Activation Metric
An activation metric is the specific in-product behavior that statistically predicts a new user will become a long-term retained user. It's not 'signed up' (that's registration) and it's not 'logged in twice' (that's a vanity metric). It's a precise behavior — Slack's '2,000 messages sent in a team,' Facebook's '7 friends added in 10 days,' Twitter's '30 follows,' Pinterest's '7 pins repinned.' The activation metric is identified by analyzing cohort retention data: which behaviors, performed in the first session or week, correlate with users still being active 30/60/90 days later? Once identified, the entire onboarding flow gets redesigned to push new users toward that behavior as fast as possible.
The Trap
The trap is choosing an activation metric by intuition instead of cohort data. PMs pick 'completed onboarding' or 'invited a teammate' because they sound right. Real activation metrics often look counterintuitive (Slack's 2,000 messages took weeks to discover from data). The second trap: optimizing for the activation metric without verifying that the correlation is causal. If 7-friend Facebook users retain 90% but 2-friend users retain 30%, pushing users to add 7 friends might not lift retention if the underlying cause was 'users who naturally add 7 friends are more invested.' You need cohort experiments, not just correlation.
What to Do
Run a cohort analysis on 30-day retained vs churned users. For each in-product behavior in the first 7 days, compute the difference in retention between users who did and didn't do it. Rank behaviors by lift. The top behavior — adjusted for ease of nudging — is your candidate activation metric. Validate causality: run an experiment where you actively push half of new users toward the behavior; if the experimental group's retention rises, the metric is causal. Then redesign onboarding to make the activation metric the central goal of the first session.
Formula
In Practice
Sean Ellis is most known for the 'must-have' survey ('How would you feel if you could no longer use this product?'), but his deeper contribution to growth practice was the discipline of identifying activation metrics through cohort retention analysis. At Dropbox, LogMeIn, and Eventbrite, Ellis built the playbook of: (1) define long-term retention, (2) find behaviors that predict it, (3) instrument the funnel, (4) redesign onboarding to drive the behavior, (5) measure lift. The Slack '2,000 messages,' Facebook '7 friends in 10 days,' Twitter '30 follows,' and Pinterest activation metrics all came from variants of this playbook. (Source: Sean Ellis & Morgan Brown, Hacking Growth, 2017)
Pro Tips
- 01
Activation metrics often combine action AND time. 'Sent 5 messages' is weaker than 'sent 5 messages within first 24 hours' — the time constraint forces the user past the cold-start friction. The Facebook '7 friends in 10 days' framing is precise on both axes for a reason.
- 02
Multi-player products usually have multi-player activation metrics. Slack's was about TEAM message volume, not individual user behavior. If your product creates value through collaboration, the activation metric must capture the collaborative behavior, not the individual one.
- 03
Re-validate the activation metric every 6-12 months. As your product evolves, the behaviors that predict retention shift. A metric that drove your last $10M of growth may be the wrong one for your next $10M.
Myth vs Reality
Myth
“Every product has an activation metric — you just need to find it”
Reality
Some products have weak or no clear activation signal because users hit value through wildly different paths. Personalization tools, exploratory analytics, and creative apps often resist a single activation metric. In those cases, segment-specific activation metrics (one per major use case) work better than a forced single number.
Myth
“You should pick the activation metric with the highest correlation to retention”
Reality
Highest correlation isn't always best. The metric must also be (1) influence-able by the product (you can nudge users toward it), (2) early enough to matter for onboarding, and (3) causal not just correlated. A metric with 0.85 correlation that you can't influence is a measurement artifact. A metric with 0.65 correlation that you CAN influence drives real lift.
Try it
Run the numbers.
Pressure-test the concept against your own knowledge — answer the challenge or try the live scenario.
Scenario Challenge
You're a PM at a note-taking app. You analyze 90-day retention and find users who created 3+ notes in their first session retain at 68%; users who created 0-2 notes retain at 14%. Your designer suggests making the onboarding force users to create 3 notes before they can use the app.
Industry benchmarks
Is your number good?
Calibrate against real-world tiers. Use these ranges as targets — not absolutes.
