01
Users could not see what the tool gave them.
The home screen was a planning grid: a chore, not a reason to come back.
“I come on Tuesdays because I assume my team is there. But I don’t really know.”User, client company
CASEat.Café (Y Combinator) · B2B SaaS · Hybrid work
Cutting new-user churn from 55% to 32%
Redesigned the first session around people and events instead of a schedule.
- Role
- Research-led redesign of onboarding and home
- Scope
- Onboarding and home experience
- Duration
- Dec 2021 to Dec 2022
- What I did
- Interviews after drop-off, a continuous 5-question survey, card sorting and prototype tests on social features, user tests of the new “Lounge” before rollout.
- new-user churn
- 55% → 32%
- survey answers / week
- ~50
After Y Combinator, at.Café onboarded ~300 new users a week, but 55% never came back after planning their first week. The home screen showed a weekly planning grid with no immediate value.
How I went to the field.
- 01
Interviews after drop-off
Moderated interviews with users who stopped coming back, to understand why the first week did not stick.
- 02
Continuous survey
A 5-question survey after the first login (~50 answers a week) that also recruited the next interviewees.
What the field told us.
02
It felt like an imposed constraint.
Rolled out by leadership, configured by managers and HR: one more task to fill in.
03
The real need: finding colleagues.
In hybrid work, people from the same company did not know each other despite shared interests. Connections and events made them come in.
“We met by chance at the coffee machine. That’s where the real conversations happen.”User, client company
From evidence to screens.
- 03
Card sorting
Card sorting on the social features to find what users expected to see first.
- 04
Prototype tests
Prototypes of the social features tested with users, iterating on what they understood and used.
- 05
“The Lounge”
The home rebuilt around colleagues, company events and interest clubs; planning a month in a few clicks.
- 06
Tests before rollout
User tests of the new version before it shipped to every client company.
What I chose, and what it cost.
01
Lead with people, not the planner
- Options
- Keep the planning grid and add reminders · or rebuild the home around connections and events
- Trade-off
- Planning stays one tap away but loses the hero spot
- Choice
- The home became “The Lounge”: colleagues, events and interest clubs first.
02
Plan a month in a few clicks
- Options
- Day-by-day entry · or recurring patterns
- Trade-off
- Less granular, far less effort
- Choice
- Plan a whole month in a few clicks so the chore no longer blocks the value.
Outcome
The new version was tested with users before rollout. After the redesign, new-user churn dropped from 55% to 32%.