Why the gap is invisible
Acquisition and retention are both well-instrumented in most companies. Signups are counted, churn is reported, and both have owners. The step between them frequently has neither.
That's a problem because the step between them is where most of the loss happens. Users who never reach the point where the product does something useful don't churn for product reasons — they leave because nothing happened, and no win-back campaign addresses that.
The symptom is a stubborn disconnect between top-of-funnel growth and revenue. More signups, same revenue. That pattern almost always means the added users are failing to activate.
Choosing the event
The defining work is picking which action counts, and the temptation is to choose something that feels like commitment: completed profile, invited a teammate, connected an integration.
Feel is the wrong criterion. Derive it instead: take users who retained past 90 days and users who churned early, and find the action that most separates the two cohorts. Frequently it's something modest — created one project, ran one report, completed one workout.
Then check it holds. If activated users retain dramatically better than non-activated ones, you've found a real predictor. If both cohorts churn at similar rates, the event isn't measuring what you think and you should look again.
If your activation event doesn't produce a sharp retention split between activated and non-activated users, it's the wrong event.
The window is not optional
Activation without a time window is close to meaningless. 'Reached the event eventually' tells you nothing about whether onboarding works, because a user who activated in month four did so despite your first-run experience rather than because of it.
Pick the window from your product's natural rhythm. Consumer tools with immediate value should measure 24 hours. B2B software with an implementation step is reasonable at 7–14 days. Anything longer and you're measuring persistence rather than onboarding.
The window is what makes the metric actionable. It turns a vague quality question into a specific one: does a new user reach value while intent is still high?
Where activation actually breaks
Once you're measuring it, the failures cluster into a few recognisable shapes.
Setup before value
Products that require configuration before showing anything useful lose users during setup. Showing value on sample or default data first, then configuring, consistently outperforms.
Expectation mismatch at acquisition
Activation frequently differs sharply by channel. If one source activates at half the rate of others, the ads or landing page are promising something different from what the product does.
The empty state
A new account with no data is the least persuasive screen in most products, and usually the least designed. Templates, samples, and a single obvious next action all beat a blank page.
Requiring the wrong first action
Onboarding that pushes users toward what the business wants — invite colleagues, connect billing — rather than toward what delivers value delays the moment that decides retention.
Fixing it usually isn't a marketing job
This is where the metric gets organisationally awkward. Activation sits in the product experience, which marketing typically doesn't own, while the consequences land squarely on marketing's numbers.
The productive framing is that activation is a shared metric rather than a handoff. Marketing controls the expectations users arrive with, which materially affects whether they activate. Product controls what happens once they do arrive. Neither can fix it alone.
In practice, the highest-leverage work we do in SaaS engagements is frequently in onboarding rather than in anything traditionally called marketing. We'd rather say that than optimize a signup form feeding a funnel that loses four users in five.
Terms used in this piece
We do this for a living
If you'd rather not build this yourself, these are the services where it lives.