Listen closely.
Change the question.

In an Oppia lesson-player review, I flagged conflicting exit-dialog copy and button emphasis, then asked engineering to clarify when progress actually saves. I separated those open questions from recommendations about narrow layouts. That gave the team a focused set of decisions to resolve before implementation.

When the evidence challenges the premise. Oak’s privacy-community material raised concerns about control and company behavior, beyond understanding policy language. I used those themes to question whether deeper work with the same audience was the right next investment. The partial record supports that question, not a market-fit verdict.

I separate the observation, my interpretation and the product decision it might inform. In the audience survey, reported barriers led to proposals for clearer steps, templates and reminders. Those proposals are visible alongside the sample limits; the survey does not prove that the proposed product changed behavior.