GuideExit intent popups, and the three signals behind them.
“Exit intent” is not one trigger. It is a cursor heuristic that only works on desktop, a scroll heuristic standing in for it on mobile, and an idle timer doing something different again. Knowing which one is firing is most of knowing why it converts.
What is actually being detected.
Three signals, three failure modes.
Cursor leaves toward the browser chrome
The page watches the pointer's velocity and direction as it approaches the top edge. A fast upward move toward the address bar or the tab strip reads as leaving.
What it misses: Every touch device, because there is no cursor. Also anyone who leaves by keyboard shortcut, by switching tabs, or by closing the laptop.
A fast scroll upward
On a phone, the gesture that precedes leaving is a quick flick back toward the top of the page — reaching for the back button or the tab switcher.
What it misses: Someone re-reading a section they scrolled past. This is the signal most likely to fire on a visitor who was not leaving, which is why it deserves a tighter frequency cap than the desktop one.
Going idle
No scroll, no pointer movement, no keystroke for a set period. The visitor has not left the page but has left the moment.
What it misses: Nothing about intent — it is a proxy. It catches the reader who wandered off, which is a different person from the one about to bounce, and often a better one to interrupt.
Four times an exit popup is the wrong tool.
The part vendor guides tend to leave out.
On a page someone arrived at from Google
Google's intrusive interstitial guidance is about content being covered immediately after a search click, on mobile. Exit intent is not that — it fires on the way out, not the way in — but stacking an entry popup and an exit popup on the same page gets you there anyway.
In checkout
A visitor with their card out is not a lead to capture. Interrupting the payment step to ask for an email trades a sale for an address. Fire on the cart, not on the checkout.
When you have nothing to say
An exit popup with no offer, no answer and no reason to stay is a tax on leaving. If the honest content of the popup is “please don't go”, the page has a different problem and the popup will not fix it.
To the same person, repeatedly
The second exit popup in a session is noise, and the third is a reason not to come back. Frequency capping matters more on this trigger than on any other.
How this works in BetterPopup.
All three signals are separate trigger conditions rather than one “exit intent” switch, so you choose which to fire on and can give the mobile one a different frequency cap from the desktop one. Combine them with a scroll depth or a cart condition if you only want the popup on visitors who got far enough to matter, and every signup still lands in whichever of the 6 connected providers you use.
- When mouse leaves the page
- When user scrolls up quickly
- After user is idle
Exit intent, answered plainly.
Including the two questions asked by people trying to close one.
A popup that appears when a visitor looks like they are about to leave the page, rather than after a timer or a scroll depth. On desktop it is usually triggered by the mouse moving quickly toward the top of the browser window; on mobile, where there is no cursor, it is normally a fast upward scroll or a period of inactivity. The point is to make the offer at the last useful moment instead of interrupting someone who is still reading.
You need something that can watch for the signal and something to show. In BetterPopup: build the popup, add a trigger, and pick one of the 3 exit conditions — mouse leaving the page, a fast scroll up, or an idle timeout. Set a frequency cap so the same visitor does not meet it twice, choose who it applies to, and publish. No theme code.
Not in the desktop sense — there is no cursor to track, so true exit intent is impossible on a touch screen. What tools call mobile exit intent is a proxy: a fast scroll upward, or inactivity. Both are useful and both fire on some people who were not leaving, so they need tighter frequency limits than the desktop trigger.
Not on their own. Google's intrusive interstitial guidance targets content being obscured immediately after a visitor arrives from search, particularly on mobile. A popup that fires on the way out does not match that pattern. The risk comes from stacking it with an entry popup so that a visitor meets an overlay both coming and going.
As a visitor: reload the page, or close and reopen the tab — most popups store a per-session flag that a reload clears. If it survives that, the site has set a cookie; clearing site data for that domain removes it. As a store owner seeing this on your own site, it usually means two campaigns are targeting the same visitor with no frequency cap between them.
We are not going to publish a number we cannot source. Rates vary enormously by traffic mix, offer and industry, and every benchmark you find on a vendor site is drawn from that vendor's own customers. Measure your own baseline first, change one thing, and compare against it.