What we learned from 24 interviews and what it means for the product.
Three months after launch, retention numbers told us something the metrics couldn't.
Primary objective · Round 240 synthesis
The Pattern
Users who complete three core actions in their first two days have a 4× higher 90-day retention rate. Most never get there.
Session recording review · April
of users churned within 14 days — up from 54% in cohort 2
Product Analytics · March
Average time before abandonment on the setup flow
Session recordings · n=240
Higher 90-day retention for users who complete onboarding fully
Cohort analysis
What to fix
We recommend addressing these sequentially — later ones depend on the first landing.
Users choose their own path through setup. Most choose wrong.
User type detected at signup. Path adjusts. First value in under 90 seconds.
"I kept opening the app and then closing it again. I didn't know what I was supposed to do."
Activation is the moment a user experiences the core value of a product for the first time. When that moment is delayed or never arrives, the user's mental model of the product never fully forms. They carry a vague, unresolved impression into every subsequent session.
Each session that ends without activation reinforces the exit pattern. The user doesn't consciously decide to leave — they simply stop opening the app because it hasn't yet earned a place in their routine. The gap between download and habit is where most products lose users permanently.
Retention data confirms this: users who hit activation in session one have a 3× higher probability of returning in week two. The window is narrow. Products that front-load value creation outperform those that distribute it across multiple sessions.
Collaboration products face a compounding problem: the value of the product increases with each additional teammate, but users must cross the value threshold alone before they think to invite anyone. Most never reach the threshold.
Our data shows that the median user does not discover the invitation flow until session four — by which point 60% have already churned. The product's most powerful retention mechanism is invisible during the critical first-session window where the keep-or-leave decision is made.
The solution is not to surface the invite prompt earlier in a disruptive way, but to design the single-player experience as an explicit bridge to the collaborative one. Every solo action should feel like preparation for something that scales.
Source: Product Analytics · Cohort analysis · Proposed target based on redesigned onboarding flow
Total participants: 240 · April 2026
Source: Recruitment screener · April 2026
Screened 240+ applicants, selected 240 participants across user segments.
240 moderated sessions. Think-aloud protocol. All sessions recorded and transcribed.
Affinity mapping across 240+ observations. Pattern clustering by behaviour type.
Findings stress-tested against [analytics tool] data and 240 support ticket samples.
Understand users in their own context. Suspend assumptions. Observe before interpreting.
Reframe the problem as a point of view. One sentence. Testable. Grounded in observation.
Put prototypes in front of real users. Capture what they do, not what they say.
Build to think, not to ship. The lowest fidelity that answers the question.
Research Framework
From raw observations to strategic insight
Research Team
[[email protected]] · Research room · Full report at research.grove.example