An exit intent popup is the last thing a visitor sees before they leave. Done right, it quietly recovers 2% to 8% of abandoning traffic as email subscribers or orders. Done wrong, it inflates your bounce rate, trains people to hate your brand, and on mobile it can even put you at odds with Google’s intrusive interstitial guidelines.
The difference is almost never the design. It is the trigger configuration: how long you wait, how far the visitor scrolled, how often the popup is allowed to fire, and which pages it is banned from. This guide gives you the exact settings we deploy for clients, plus the mistakes that cause measurable damage.
What is an exit intent popup?
An exit intent popup is an overlay that appears when a visitor shows behavioural signals that they are about to abandon the page. On desktop, exit intent technology tracks cursor velocity and direction; when the mouse accelerates toward the top of the viewport (the address bar, the back button, the tab close icon), the script fires the overlay. On mobile, where there is no cursor, detection relies on other signals such as back button presses, rapid upward scrolling, or inactivity.
In plain terms: it is a last-chance offer shown to traffic you were going to lose anyway. That single fact should shape every setting you choose.
How exit detection actually works
- Mouse-out detection (desktop): the script listens for a
mouseleaveevent on the document or a cursor crossing the top boundary of the viewport above a minimum speed threshold. - Velocity filtering: good tools ignore slow, accidental cursor drifts and only fire above a set pixel-per-second threshold, which cuts false positives dramatically.
- History manipulation (mobile): a pushState entry means the first back-button press shows the popup instead of leaving the page. Use this with extreme care, it frustrates users and can look like a navigation trap.
- Scroll-velocity and idle triggers (mobile): a fast scroll back to the top or 15 to 30 seconds of inactivity are softer, more honest proxies for exit intent.
Why most exit intent popups hurt conversions
Before the rules, here are the configuration errors that reliably produce bounce spikes, session-length drops and complaint emails. If your popup “stopped working”, you are probably doing at least two of these.
- Firing on first scroll, within 2 seconds. The visitor has read nothing, has no reason to care, and the only possible action is to dismiss.
- No frequency cap. Showing the same overlay on every page view of every session is the single fastest way to burn goodwill.
- Running on checkout, cart and form pages. You interrupt people who are already converting, often with a discount they were not going to ask for.
- Using an aggressive mobile full-screen overlay on landing pages that receive search traffic. This is where Google’s intrusive interstitial guidance bites.
- Hidden or tiny close buttons. Hard-to-dismiss overlays increase rage clicks and back-to-SERP behaviour.
- One generic offer across the whole site. A “10% off your first order” popup on a B2B comparison article converts nobody.
- Measuring only popup conversion rate. A 9% capture rate means nothing if sitewide revenue per session fell.
The 7 rules for exit intent popups that capture instead of annoy
Rule 1: Add a qualification delay before the popup is allowed to fire
Never arm the trigger the moment the page loads. Require a minimum time on page so that only engaged visitors are eligible.
- Blog and content pages: arm after 15 to 20 seconds.
- Ecommerce product pages: arm after 8 to 12 seconds.
- Landing pages and SaaS pricing: arm after 10 to 15 seconds.
- Homepages: arm after 20 seconds, homepage traffic is exploratory and early exits are normal.
This one setting typically removes 30% to 50% of impressions while barely touching conversions, because the removed impressions were never going to convert. Fewer impressions, same leads, better experience.
Rule 2: Require a minimum scroll depth
Time alone can be faked by an open tab in the background. Combine it with scroll depth so the trigger only arms for people who actually consumed something.
- Long-form articles: 30% to 40% scroll depth.
- Product pages: 25%, enough to reach the price and variants.
- Short landing pages: 50%, since the page is short to begin with.
- Category pages: 20%, or two to three product rows.
Set the condition as time AND scroll, not OR. The AND logic is what separates a polite offer from a reflex interruption.
Rule 3: Cap frequency hard
Frequency capping is the rule most teams skip and most users feel. Our standard baseline:
| Setting | Recommended value | Why |
|---|---|---|
| Max per session | 1 | One ask per visit, never per page view |
| After dismissal | Suppress 14 to 30 days | A close click is an answer, respect it |
| After conversion | Suppress permanently | Never ask a subscriber to subscribe again |
| Second dismissal | Suppress 90 days or forever | Two refusals is a clear signal |
| Known customers / logged-in users | Exclude or swap offer | Discounting existing buyers erodes margin |
Store the suppression flag in a first-party cookie or localStorage, and remember that privacy settings on Safari can shorten cookie life to seven days. If your audience is Safari-heavy, use a server-side or account-level flag for the “already converted” state.
