Skip to main content

Funnels

A funnel measures how many users progress through an ordered sequence of events — and where they drop off.

What a funnel is​

A funnel is defined by:

  • Steps — an ordered list of events, e.g. signed_up → created_project → sent_first_event
  • Conversion window — the maximum time a user has to complete all steps (optional; unlimited if unset)
  • Date range — which cohort of users to look at
  • Breakdown — optionally segment results by a property (e.g. platform)

For each step you get the number of users who reached it and the conversion rate, plus an overall first-step-to-last-step conversion:

Signed Up          1000   100%
Created Project     680    68%
Sent First Event    450    45%

A user counts for a step only if they performed the step events in order within the window. The dashboard can also suggest funnels based on common paths in your data.

How to instrument for funnels​

  1. Track a distinct event per step. Each funnel step matches one event name. If your onboarding is one onboarding_step event with a step property, you can't build a step-per-screen funnel from it — track onboarding_started, profile_completed, onboarding_finished instead (or both).
  2. Keep identity consistent across steps. Funnels join steps on the user ID. Call identify() at login/signup so pre- and post-signup steps link up — the anonymous ID aliasing handles the signup boundary automatically.
  3. Fire step events at the moment of success, not intent. Track purchase_completed when the receipt confirms, not when the buy button is tapped (that's its own step).
// A checkout funnel worth building
track('product_viewed', { product_id: 'SKU123' });
track('added_to_cart', { product_id: 'SKU123' });
track('checkout_started', { cart_value: 89.99 });
track('purchase_completed', { order_id: 'ORD123', total: 89.99 });

Availability​

Funnels are available on paid plans.

Next steps​

  • Retention — what happens after they convert
  • Experiments — A/B test the drop-off points you find