the team runs an A/B test, and the point of the test is to cross the idea off for good. That habit sums up how this founder uses product data. The numbers are a second opinion on a hypothesis the team already holds. They are rarely where the hypothesis comes from. **The short answer:** Product data confirms or kills a hypothesis that people formed first, and the expensive gap is the number nobody collects. ## What this founder trusts, and for what | Evidence | Trusted for | Not trusted for | | --- | --- | --- | | Domain knowledge, interviews, watching users | Forming the hypothesis in the first place | Settling a recurring argument | | Cohort funnels and A/B tests | Confirming or retiring a hypothesis | Explaining why a user left | | Session recordings | Seeing the why, one user at a time | Anything systematic, because watching does not scale | | Backend events joined into BI tables | The hard numbers, once someone builds them | Speed: each new one waits in a queue | ## People first, data on both sides The founder describes the order plainly. Understanding comes from domain knowledge, customer interviews and watching people use the product. Data comes in afterwards, on both sides of the question, to see whether the observation holds. By the time something shows up in a chart, the team has usually discussed it already. If you are deep enough in the product, they say, the answer is partly obvious already, and the data confirms or rules it out without much effort. The test that retires a recurring idea is the clearest case: its job is to end an argument, not to discover anything. ## The one time the numbers moved the roadmap There was one genuinely data-led change. A secondary platform the team had always meant to improve, and kept parking, turned out to account for a much larger share of usage than anyone assumed, and its retention was many times what the team had believed. It moved from roughly fifth on the priority list to second. Note what the data did there. It did not invent a new idea. It re-ranked a known one, which is a narrower and more believable job than most dashboards are sold for. ## The number that does not exist Cohort funnels show when a user activated or dropped off and what they clicked before it. They do not show that the user searched for the same thing eight times, never got the right result, hovered for five minutes and left. For now, those moments surface only when someone sits through session recordings by hand. The founder calls the gap emotional. The moments where users give up are often not a data point at all. That is the expensive gap in this interview: not an inaccurate number, but one nobody collects. ## Where trustworthy data stalls The hardest tracking fires deep in the backend. Dozens of events have to be joined into large tables before anyone can build a dashboard on them, and the team has one data engineer and one automation person to do it. The founder is clear that the team can do it. It is a big project, so it waits in a queue. ## What to do with this Our reading, not the founder's: a test is only as good as the events it reads. Before a test settles an argument, check that those events still fire the way the dashboard assumes, because a test that reads a broken event retires the wrong idea with full confidence. For seven checks drawn from across the series, see the [questions to ask before trusting a number](https://www.skene.ai/resources/the-question#before-you-trust-a-number). ## Frameworks this interview tests - [Cohort retention analysis](https://www.skene.ai/resources/frameworks/cohort-retention-analysis): Cohort retention analysis tracks each signup group's curve, so you can see whether retention is improving. It assumes 'active' means the same thing every month. - [Hook Model](https://www.skene.ai/resources/frameworks/hook-model): Nir Eyal's Hook Model builds habits through trigger, action, variable reward and investment. Most steps leave events; the reward step barely does, so measure with care. - [AARRR (pirate metrics)](https://www.skene.ai/resources/frameworks/aarrr-pirate-metrics): Dave McClure's AARRR pirate metrics count acquisition, activation, retention, referral and revenue. Each stage rests on one or two events; activation breaks most often. ## Read next in The Question - [Maria Petrova: data tells you what happened, never why](https://www.skene.ai/resources/the-question/maria-petrova) - [Shahin Katebi: data finds the stuck user, a person asks why](https://www.skene.ai/resources/the-question/shahin-katebi) - [Arash Farhani: the first data he trusts is a testimonial](https://www.skene.ai/resources/the-question/arash-farhani) - [What all fourteen interviews add up to](https://www.skene.ai/resources/the-question) - Glossary: [Funnel](https://www.skene.ai/resources/glossary/funnel) - Glossary: [Cohort](https://www.skene.ai/resources/glossary/cohort) ## Source and disclosure [Read Issue 1 on A Founder's Friend](https://afoundersfriend.substack.com/p/aff-the-question-issue-1-making-decision). The full interview is for subscribers. This page paraphrases it and quotes nothing that sits behind the paywall. Skene sponsored this research and commissioned it on a single question. A Founder's Friend ran the interviews, wrote them up and published them, and the series states that Skene paid for the research and did not direct the questions, the analysis or the conclusions. Each of the thirteen interviews was reviewed and signed off by the person in it, named or anonymous by their choice. Appearing in the series is not an endorsement of Skene, and nobody here is presented as a Skene customer. ## Sitemap - [Markdown sitemap](https://www.skene.ai/sitemap.md): every page on this site, with its markdown counterpart. - [Index for language models](https://www.skene.ai/llms.txt): the same set as a link-dense index. - [Full corpus](https://www.skene.ai/llms-full.txt): reference content concatenated as plain text.