How this instrument works
Exit rate is the share of one screen's own views that were also the last thing a visitor looked at before leaving the site entirely. The denominator is every recorded view of that single URL, not every session across the whole site, and the numerator counts a visit no matter how many other screens it crossed first — a shopper who browsed six categories and then left from the seventh counts here exactly the same as one who arrived on it directly and left within seconds.
A product manager reads exit rate step by step through a checkout flow to find which screen is quietly losing customers before payment, an editor reads it across a series of articles to see which entry the audience treats as a natural stopping point, and a researcher pulls it for a single screen inside an app-like flow to decide whether a redesign moved the leak or just relocated it. The figure names a location, not a person, so it says where attention stopped without explaining why — that answer comes from session recordings, exit surveys, or a look at what that screen actually offers once someone lands there.
Typical figures vary enormously by a screen's job. A thank-you or order-confirmation screen commonly exits near 100% because the visit is finished by design, and a homepage or a hub screen usually sits far lower because its whole purpose is sending traffic onward. The common misreading treats any exit rate as a verdict on quality, when the number only becomes meaningful once you know what that screen was built to do — a high reading on a step meant to hold visitors is a signal worth chasing, and an identical reading on a screen meant to end a journey is just the plan working.
- Enter Exits from this page — the count of visits for which this was the last screen viewed before the session ended.
- Enter Total page views — every recorded view of this same screen across the same date range.
- Read Exit rate, % — the instrument divides exits by views and multiplies by 100.
- Compare the figure against the screen's role in the funnel rather than against one universal benchmark.
Worked example — 180 exits out of 1,200 views
A mid-funnel product category screen logs 1,200 views over a month, and 180 of those views were the last screen a visitor saw before leaving the site. Enter 180 as Exits from this page and 1,200 as Total page views, and the instrument returns 180 ÷ 1,200 × 100 = 15%.
Fifteen percent reads as unremarkable for a category screen, since most shoppers are expected to click through to an individual product before they buy or leave; a step deeper in the same funnel, say a shipping-details screen inside checkout, posting that same 15% would be worth investigating, because almost everyone who reaches that step should be seeing a payment screen next, not leaving the site.
Questions
How is exit rate different from bounce rate?
The two metrics use different denominators. Bounce rate looks across every session on the whole site and flags only the ones that stopped after a single view; exit rate looks at one screen in isolation and asks what fraction of its own views were the final view of that visit, regardless of how long the visit ran beforehand. A screen deep inside a five-step flow can post a high exit rate while barely moving the bounce number, because almost nobody arrives there as their very first stop.
Is a 100% exit rate a problem?
Not on its own. A checkout confirmation, a downloaded receipt, or a support article that fully answers a question can legitimately end almost every session that reaches it — that is the screen doing its job, not losing anyone. Treat a high reading as a prompt to check the screen's intended role before treating it as a leak.
Why does exit rate use page views instead of sessions?
Because the question being asked is specific to one screen: out of everyone who looked at it, how many left the site from there? A single session can view the same screen more than once, and each of those views is a separate opportunity to exit, so counting sessions instead would understate how often that screen is the last thing seen.
How do I find which step in a funnel is losing the most visitors?
Run this calculator once for every screen in the sequence using each one's own exit and view counts, then compare the results against what each step is meant to do. A step positioned mid-funnel that exits far above the steps around it, rather than above some fixed number, is the one worth investigating first.
Can exits exceed total page views?
No — an exit is a subset of a screen's own views, so the count entered as exits should never be larger than total views for the same period. If your analytics export shows otherwise, check that both figures cover the identical date range and traffic segment before trusting the result.
Does a high exit rate mean the screen caused visitors to leave?
Not necessarily. The figure records where a session ended, not why — a visitor might have found exactly what they needed and left satisfied, been interrupted by something off-site, or hit a genuine problem with the screen itself. Pair the number with session recordings, exit surveys, or a support-ticket check before assuming the screen is at fault.
References
- GSA Digital Analytics Program — analytics.usa.gov
- U.S. Small Business Administration — Business Guide
Read this first: This instrument shows arithmetic, not advice. Real offers add fees, taxes and terms that vary by lender and place — verify the figures against your actual paperwork before deciding anything.