Two Numbers Hide Behind One Drop-off Rate

What is drop-off rate?

In funnel analysis, drop-off rate is the percentage of users who enter a defined multi-step flow and leave before finishing it, calculated as users who exited a step divided by users who reached that step. The metric only exists once a team has written down the flow: an ordered list of steps, a first step that defines entry, and a last step that defines completion. Fix that sequence before pulling any number, because two analysts who pick different step lists report different drop-off for the same week of traffic.

What is the drop-off rate formula?

Per-step drop-off rate divides the users who reached a step but never reached the next one by the users who reached that step.

per-step drop-off rate = users who exited a step / users who reached that step x 100

End-to-end drop-off rate uses the whole flow as its denominator instead.

end-to-end drop-off rate = (users who entered - users who finished) / users who entered x 100

Take a checkout flow that 10,000 users enter, of whom 1,440 reach the confirmation page. End-to-end drop-off rate is (10,000 - 1,440) / 10,000, or 85.60%. If 6,000 users reach the product page and 3,000 of them go on to the cart, that single step's drop-off rate is 3,000 / 6,000, or 50.00%.

Why do per-step and end-to-end drop-off rate differ?

Per-step drop-off measures loss against the users who survived to that step, while end-to-end drop-off measures loss against everyone who entered, so the two numbers come from different denominators and never match. The same five-step funnel makes the gap concrete:

No single step in that table loses more than 50.00% of the people who reach it, and the flow still loses 85.60% of everyone who entered. Someone quoting "our drop-off rate is 20%" from the payment row and someone quoting "our drop-off rate is 85.60%" from the cumulative column are both reading the same funnel correctly. Label every drop-off figure with its step and its denominator in the same sentence, or the number carries no information.

The table also settles which step to fix first. Step 2 has the worst percentage at 50.00%, but step 1 loses 4,000 people against step 2's 3,000. Percentage ranks the leak, headcount ranks the revenue, and a conversion funnel review needs both columns visible.

Is drop-off the same as abandonment?

Drop-off and abandonment describe the same event under two vocabularies, and the split is about intent, not arithmetic. Analysts who separate the words reserve abandonment for flows where the user declared intent and then walked away, such as a cart holding an item or a partly filled form, and use drop-off for any exit from any step. Analysts who treat the words as synonyms compute the identical ratio and just name it differently.

Tool vendors have not settled this either. Google's GA4 funnel exploration documentation uses neither word and says users "fall out of the funnel", so a team moving between GA4, a product analytics tool and an ecommerce platform meets three labels for one calculation. Pick one word per team and write its denominator next to it.

How is drop-off rate different from bounce rate and exit rate?

Drop-off rate measures a flow, while bounce rate and exit rate measure a page, so the three metrics answer different questions and cannot be substituted for each other.

A page can post a low exit rate and still sit on a brutal drop-off rate, because users move on to somewhere other than the next funnel step. The bounce and exit split has its own denominators and its own worked example in bounce rate vs exit rate.

What decisions change the number before you calculate anything?

Four setup choices move drop-off rate without any change in user behavior. The first is the completion window: a funnel that gives users 30 minutes to finish reports higher drop-off than the same funnel measured over 7 days. The second is scope, because counting a flow per session splits every user who returns tomorrow into two entries, while counting per user merges them into one journey.

The third is ordering: a strict funnel discards a user who reaches checkout without the product page, an open funnel keeps them, and that shifts every later denominator. The fourth is re-entry, which decides whether a user who restarts the flow next week counts as a fresh entry or a continuation. Record all four rules alongside the funnel definition, because changing one silently is the most common reason a drop-off chart appears to improve.

What do published abandonment benchmarks tell you?

Public benchmarks describe abandonment in one specific flow, not drop-off in yours, so they set context and cannot set a target. Baymard Institute's running average of 50 different studies puts the documented online shopping cart abandonment rate at 70.22%, last updated September 22, 2025. That figure spans stores with different products, prices and checkout designs, which makes it an average of averages rather than a measurement of any one funnel. A store reporting 68% cart abandonment has therefore proven nothing about its own checkout, and a SaaS onboarding flow has no comparable published number at all. Compare your funnel to its own trailing weeks, and use cart abandonment rate when the flow in question is specifically a cart.

