Define your activation metric before you redesign onboarding

Teams often redesign onboarding by taste: fewer steps, nicer tooltips, a welcome video. Without a definition of what “activated” means, there is no way to tell if any of it worked. An activation metric gives onboarding a single job and a number to move.

Why it works

New users decide quickly whether a product is worth their time. If they reach the point where it does something useful for them, they have a reason to come back. If they never get there, they drift away however polished the tour was. An activation metric names that point so the whole team can see it.

A single number also changes how onboarding work is prioritised. Every idea can be judged by one question: does it get more people to the action, faster?

How to do it

1. Define retained. Pick a retention signal that matches how often the product should be used, for example, active in week four for a weekly-use tool.
2. List candidate early actions users can take in their first few days: created a first item, connected an integration, invited a teammate, imported data.
3. Compare retained vs. churned users. For each candidate, look at what share of retained users did it early versus churned users. You are looking for actions retained users did far more often.
4. Add a threshold and a time window. “Created 3 projects in the first 7 days” is actionable; “uses projects” is not.
5. Validate with a test. Change onboarding to push more users towards the action and check whether retention actually moves, not just the action count.
6. Make it the onboarding north star. Every onboarding step should either get users closer to that action or get out of the way.

Example

A B2B reporting tool finds that accounts which connect a data source in the first two days retain much better than those that don’t. The team cuts the onboarding tour, moves “connect your data” to the first screen, and offers a sample dataset for users without access yet. Activation is now tracked as “data source connected within 48 hours” on the weekly dashboard.

Metrics to track

• Activation rate: the share of new signups who complete the action inside the time window.
• Time to activation for users who get there. Shorter is usually better.
• Week-four retention of activated and non-activated users. The gap should stay large. If it narrows, the metric may have stopped predicting retention.
• Activation rate by signup source, to spot channels that bring users who never reach value.

When it backfires

• Picking a vanity action. “Viewed the dashboard” is easy to inflate and rarely predicts retention.
• Forcing the action. Making a step mandatory raises the metric while frustrating users who were not ready. Retention may not follow.
• Never revisiting it. As the product and audience change, re-check that the metric still predicts retention.

Questions people ask

What is an activation metric?

An early, measurable action, usually within the first days, that new users who go on to retain tend to complete, such as creating a first project or inviting a teammate.

Is correlation enough to pick an activation metric?

It is a starting point, not proof. Treat the action as a hypothesis, then test whether nudging more users towards it actually improves retention.

Can a product have more than one activation metric?

It can if you serve very different user types, such as account admins and invited team members. Keep one metric per segment so each group has a single clear target.