Rule 4: Exclude the pages where interruption costs you money
Page exclusions are the cheapest insurance policy in conversion rate optimisation. Our default blocklist:
- Cart, checkout and payment pages (unless you are running a dedicated cart-abandonment offer tested against a holdout)
- Order confirmation and thank-you pages
- Account, login and password reset pages
- Any page containing an open form (contact, demo request, quote)
- Support, documentation and help-centre pages
- Legal pages: privacy policy, terms, cookie policy
- Paid landing pages with a single conversion goal, where a competing offer splits attention
Also exclude by behaviour, not just URL: suppress the popup if the visitor has started typing in a form field, has an item in the cart and is on a product page, or arrived from your own email campaign with a tracked link.
Rule 5: Match the offer to the page intent
The biggest conversion gains come from segmenting the offer, not from redesigning the box. One global popup is a wasted asset.
| Page type | Best-performing offer | Typical capture rate |
|---|---|---|
| Blog article | Content upgrade tied to the article topic (checklist, template, swipe file) | 4% to 9% |
| Ecommerce product page | First-order incentive, free shipping threshold, or “save this cart” email | 2% to 6% |
| Category page | Quiz or product finder, back-in-stock alert | 2% to 5% |
| SaaS pricing page | Extended trial, pricing-objection FAQ, book-a-call slot | 1% to 4% |
| Service or agency page | Audit, teardown, benchmark report | 2% to 5% |
| Comparison or alternatives page | Side-by-side comparison PDF, migration guide | 3% to 7% |
Treat these ranges as planning benchmarks, not promises. Your traffic quality, offer strength and industry will move them.
Offer rules that consistently outperform
- Specificity beats generosity. “The 12-point pre-launch checklist we use for client stores” beats “Join our newsletter”.
- One field. Email only. Every extra field costs you conversions on an exit screen.
- Zero-party data instead of discounts when margins are thin: a two-click quiz (“What are you shopping for?”) captures intent and segments the list.
- Avoid blanket discounting on high-intent product pages. You will discount orders that were coming anyway. Always run a holdout group.
Rule 6: Treat mobile as a different channel entirely
True exit intent does not exist on touch devices. There is no cursor leaving the viewport. Everything else is a proxy, and the aggressive proxies (back-button hijacking, instant full-screen overlays) are exactly what Google’s page experience guidance warns about.
Google’s intrusive interstitial guidance targets pages where content is inaccessible to a user arriving from search, specifically: a popup covering the main content immediately after arrival, a standalone interstitial the user must dismiss before reading, and a layout where the above-the-fold area is occupied by the interstitial while the real content sits underneath.
What stays safe:
- Banners that use a reasonable amount of screen space and are easy to dismiss, such as a slim bottom bar.
- Legally required notices such as cookie consent and age verification.
- Overlays triggered by user action or appearing well after the visitor has engaged, rather than on arrival from the SERP.
Our mobile configuration:
- Use a bottom slide-in bar covering 25% to 30% of the viewport instead of a full-screen modal.
- Trigger on scroll depth 50% plus 20 seconds, or on a fast upward scroll, not on arrival.
- Do not hijack the back button on pages that receive organic search traffic.
- Keep the close affordance at least 44 by 44 px and visually obvious.
- Give mobile its own frequency cap and its own creative, then measure it separately.
Rule 7: Make dismissal effortless and measure the right numbers
A popup that is hard to close converts worse, not better, because the dominant emotion becomes irritation. Checklist:
- Visible X in the top right, high contrast, minimum 24 px icon inside a 44 px tap area.
- Click on the overlay background closes it.
- The Escape key closes it on desktop.
- No shaming copy on the decline link. “No thanks” is enough. “No, I prefer losing money” is a brand liability.
- Focus trap and
aria-modalattributes for accessibility, with focus returned to the page on close. - Lazy-load the popup script so it does not affect Largest Contentful Paint or Interaction to Next Paint.
Then track more than the capture rate:
| Metric | Healthy range | Red flag |
|---|---|---|
| Capture rate (submissions / impressions) | 2% to 8% | Below 1% |
| Impression rate (impressions / sessions) | 10% to 25% | Above 40%, triggers too loose |
| Sitewide conversion rate vs holdout | Flat or up | Any measurable drop |
| Pages per session | Stable | Drop after launch |
| Email list quality (open rate at 30 days) | Within 10% of other sources | Far below other sources |
| Mobile rage clicks on the overlay | Near zero | Any visible cluster in session replay |
Always keep a 5% to 10% holdout that never sees the popup. Without it you cannot tell whether the emails you captured came with a revenue cost.
