What Is a Product Funnel?

What does product funnel mean?

A product funnel is the sequence of steps a user moves through toward a specific goal inside a product, such as completing signup, finishing setup, reaching a first moment of value, or making a purchase. Sam Edwards, writing in Amplitude's funnel analysis guide, describes the method this way: "Funnel analysis is a method used to visualize, measure, and understand key user behaviors throughout the customer journey." At each step some users continue and others leave, and the funnel exists to show how many, where and how fast.

Even the last, shortest step leaks. The Baymard Institute calculates an average documented online cart abandonment rate of 70.22 percent, based on 50 different studies of e-commerce checkouts. That is a checkout benchmark, not a SaaS one, and the studies define abandonment in different ways, but it makes the point every funnel makes: people who were one step from the goal still walk away, and the only way to know why is to measure each step separately.

The shape that gives the funnel its name, wide at the top and narrow at the bottom, comes from that drop off. A funnel definition has three parts: the ordered steps, a time window for completing them, and a rule for what happens when someone completes steps out of order.

How do you read a product funnel?

Read every funnel on two axes: conversion from each step to the next, and time between steps. Amplitude's guide lists time to convert as its own dimension ("An important factor in analyzing a funnel is time"), and Mixpanel's funnel documentation frames the report around conversions "within a particular time window". Conversion tells you where people leave. Time tells you where people get stuck before they leave, or before they eventually convert.

The table below is an illustrative funnel for a fictional invoicing tool for small agencies, not benchmark data. It shows why one axis is not enough.

Step	Accounts	Conversion from previous step	Median time from previous step
Signed up	1,000	n/a	n/a
Setup done: client list imported, bank connected	580	58 percent	1 day
First invoice sent	430	74 percent	9 days
First invoice paid through the tool (activation)	250	58 percent	6 days
Upgraded to a paid plan	70	28 percent	21 days

Read on conversion alone, the biggest loss is between signup and setup: 420 accounts gone. Read on time, the step that hurts most is setup to first invoice: 74 percent of accounts get there, but the typical one waits nine days, and every one of those days is a day the account can forget the product exists. Those are two different problems with two different owners. The signup-to-setup loss is a question about fit and friction (are these the right accounts, and is the bank connection too hard?). The nine-day wait is a question about what happens after setup (does the product give the account a reason to send an invoice today?).

What is the difference between a drop and a stall in a funnel?

A drop is a step where a large share of people never arrive at the next step. A stall is a step where people do arrive, but slowly, so the median time between steps is long compared with the rest of the funnel. Drops show up in the conversion column, stalls in the time column, and a funnel chart that only shows bars hides every stall.

The two call for different fixes. A drop usually points to friction or a mismatch: a form that asks for too much, a permission the user does not have, a promise on the landing page the product does not keep. A stall usually points to a missing trigger: nothing in the product, or in the email that follows it, gives the user a reason to take the next step now. Removing a field from a form does nothing for a stall, and a reminder email does nothing for a user who cannot connect their bank.

A useful habit is to write the funnel review as two sentences every time: "The biggest drop is between X and Y, where Z percent leave." and "The longest stall is between A and B, where the median wait is N days." If you cannot fill in both, the funnel is not instrumented well enough yet.

Why should setup and activation be separate funnel steps?

Setup is the point where an account is ready to get value; activation is the first time it actually gets value. Many funnels merge the two, which makes the setup rate look like an activation rate and hides the stall that usually sits between them. In the illustrative table above, 58 percent of signups finished setup but only 25 percent of signups reached a paid invoice, the moment the product proved itself.

Keep them as separate steps with separate events, and name the activation step after the value moment, not after the last configuration screen. The guide on how to define an activation metric covers how to pick that event and test it against retention, and the activation rate entry covers the resulting number.

Should a funnel track failures as well as completions?

Yes. A funnel built only from success events tells you how many people made it through each step, not what happened to the rest. Track the failure alongside each step (a bank connection that errored, an import that was rejected, a payment that bounced) as its own event with properties that say why. Then the drop between two steps splits into people who tried and failed and people who never tried, which are different problems.

Segment before you conclude anything. A funnel's overall conversion can hold steady while one acquisition channel collapses and another improves, so compare the same funnel by channel, plan, company size and signup cohort before deciding which step to fix. For drafting those comparisons faster, see AI for funnel analysis.

Product funnel example

The practical challenge in Builders Camp's Product Analytics bootcamp starts from a B2B field service tool, Fieldly, that tracks exactly one event across its product: user_login. To build a real funnel, the PM has to define six steps, from viewing the signup page through completing account creation, connecting a first field team member, creating an inspection template, assigning an inspection, and finally submitting a completed inspection.

For each step, the challenge asks for the exact event name, the properties to attach, the account group it belongs to, and whether a failure variant should be tracked as well. That turns a single vague login event into a funnel that can show where field service teams drop, and where they stall, before reaching real value.

Where does the product funnel fit in Builders Camp's bootcamps?

The Product Analytics bootcamp is a one-week program with 2 live sessions and 9 microlessons, directed by Mário Araújo and part of the Growth Specialist Track and the Data & Analytics Specialist Track. Its published topics include B2B analytics instrumentation, activation metrics and aha moments, account-level analysis and cohort retention analysis. For the funnel as one part of a full growth model, Growth for Product Managers is a two-week bootcamp whose published topics include acquisition loops, retention and engagement loops, and monetization.

Builders Camp runs live and self-paced bootcamps in product management and AI product building. See the Product Analytics bootcamp for the next cohort dates.