AI response
Recommended AI: ChatGPT, Claude
🗂 These notes can't stay as-is — let's pull out just the decisions and the to-dos.
Meeting summary · A session to set the direction for the new app's onboarding redesign; landed on cutting it from 3 steps to 2.
By topic
1. Cutting onboarding steps
- Discussion: Data shared that drop-off is highest on the 2nd screen. One side said there are too many input fields; the other said taste data is still needed for recommendations.
- Decision: Cut 3 steps to 2, and move taste input into an optional, skippable step.
- Left open: Where to re-collect taste input later is undecided.
2. Recommendation-quality concern
- Discussion: Making taste optional may lower early recommendation quality. A compromise: start with a popularity-based default.
- Decision: When taste is skipped, default to a "popular + new" mix.
3. Schedule
- Discussion: When the design draft can be ready.
- Decision: First draft by next week, dev kickoff the week after.
Action items
| Item | Owner | Due | Priority |
|---|---|---|---|
| Wireframe the 2-step onboarding | Planning lead | Next Mon | High |
| Define the skip-taste-input flow | Planning lead | Next Wed | High |
| Verify "popular + new" default logic | Dev lead | TBD | Medium |
To decide next
- At which point in the app do we re-prompt the skipped taste input?
- What ratio should the "popular + new" mix be?
One-line take — Direction is set, but "when to collect taste data" got pushed wholesale to next time. Density ★★★☆☆ (3 decisions / 2 open)