Trigger settings cheat sheet by site type
| Site type | Min time on page | Min scroll depth | Frequency cap | Mobile format |
|---|---|---|---|---|
| Content / media blog | 15 to 20 s | 35% | 1 per session, 30-day suppression | Bottom bar at 50% scroll |
| Ecommerce store | 10 s | 25% | 1 per session, 14-day suppression | Slide-in, cart page excluded |
| SaaS website | 15 s | 40% | 1 per session, 30-day suppression | Bottom bar, logged-in users excluded |
| Lead-gen / local service | 12 s | 30% | 1 per session, 21-day suppression | Click-to-call bar instead of form |
| Paid landing page | 20 s | 50% | 1 per session, 7-day suppression | Usually off, test against holdout |
A testing sequence that actually finds the winner
Do not test five variables at once. Run this order, two to four weeks per test depending on traffic:
- Offer first. It moves results more than anything else. Discount vs content upgrade vs quiz.
- Headline and value proposition. Benefit-led and specific beats clever.
- Trigger thresholds. Tighten or loosen time and scroll by 50% and compare leads per 1,000 sessions, not capture rate.
- Format. Centre modal vs slide-in vs bottom bar, especially on mobile.
- Form fields and button copy. Smallest effect, test last.
Need statistical confidence? With a 3% capture rate you need roughly 10,000 to 20,000 impressions per variant to detect a meaningful lift. If your traffic cannot support that, prioritise obvious fixes from this guide over micro-testing.
Pre-launch checklist
- Triggers use time AND scroll, never a single condition
- Frequency caps and post-conversion suppression are configured and tested in a private window
- Checkout, cart, forms, account and legal pages are excluded
- Mobile uses a partial-screen format and a later trigger
- Close button is obvious, Escape key works, background click closes
- Offer is relevant to the page template it appears on
- Script is lazy-loaded and Core Web Vitals are unchanged
- Consent and data handling comply with GDPR and your privacy policy
- A holdout group is live and reporting is segmented by device
- Double opt-in or at minimum a confirmation step to protect list quality
FAQ
Can you give me an example of an exit intent popup?
A reader is 60% through an article about email deliverability and moves the cursor toward the browser tabs. A centred overlay appears: “Before you go, grab the 15-point deliverability checklist we use before every send.” One email field, one button, a visible X and a plain “No thanks” link. That is a textbook exit intent popup: contextual offer, engaged visitor, easy escape.
Do exit intent popups work on mobile?
Not in the same way. There is no mouse to track, so mobile tools infer exit intent from back-button presses, rapid upward scrolling or inactivity. These proxies are less accurate and more intrusive, and full-screen overlays on pages that receive search traffic can conflict with Google’s intrusive interstitial guidance. The safer approach is a bottom slide-in bar covering a small portion of the screen, triggered after real engagement.
Do exit intent popups hurt SEO?
The popup itself does not harm rankings. The risk comes from mobile interstitials that block content for users arriving from search, and from poor engagement signals created by over-triggering. Keep mobile overlays small and late, keep dismissal easy, lazy-load the script, and you avoid both issues.
What is a good exit intent popup conversion rate?
Most well-configured popups capture 2% to 5% of impressions. Highly relevant content upgrades on blog posts can reach 6% to 9%. Under 1% usually points to a weak or irrelevant offer rather than a design problem.
How often should an exit intent popup appear to the same visitor?
Once per session, once per 14 to 30 days after a dismissal, and never again after a conversion. If someone dismisses it twice, stop showing it entirely.
Should exit intent popups run on checkout pages?
By default, no. Interrupting an active checkout risks cancelling sales you had already won. If you want to test a cart-abandonment offer, run it against a holdout group and measure revenue per session, not just recovered carts.
How do I turn off a popup I keep seeing on a site?
As a visitor, close it with the X, the Escape key or a click outside the overlay. Clearing cookies and site data for that domain resets most frequency caps, and browser content blockers can suppress common popup scripts. As a site owner, remember that this friction is exactly what you create when your caps are too loose.
The takeaway
An exit intent popup is a tool for recovering visitors who are already gone, not a device for interrupting people who are still deciding. Arm it late, require real engagement, cap it hard, exclude the money pages, give mobile its own softer format, and always keep a holdout so you can prove the lift. Configure it that way and the popup becomes one of the cheapest acquisition channels you own.
Want your popup and content strategy audited end to end? King Content Agency builds and tests these systems for ecommerce and SaaS teams. Get in touch and we will review your current setup.