How funnel analysis works
You define an ordered set of steps, each one an event, for example signup, onboarding_done,
order_completed. The funnel then shows, for a chosen time window, how many users completed each step,
the conversion rate between steps, and the overall conversion from first step to last. The biggest single drop-off
is usually where to focus.
Why it matters
Funnels turn a vague goal (“improve activation”) into a precise question (“why do 60% of users drop between signup and first action?”). Because each step is a real event, you can segment the funnel, by acquisition source, device, country, or any user trait, to see which groups convert and which get stuck.
Time between steps
A good funnel also shows the average time between steps, which separates “users abandoned” from “users just haven’t gotten there yet.” A step that takes days, not minutes, is a different problem than one people never reach.
A worked example
Say 10,000 people enter a signup funnel in a week:
signup: 10,000 (100%)onboarding_done: 6,200 (62% of the step before)order_completed: 1,488 (24% of the step before)
Overall conversion is 14.9% (1,488 ÷ 10,000), but the story is the middle step. The 38% drop at onboarding dwarfs the others, so a day spent there returns the most. Segment that step by source and you might find the loss is almost all paid traffic: a targeting problem, not a UX one.
How Pug does funnels
Pug’s Funnels insight measures conversion across ordered steps, with drop-off and the average time between steps, filterable by any property. Because Pug ties events to a person, you can build the funnel over real user behavior rather than anonymous sessions. You can also sketch the math first with the free funnel conversion calculator.