Find the moment people get your app.Get every new user there.
Omni reads what new users do, finds the early action that separates people who stay from people who leave, and shows where onboarding loses them before they get there. Then it writes the fix up as a test sized to your traffic, and reads your next release against the last one.
Activation loopRound 1Omni reads and measures; you ship the change.
Scribo · new installs0 eventsOnboarding, step by step
Numbers in the loop and the chart are Scribo's real data, October 17–21, 2026.
From first open
to the aha.
Omni does the reading, the math and the follow-up. You decide what ships.
Reads what your new users do
Connect PostHog or add Omni's SDK. Omni reads your events, works out what your app is about, and sets up the numbers that matter for you: activation, your onboarding funnel, and day-1 and day-7 retention.
Set up for you
Finds the early actions that predict staying
For every action new users take in their first three days, Omni compares the people who did it with the people who never did: how many are still around in week two. It ranks them by the gap, shows how many people are behind each, and flags the thin ones.
Still here in week 2 · did it vs never
- Saved a note48% vs 12%
- Finished a lesson41% vs 18%
- Turned on reminders30% vs 22%
- Shared a card12 people52% vs 24%
Correlation, not cause: the top one becomes a test.
Shows where new users drop before getting there
Your onboarding as a funnel, step by step, so you see exactly where people leave before they reach the moment that keeps them.
Onboarding · example
Writes the fix as a test sized to your traffic
One change, a main metric and guardrails, and how long it needs at your number of new users. If a test would take months, Omni says so and suggests a bigger change instead.
Experiment · Activation
- Change
- Ask for permissions after the first save
- Main metric
- 3+ saves in the first 3 days
- Guardrail
- Trial starts
Reads your release like an experiment
You ship the change. Omni compares people who started on the new version with people who started on the one before, with a 95% interval, and calls it: up, down, no clear change, or not enough data yet.
Activation · new users
Up · +13 pts, 95% interval +5 to +21
Cards above are examples. The Scribo numbers below are real.
Scribo asked for money before the aha.
Scribo's aha is writing an unknown word with the pen and seeing its meaning. Five days of its events, October 17–21, showed new installs meeting the paywall at step 8 of 15, before they had written a single word: 110 reached the paywall, 68 got past it, and 62 wrote their first word.
157
new installs
39%
wrote a word during onboarding
−38%
at the paywall
An early signal, not a verdict
New installs who saved 10 or more words on day one came back 59% of the time (13 of 22). Everyone else: 22%. Omni's week-two check needs two weeks of data, so it's still running.
What's next
Scribo 2.5 lets people write their first word before they see the price. Omni will read 2.5 against 2.4 on new users, with trial starts as the guardrail.
Scribo on iOS · PostHog events, October 17–21, 2026 · new installs on 2.3 and 2.4
Onboarding · new installs
−38% at the paywall
Came back after day one
13 of 22 new installs
25 of 116 new installs
Connect what your app already sends.
Product events tell Omni what people do. Your store tells it which version they're on.
To see what people do
PostHog, or Omni's own SDK for iOS and web.
To read each release
So every version is a before and after.
To keep paying in view
Trials and payments as guardrails on every change.
Questions about activation with Omni.
It looks at everything new users do in their first three days and compares the people who did each action with the people who never did: how many of each are still active in their second week. Actions are ranked by that gap, with the number of people behind each, and thin ones are flagged. It's a correlation, not a cause, so the top candidates become tests.