How to Set Up Product Metrics in PostHog
PostHog is an open-source product analytics platform that combines event tracking, session recording, feature flags, and A/B testing in a single tool. For early-stage teams, PostHog reduces the complexity of managing multiple analytics vendors.
This guide covers setting up PostHog from initial instrumentation through building your first metrics dashboard.
Step-by-step guide
Install the PostHog snippet and enable Autocapture
Add the PostHog JavaScript snippet to your application. In the initialization config, set autocapture: true. Autocapture automatically tracks clicks, form submissions, and page views without any additional instrumentation. This gives you immediate data while you plan your custom event schema.
Define and instrument custom events for core actions
Identify your 10-15 most important user actions. Use posthog.capture() to track each one with a consistent naming convention: noun_verb (project_created, ticket_exported, prd_generated). Add properties to each event: the object ID, the user role, and any relevant metadata. Custom events are more reliable than Autocapture for analysis.
Set up Person Properties for segmentation
Use posthog.identify() when a user signs in, passing their user ID and key properties: email, plan, company name, role, and signup date. Use posthog.people.set() for properties that change over time (plan upgrades, feature flags). Person properties enable segmenting all metrics by user characteristics.
Build a Trends dashboard for core metrics
In PostHog, go to Insights > New Insight > Trends. Add your core active event (project_created or similar). Set the frequency to Weekly and the date range to Last 90 days. Save this as a Trend chart. Repeat for your 4-5 most important events. Add all charts to a new Dashboard named [Product] Metrics.
Create a Funnel insight for activation
Go to Insights > Funnel. Add your activation steps in order: user signed up, user completed onboarding, user created first project. Set the conversion window to 7 days. PostHog calculates the conversion rate at each step and shows the drop-off. Add this Funnel to your dashboard as the activation chart.
Set up Retention analysis for your core action
Go to Insights > Retention. Set the Starting event to your sign-up event and the Returning event to your core engagement action. PostHog displays a retention table by cohort. This shows the percentage of users who return each week after their first action — your key stickiness metric.
Use Session Recordings to diagnose drop-offs
In PostHog, go to Session Recordings > Filters. Filter by users who triggered your sign-up event but not your onboarding completion event. This shows you recordings of users who dropped off during onboarding. Watch 5-10 recordings to identify common friction points that quantitative data cannot explain.
Common mistakes
Relying only on Autocapture for analysis
Autocapture provides a good starting point but is unreliable for rigorous analysis — button text changes, CSS class renames, and DOM restructuring break Autocapture selectors. Always add custom events for actions that matter for product decisions.
Not calling posthog.identify() on login
Without identify(), all events before login are anonymous and cannot be attributed to a user. This breaks funnel analysis (you cannot see the full path from anonymous visit to conversion) and retention (you cannot track returning users). Always call identify() on successful login.
Ignoring the Feature Flags integration with analytics
PostHog feature flags automatically add a flag property to all events when a flag is active. This enables analyzing behavior of users in a flag group versus those not in it. Use this for every experiment and feature rollout.
Tips
Use PostHog Groups to track events at the organization level, not just the user level — essential for B2B SaaS
Set up PostHog Alerts (Webhooks) to notify Slack when a key metric drops below a threshold
Use PostHog Surveys to collect in-product feedback directly tied to user behavior data
Enable PostHog Heatmaps alongside Session Recordings to see aggregate click patterns on key pages
How Vantage helps
Vantage connects to product analytics tools to ground PRD generation in real user behavior data. When you add analytics context to a Vantage project, the AI generates specs that reference your actual activation and retention metrics rather than generic benchmarks.