How do you find the cause behind a bad step?

A drop-off rate names the step that loses users, and session replay of that exact step names the reason. Flowsery records every user session and detects rage clicks, dead clicks, JavaScript errors, drop-offs and broken flows automatically, then groups matching sessions into one issue and ranks issues by how many users hit them. Its funnel analysis sits on the same event stream, so the worst step and the replays behind it come from one dataset instead of two tools that disagree.

A 50.00% step drop can be a broken button, a validation error nobody sees, or a price that surprises people, and those three look identical in a funnel chart. Repeated clicking on a control that does nothing shows up as a rage click cluster, which points at the element instead of the step. Each Flowsery issue lands in Slack, Linear or Jira with the replay and steps to reproduce attached, and tagging @flowsery in Slack opens a draft pull request in the team's GitHub repo.

Frequently Asked Questions

What is a good drop-off rate?

There is no cross-industry threshold, because drop-off rate depends on how many steps a team defined and where they placed the entry point. A two-step flow and a seven-step flow covering the same journey report different end-to-end numbers. Judge a step against its own history and the steps beside it.

Is drop-off rate the same as one minus conversion rate?

End-to-end drop-off rate and conversion rate sum to 100% when both use the same denominator and the same completion event. In the worked example, 1,440 of 10,000 entrants finish, giving a 14.40% conversion rate and an 85.60% end-to-end drop-off rate. Per-step drop-off does not have this relationship, since each step uses its own denominator.

Can drop-off rate exceed 100%?

No, because the numerator counts a subset of the denominator by construction. A figure above 100% means users are counted at a step they never reached, which points at a duplicated event or a step definition that fires on more than one page. Audit that step's event before trusting any other row.

Why does the same funnel show different drop-off in two tools?

Two tools apply different session timeouts, completion windows, and rules about whether steps must occur in order. Any one of those changes the denominator at every step after the first. Compare the two funnel definitions field by field before assuming either tool is wrong.

Should drop-off rate be measured by users or by sessions?

Measure by users when the flow can span more than one visit, such as a signup that people finish the next day, and by sessions when the flow is meant to complete in one sitting. Mixing the two inside one funnel produces steps whose denominators are not comparable. State which unit the funnel uses in the funnel name itself.

Does Flowsery detect drop-offs without extra setup?

Flowsery detects drop-offs and broken flows automatically alongside rage clicks, dead clicks and JavaScript errors, and it connects to PostHog or Amplitude replays already recorded with no re-instrumentation. The script is under 10 KB, cookie-free and EU-hosted, with no data sampling. Defining a named funnel is still worth doing, because it fixes the step list the drop-off numbers report against.

What counts as a step in a funnel definition?

A funnel definition needs an ordered list of steps, a first step that marks entry, and a last step that marks completion. Two analysts who pick different step lists report different drop-off rates for the same week of traffic. Write the step list down before pulling any number.

Why can a page have a low exit rate but a high drop-off rate?

Exit rate divides exits by pageviews of a single page, with no requirement that a sequence exist. Drop-off rate divides exits by the users who reached a step inside an ordered flow, so a user who leaves for anywhere other than the next funnel step counts against drop-off in a way exit rate does not capture. A page can therefore post a low exit rate while sitting on a brutal drop-off rate.

Is cart abandonment the same thing as drop-off rate?

Drop-off covers any exit from any step, while some analysts reserve abandonment for flows where the user declared intent first, such as a cart holding an item or a partly filled form. Baymard Institute's cart abandonment benchmark of 70.22% measures that specific flow, not a general funnel. Use cart abandonment rate only when the flow in question is a cart, and compare your own funnel to its own trailing weeks.

How does session replay help explain a bad drop-off step?

Flowsery records every session and automatically detects rage clicks, dead clicks, JavaScript errors, drop-offs and broken flows, then groups matching sessions into one issue. A 50.00% drop at a single step can come from a broken button, a validation error nobody sees, or a price that surprises people, and those three causes look identical in a funnel chart alone. Session replay of that exact step shows which one actually happened.