Fieldnote made onboarding a weekly experiment
A note taking app turned its activation funnel into a standing test queue and tripled week one retention gains.

A small team with a leaky funnel
Fieldnote has 120 employees and a free tier used by hundreds of thousands of people. Most of them sign up, open a blank page and leave. The team knew activation was the lever, and had no experiment process to pull it.
A standing queue
Instead of planning tests per quarter, the team keeps a queue of onboarding ideas and runs one at a time. Each idea is a flag with a hypothesis, a primary metric and a guardrail written in the flag description before anything ships.
- Every test uses the same primary metric, a second note written within 24 hours.
- Sample size is estimated up front so nobody peeks early.
- A test that does not reach significance by its end date is closed, not extended.
What won
Starting new users from a template rather than a blank page won clearly. A tour of the interface lost. A prompt asking what they wanted to write about, before showing any template, beat both.
A blank page is the worst thing you can show a new user. We needed a test to believe our own instinct.
Founding designer, Fieldnote
Results
Week one activation rose 24% over six months of stacked wins. Each individual gain was small, between 2% and 6%, which is why a team that waited for big ideas would have found nothing.
Two engineers now run the whole program, using AI readouts to write the weekly summary the whole company reads.
More customer stories

Northwind cut checkout drop-off by 18%
A retailer with 2,400 employees moved every checkout change behind a flag and a test, and found the fix was a smaller form, not a redesign.

Halcyon ships weekly instead of monthly
A health scheduling platform replaced a manual release process with gradual rollouts and automatic rollback.

Brightpath replaced three tools with one
An education platform consolidated flags, experiments and session replay, and the arguments about whose numbers were right stopped.