Power users drive disproportionate value for your product—but only if they stay engaged. In Mixpanel, you need a systematic way to identify who they are, track whether they're staying active, and catch churn signals before they leave. We'll show you how to build cohorts, set up retention tracking, and create dashboards that alert you to changes in power user behavior.
Define and Segment Power Users with Cohorts
The first step is creating a clear definition of what "power user" means for your product. In Mixpanel, you do this with cohorts.
Create a cohort based on event frequency
Go to Cohorts in the left sidebar and click Create Cohort. Set up a rule that captures users with high activity. For example, find users who fired your core event—like opened_editor or ran_analysis—at least 10 times in the last 30 days. This becomes your baseline power user segment.
// Track the core engagement event consistently
mixpanel.track('opened_editor', {
'editor_type': 'sql',
'session_id': 'sess_123abc',
'timestamp': new Date().toISOString()
});
// Or use Mixpanel's retention API to query cohort members:
// GET /api/2.0/cohorts
// GET /api/2.0/cohorts/[cohort_id]/membersRefine the cohort with property filters
Beyond event count, layer in property filters. Click Add filter and chain conditions: subscription_tier equals 'premium', account_age is greater than 90 days, or total_features_used is at least 5. This narrows your power user definition to the users who actually matter to your business.
// Set user properties that define power users
mixpanel.people.set({
'subscription_tier': 'premium',
'account_age_days': 145,
'total_features_used': 7,
'last_active_timestamp': Date.now()
});
// Update properties on the fly
mixpanel.people.increment('total_features_used', 1);Track Retention and Spot Churn Signals
Now that you have your power user cohort, monitor whether they're staying engaged or showing warning signs of churn.
Run a retention analysis on the power user cohort
Go to Retention and select your power user cohort in the filter pane. Choose a key engagement event—like ran_query or exported_report—as your return event. Set the interval to weekly. This shows you what percentage of power users are returning each week. A declining retention line means you're losing engagement.
// Track return visits with consistent event naming
mixpanel.track('ran_query', {
'query_type': 'saved_report',
'duration_ms': 2400,
'user_cohort': 'power_user'
});
// Query retention via Mixpanel API:
// GET /api/2.0/retention?from_date=2026-03-19&to_date=2026-03-26&retention_type=birth&event=ran_query&interval=week&unit=daySet up a funnel to catch drop-off points
Create a funnel: 1) opened_editor, 2) created_query, 3) ran_query, 4) exported_results. Apply your power user cohort filter. Watch the completion rate week-over-week. A sharp drop at any step tells you where power users are stalling.
// Track each funnel step
mixpanel.track('opened_editor', { 'funnel_step': 1 });
mixpanel.track('created_query', { 'funnel_step': 2 });
mixpanel.track('ran_query', { 'funnel_step': 3 });
mixpanel.track('exported_results', { 'funnel_step': 4 });
// Define funnel via API:
// POST /api/2.0/funnels
// { "event_selectors": ["opened_editor", "created_query", "ran_query", "exported_results"] }Flag at-risk users with a custom property
Run a daily server-side check: if a power user hasn't fired any event in 14 days, set risk_status: 'at_risk'. Then segment by this property in any Mixpanel chart. It's a simple flag that isolates users needing immediate attention.
// Server-side check (Node.js, run daily)
const daysSinceActive = Math.floor((Date.now() - user.last_active_ms) / (1000 * 60 * 60 * 24));
if (daysSinceActive > 14 && user.cohort === 'power_user') {
mixpanel.people.set(user.id, {
'risk_status': 'at_risk',
'days_inactive': daysSinceActive,
'risk_flagged_date': new Date().toISOString()
});
mixpanel.track('user_flagged_at_risk', {
'user_id': user.id,
'days_inactive': daysSinceActive
});
}Build a Monitoring Dashboard
Consolidate everything into a dashboard you check weekly.
Create a power user dashboard with key metrics
Go to Dashboards and create a new one. Add cards for: 1) Count of active power users (segmented by your cohort), 2) Weekly retention rate, 3) Funnel completion percentage, 4) Count of at-risk users. Arrange them in a grid so trends jump out.
// Log dashboard interactions for tracking
mixpanel.track('dashboard_opened', {
'dashboard_name': 'power_users_monitoring',
'viewer_cohort': 'power_user'
});
// Query dashboard data via API:
// GET /api/2.0/events?event=opened_editor&unit=day&interval=7
// GET /api/2.0/retention?event=ran_query&retention_type=birth&interval=week
// GET /api/2.0/funnels?event=["opened_editor","created_query","ran_query"]Set up alerts for sudden changes
Configure an alert rule: if power user retention drops below 55% week-over-week, or if the count of at-risk users spikes by more than 25%, send a Slack notification. This gives you early warning before churn becomes a serious problem.
// Alert logic (run daily on your server)
const currentAtRisk = await getAtRiskUserCount();
const previousAtRisk = await getAtRiskUserCount('7_days_ago');
const percentChange = ((currentAtRisk - previousAtRisk) / previousAtRisk) * 100;
if (percentChange > 25) {
sendSlackAlert({
'channel': '#product-alerts',
'text': `⚠️ At-risk power users up ${percentChange.toFixed(1)}%. Check dashboard.`
});
}Common Pitfalls
- Defining power users too narrowly (50+ events/month). You might miss high-value users who use intensively but less frequently. Test multiple thresholds and validate against revenue or satisfaction data.
- Forgetting to apply your power user cohort filter in retention and funnel analyses. Without filtering, you're measuring the entire user base—the signal gets drowned out.
- Chasing engagement drops without understanding why. A power user who stops using your product might have solved their problem, switched tools, or hit a bugs. Add a post-churn survey.
- Skipping the automated at-risk flagging. If you only check dashboards manually, you'll miss the warning signs until it's too late. Automate the flagging daily.
Wrapping Up
You now have a system to identify power users, track their health, and catch churn signals early. Monitor retention weekly, check your dashboard, and act on alerts. If you want to track this automatically across tools, Product Analyst can help.