An exit-intent popup is an overlay that appears when a visitor's behavior suggests they are about to leave: on desktop, the cursor moving toward the browser bar; on mobile, a quick scroll back up the page.
It gives you one last chance to save the visit with a cart reminder, a sign-up offer, or a way to save progress.
In our database, exit-intent popups usually help when they solve the reason someone is leaving, and the evidence is thinner than vendor case studies suggest.
Below is how the trigger works, what the A/B tests in our test database show, and the best practices those tests back up.
For popups shown at other times (on arrival, on a timer, after an add to cart), see the popups hub.
Exit intent: what it is and how the trigger works
"Exit intent" means guessing that a visitor is about to leave before they actually do.
The popup has to fire before the tab closes, so every tool works from early warning signs, not the exit itself.
Desktop: cursor movement
On desktop, the standard signal is the mouse heading for the tabs, address bar, or close button.
OptinMonster's documentation describes its trigger as firing "when a visitor's mouse motions towards the browser bar", with a sensitivity setting that sets how close to the window edge the cursor must get: 1 pixel on low, 50 on medium, 150 on high.
Higher sensitivity catches more real exits but also more false alarms, such as someone reaching for a bookmark or another tab they plan to come back from.
Mobile: no cursor, so weaker signals
Phones have no cursor, so mobile exit intent is a rougher guess.
OptinMonster says its mobile trigger fires "when the visitor quickly scrolls up the page", with sensitivity set as how far back up the visitor scrolls, measured as a share of screen height (50%, 30%, or 10%).
Its docs also say the mobile trigger does not touch the browser history or interfere with the back button.
A fast scroll up can also mean "I want to reread the price."
That is why many mobile setups use inactivity or time on page instead.
Two of the strongest lead-gen results in our database used idle-visitor triggers, not exit detection (covered below).
Exit intent and Google
Google's 2016 announcement on mobile interstitials, effective January 10, 2017, said pages whose content is not easily accessible "on the transition from the mobile search results may not rank as highly."
The examples it named were popups that cover content right after arriving from search or while reading the page, and standalone interstitials that have to be dismissed first.
A popup that fires only as someone leaves is not one of those named examples.
That does not make every overlay safe: Google's current guidance still recommends small banners over dialogs that block content.
Keep exit popups as exit popups, and don't let the same tool also fire a full-screen overlay on mobile arrival.
What exit-intent popups are used for
| Use | What the popup offers | Test in our database |
|---|---|---|
| Cart recovery | A reminder of the items already in the cart | Gelpro, +6% conversion rate (before-after) |
| Email capture | A discount in exchange for an email | Plant brand, +4% AOV |
| Form rescue | Save an application and finish later | IMB Bank, +52% applications saved |
| Channel capture | A different way to stay in touch | BlendJet, about 20% more sign-ups via Messenger |
What the exit-intent tests actually show
Our database has only a handful of tests where exit intent was the trigger.
They point the same way, but each comes with a catch worth knowing before you copy it.
An offer beat content. For a plants brand, Blend Commerce tested two exit-intent popups.
The control sent leaving visitors to relevant content; the variant offered a discount for an email sign-up.
The variant won: +4% AOV, +2.9% conversion rate, and +37% popup CTA clicks.
This was popup against popup, so it tells you which exit offer works better, not whether an exit popup beats having none.
A cart reminder helped, by a vendor's count. Gelpro Australia's exit-intent cart reminder, reported by OptiMonk, showed visitors the products already in their cart.
The result was +6% ecommerce conversion rate. It is a before-after comparison from the popup vendor, and the case study does not state the comparison basis, so treat it as a direction rather than a benchmark.
Saved progress went up, completions did not. IMB Bank added an exit popup to its personal loan form inviting applicants to save or continue.
Saved applications rose 52%, but there was no clear effect on completed applications during the test.
A popup can move its own metric without moving the one you're paid on.
