Funnel conversion calculator

A funnel is a sequence of steps, each passed by a share of the users from the previous step. Let U₁, U₂, …, U_n be the number of unique users at each step. The calculator computes:

The last line is the key property of a funnel: overall conversion is the product of step conversions. So the same relative lift on any step gives the same lift at the end: raising 50% to 60% on the third step is exactly as valuable as raising 70% to 84% on the first — both steps grew by 20%.

The what-if scenario

If the conversion of step k grows by a factor of (1 + u) and later steps keep their conversion, users at step k and at every later step are multiplied by the same factor:

The “later steps stay the same” assumption is not always realistic: if you push less motivated users through a step, they may drop off more often later. Treat the result as an upper bound and validate the change with an A/B test.

The “weakest point” of a funnel has two honest definitions, and the calculator shows both.

• Weakest step — the step with the lowest conversion CR_k, highlighted red on the chart. It answers “where does the process work worst”: if a step passes fewer than half of arrivals, there is usually a clear problem — a confusing screen, an extra field, no reason to continue.
• Most users lost — the step with the largest drop-off D_k in users. Early steps almost always lose more people simply because more traffic passes through them.

To prioritize, look at both numbers and at the cost of the change. From the product formula, the same relative lift on any step has the same effect at the end. But raising a conversion from 45% to 54% is usually easier than from 90% to 108% (which is impossible) — that is why low-conversion steps have more room to grow.

The onboarding funnel

In SaaS the most expensive leak often sits between sign-up and the first key action — the moment a user first gets value: creates a project, connects a data source, invites a teammate. Typical steps: sign-up → email confirmation → profile or onboarding → first key action → return within a week. How to choose the key action and get more users to it: How to increase user activation in SaaS; the onboarding steps themselves: User onboarding checklist.

• Events instead of users. If one person opened the checkout page three times, that is one user at the step, not three. Event-based funnels easily produce a “conversion” above 100%.
• Different periods for different steps. March sign-ups and March payments are different people: some payers signed up in February. A cohort funnel is more honest: everyone who signed up in March, and how many reached each step within a fixed window, say 14 days.
• Skippable steps. If users can enter the product without onboarding, “first key action” may have more people than “completed onboarding”. The calculator warns about such steps; build the funnel from strictly sequential steps or treat skippers as a separate segment.
• An average across channels. Paid and organic, mobile and desktop funnels often differ several-fold. The overall number can hide a problem that exists in one segment only.
• A window that is too short. “Returned on day 7” cannot be counted for users who signed up yesterday. Include only users whose observation window has ended.
• Conclusions without a test. A step conversion that rises after a change may be seasonality or a traffic shift. Validate with an A/B test — the A/B test calculator tells you how many users you need.

1. Croll A., Yoskovitz B. Lean Analytics: Use Data to Build a Better Startup Faster. O'Reilly Media, 2013 — funnels, choosing the key metric, user lifecycle stages.
2. McClure D. Startup Metrics for Pirates: AARRR! Talk, 2007 — the acquisition → activation → retention → revenue → referral funnel.
3. Amplitude Docs — Funnel Analysis chart: unique users, step order and conversion window (amplitude.com/docs).
4. Mixpanel Docs — Funnels: conversion window, uniques vs totals (docs.mixpanel.com).