Finishing an onboarding checklist does not necessarily mean a new customer has experienced the reason for using a product. Someone can upload a profile photo, watch a tutorial and dismiss every welcome message without accomplishing the job that brought them there. Activation measurement tries to identify a more meaningful early milestone.

An activation metric tracks a defined action, or set of actions, intended to indicate that a new user or account has begun receiving the product’s core value. The definition is a hypothesis about value, supported by research and behavioural evidence rather than chosen solely because an event is easy to count.

Find a milestone that means something to the user

For a fictional collaborative reporting tool, creating a report may be a candidate milestone. Sharing it with a colleague who then uses it might better represent the value of collaboration, although that sequence takes longer and depends on another person’s availability. The right choice depends on what the product promises and who is expected to benefit.

Talk to recent customers about the point at which the product became useful, and examine early behaviour among users who continued and those who left. Amplitude’s guide to activating new users describes examining early engagement and later retention. Treat the resulting associations as candidates to investigate, not automatic instructions to push everyone through the same event.

Committed customers may invite colleagues because they already intend to keep using the product. Encouraging less committed customers to send invitations does not necessarily create that commitment. This is the distinction between correlation and causation, with practical consequences for onboarding design.

Define the event and the clock together

“Shared a report” leaves several questions unanswered. Does the event count at user or account level? Must another person open it? How soon after signup should it happen? Write those conditions into the metric, along with exclusions and the rule for counting the eligible population.

In a fictional example, 40 of 100 eligible new accounts share a report that another member opens within seven days. The seven-day activation rate is 40%. Accounts created yesterday have not had the full opportunity to qualify, so mixing them into that denominator would distort the comparison.

Time to activation can add useful context, but reporting it only among successful activators leaves out everyone who never reached the milestone. Read the timing measure alongside the activation rate and describe the observation window.

Improve the experience behind the event

Watch people who stall before the milestone. A missing permission, an empty account or an unclear invitation can each produce the same gap in the event data while requiring different changes. Usability testing and follow-up interviews help distinguish those cases.

After changing onboarding, examine both activation and the later outcomes it was intended to anticipate. A controlled experiment may help evaluate the effect of the intervention, but an increase in the event alone is insufficient evidence that customers received more value.

Revisit the definition as the audience and product evolve. Buyers, administrators and everyday users may reach value in different ways; forcing all of them into one milestone can reward an onboarding flow that serves only the easiest group to activate.