Activation Lift (retention pp difference)
Difference in 30-day retention between users who did vs didn't take the candidate activation behaviorStrong signal
> 30 pp
Solid signal
15-30 pp
Weak signal
5-15 pp
Noise
< 5 pp
Source: Mixpanel & Amplitude retention analytics guides
Real-world cases
Companies that lived this.
Verified narratives with the numbers that prove (or break) the concept.
Slack
2014-2016 (activation discovery)
Slack's growth team analyzed retention curves across thousands of teams that had signed up for the product. The signal was clear: teams that sent 2,000 messages had >90% probability of long-term retention; teams below 2,000 messages churned at high rates regardless of any other behavior. Slack reorganized onboarding around accelerating teams to that 2,000-message threshold — encouraging integrations, channels, and direct messages that drove organic message volume. The activation metric became the org's north star for early-team engagement.
Activation threshold
2,000 team messages
Retention if threshold hit
>90%
Retention if not hit
Sharp drop-off
Discovery method
Cohort retention curve analysis
The right activation metric is hidden in cohort data, not in product intuition. Slack's number wasn't 'sent 5 messages' or 'created 3 channels' — it was a team-level threshold that no PM would have guessed without the analysis.
2007-2010
Facebook's growth team — led by Chamath Palihapitiya — identified that users who added 7 friends within their first 10 days had dramatically higher long-term retention than users who didn't. The team rebuilt onboarding around accelerating new users to that threshold: friend recommendations, contact import, suggested connections, and notifications all served the 7-friends-in-10-days metric. The growth team optimized for this single number for years; it became one of the most-cited activation metrics in growth literature.
Activation threshold
7 friends in 10 days
Retention lift if hit
Substantial (specific numbers private)
Onboarding mechanisms built
Friend recs, contact import, suggested connections
Growth team focus
Sustained over multiple years
Once an activation metric is identified, the entire onboarding flow should serve it. Facebook didn't make 7-friends a side goal — it became the central design constraint of new-user experience.
Twitter & Pinterest
2010-2013
Twitter discovered that users who followed 30+ accounts in their first session retained dramatically better — onboarding was redesigned around suggested follows. Pinterest found that users who repinned 7+ pins early were highly likely to retain — the home feed was reorganized to surface highly repinnable content for new users. Both cases followed the same pattern as Slack and Facebook: cohort analysis surfaces a counterintuitive numeric threshold; product redesigns to push users toward it.
Twitter activation
30 follows
Pinterest activation
7 repins
Discovery method
Cohort analysis of retained vs churned users
Common pattern
Action + count + early-time-window
The action+count+timing pattern recurs across products. The specific number varies, but the structure of a real activation metric is consistent.
Decision scenario
Picking the Right Activation Metric
You're VP Product at a 50-person SaaS. Cohort analysis surfaces three candidate activation metrics, each with different trade-offs. You can only optimize onboarding around one this quarter.
Candidate A
'Logged in 3 times in week 1' — 18 pp retention lift, 60% of users naturally hit it
Candidate B
'Created 5+ items AND invited 1+ teammate in week 1' — 38 pp lift, 22% naturally hit it
Candidate C
'Connected an integration in week 1' — 52 pp lift, 8% naturally hit it
Decision 1
You have one quarter to redesign onboarding around one of these metrics. The CEO is pressuring you to pick C (highest lift). The growth lead pushes for A (easiest to hit). Engineering says B has the cleanest implementation path.
Pick A — highest natural hit rate means smallest onboarding change required and lowest implementation riskReveal
Pick C — highest lift; redesign onboarding to push users toward integration setup in week 1Reveal
Pick B — strongest combination of lift AND influence-able baseline; redesign onboarding to push users toward item creation + teammate invite✓ OptimalReveal
Related concepts
Keep connecting.
The concepts that orbit this one — each one sharpens the others.
Beyond the concept
Turn Activation Metric into a live operating decision.
Use this concept as the framing layer, then move into a diagnostic if it maps directly to a current bottleneck.
Typical response time: 24h · No retainer required
Turn Activation Metric into a live operating decision.
Use Activation Metric as the framing layer, then move into diagnostics or advisory if this maps directly to a current business bottleneck.