The channel mattered. In a BlendJet cart abandonment popup, OptiMonk reports that asking for a Facebook Messenger subscription converted about 20% better than asking for an email. The downstream sales impact was not given.
@oliverkenyon
Blend Commerce
OptiMonkOptiMonk
@DaveDiederenExit popup best practices (backed by tests)
None of the rules below is a guaranteed lift.
Each is the pattern behind the results above plus the wider set of popup tests in our database, and the tests that broke the pattern are included too.
1. Decide whether you need a popup at all, with a holdout
Popups split opinion, and our data shows both sides.
Dave Diederen removed an email popup on a client store and it lost: -20.51% revenue per visitor and -12.18% conversion rate.
Oliver Kenyon ran the opposite result: removing an email popup from a top product page for a solar lighting brand was worth +$36,721/month.
Those were general email popups, not exit-intent ones, but the point carries over.
The only way to know is to test "popup vs no popup" on your own store.
OptiMonk's NuBest holdout did this and found +20.87% orders among visitors who saw the popup compared with those who didn't.
See holdout groups for how to set one up.
2. Match the offer to why people leave
The winning exit popups each answered a specific reason for leaving.
Gelpro's reminded people what was in their cart.
IMB Bank's dealt with the fact that a loan application takes time.
The plant brand's offer gave a price-sensitive visitor a reason to buy now.
Find the reason first. Exit surveys, session recordings, and checkout data from your cart abandonment work will tell you whether people leave over price, shipping cost, trust, or simply being interrupted.
3. Test the offer, not just the design
Offer framing moves popups far more than layout.
In popups that weren't exit-triggered, Convertibles found a $30 off $150 offer beat 15% off with +43% email capture and +$45,000/month in revenue.
A three-free-treats offer beat 15% off with +26% email capture.
A bigger discount is not automatically better. In an OptiMonk test for Sassy Scents, offering 15% instead of 10% saw sign-ups fall from 29% to 21.6%.
4. Measure revenue, not popup sign-ups
Popup tools report popup conversion rate by default, which is the easiest number to move.
IMB Bank's +52% saves with no clear completion effect is the cautionary case, and the North American insurer that ran an "Email My Quote" popup for idle mobile visitors saw +340.54% emails collected and no change in applications.
Use revenue per visitor or completed orders as the primary metric, and track discount cost and unsubscribes as guardrail metrics.
A discount popup that lifts orders can still lose money once margin is counted.
5. On mobile, test idle triggers alongside exit triggers
Because mobile exit detection relies on a quick scroll up, which is a rough signal, the timing that fires the popup is worth its own test.
The insurer above targeted mobile visitors who hadn't scrolled after 20 seconds.
CleanChoice Energy's email pop-in within 15 seconds beat a footer-only sign-up button by +204% email submissions.
Timing results can differ by device. In Clean Commit's test of a promo popup delayed by 60 seconds, the overall +7.9% conversion rate was inconclusive (23,328 visitors, 24% power), with mobile up 15.0% and desktop down 10.1%.
Split your analysis by device, and see our mobile CRO guide for other mobile patterns.
6. Show it once, and make it easy to close
An exit popup that reappears on every page turns a last chance into a nag.
Cap frequency per visitor, suppress it for anyone who has already signed up or bought, and keep it off checkout steps where it can interrupt payment unless you are testing a form-rescue offer like IMB Bank's.
Keep the close button visible and large enough to tap.
If a popup feels hard to escape, you are testing frustration, not the offer.
7. Give the test enough traffic
Exit popups only reach people already leaving, so the share of visitors who see one can be small.
Size the test on everyone assigned, not just those who saw the popup, and check statistical power before you start.
Clean Commit's 24% power result shows what happens when a popup test is too small to call.
Choosing a tool
Most popup apps include an exit-intent trigger, frequency caps, and targeting by cart contents or traffic source.
Compare them in our roundup of Shopify popup apps, and see email capture for what to do with the addresses you collect.
For more ideas across the whole site, browse site-wide tests or go back to the CRO guide.