How to use Exit Intent and Bounce Recovery
A practical guide to exit-intent overlays, scroll-triggered offers, and second-chance CTAs — when they recover revenue, when they damage brand trust, and the design rules that separate the two.
Exit Intent and Bounce Recovery
Interventions that fire when a visitor signals they're about to leave — overlays, scroll-triggered offers, and second-chance CTAs designed to recover the session.
Exit intent and bounce recovery covers the set of on-site interventions that trigger when a visitor's behaviour predicts they're about to leave without converting. The classic trigger is desktop cursor velocity heading toward the browser chrome; on mobile it's scroll-up velocity, idle time, or back-button intent. The intervention itself is usually an overlay — a discount, an email capture, a reassurance about shipping or returns, or a second-chance CTA pointing to a higher-intent page.
Done well, exit intent recovers 2-5% of otherwise-lost sessions and feeds email flows that close the loop later. Done badly, it trains visitors to associate your brand with interruption and erodes the trust that drove them to your store in the first place.
Most teams treat exit intent as a single lever — install the app, set the discount, watch the email list grow. That framing hides the real decision, which is when to intervene at all. A first-time visitor browsing a hero collection is a different problem from a returning visitor staring at a half-filled cart, and the same overlay will help one and annoy the other.
This guide unpacks the mechanics underneath exit-intent tactics, the conversion ranges you can realistically expect, and the design rules that keep recovery overlays from cannibalising the brand experience. It sits inside the broader bounce-rate-reduction toolkit, but focuses specifically on the moment of departure rather than upstream page-quality fixes.
When exit-intent overlays actually help
The clearest win for exit intent is on high-consideration product pages and mid-funnel carts. A visitor who scrolled, opened the size guide, and lingered for 90 seconds has already signalled interest — an exit overlay offering free shipping or a 10% first-order code converts somewhere between 3% and 8% of those sessions in apparel and beauty.
The second strong use case is cart and checkout abandonment. Here the visitor has revealed intent through behaviour, and the cost of doing nothing is a fully-lost transaction. A reassurance overlay — return policy, customer reviews, a chat handoff — often outperforms a discount, because the friction wasn't price.
The weakest use case is the homepage of a first-time visitor. They haven't formed a preference yet, so the overlay reads as a sales pitch from a stranger. List-growth overlays still work here, but treat them as email capture economics, not conversion recovery — the value lands in the Klaviyo flow days later.
The repeat-visitor trap
If your overlay fires on every session, returning customers see it 4-6 times before they buy. Cap frequency at once per 14 days per visitor, and suppress entirely for logged-in customers and anyone with an order in the last 60 days. The lifetime-value cost of annoying buyers dwarfs the marginal email captures.
How the triggers actually work
Desktop exit intent reads cursor velocity and trajectory. When the pointer moves upward past a threshold speed toward the top edge of the viewport, the script fires. It's noisy — users reach for bookmarks, tabs, and extensions constantly — so most implementations add a dwell-time minimum (typically 5-10 seconds on page) and a scroll-depth gate (often 25%) before arming the trigger.
Mobile has no cursor, so the proxies are different: rapid upward scroll near the top of the page, back-button press intercepted at history level, or a hard idle timeout of 30-60 seconds with the tab still focused. Mobile exit-intent fires less reliably than desktop, which is why most stores see 60-70% of overlay impressions come from desktop sessions despite mobile dominating traffic.
Exit-overlay conversion rate by trigger context
The pattern above repeats across verticals: the deeper the visitor is in the funnel when intent-to-exit fires, the higher the recovery rate. This is partly selection — they've already qualified themselves — and partly stakes, because cart and checkout overlays interrupt a near-purchase, not a casual browse.
What good looks like in the numbers
Benchmarks vary widely with vertical, average order value, and how aggressive the offer is. A 20% discount converts harder than a 5% one but eats margin and trains repeat buyers to wait for the popup. The table below shows realistic ranges by platform and offer type for stores in the €1M-€15M revenue band.
Two numbers matter beyond the headline conversion rate: the incremental lift over no-overlay baseline (subtract organic email signups and returning-visitor purchases that would have happened anyway) and the downstream attributable revenue from captured emails over the next 30 days.
Exit-overlay performance ranges by offer type and platform
| Offer type | Shopify apparel | Shopify beauty | WooCommerce home |
|---|---|---|---|
| 10% first-order discount | 3.8 - 6.2% | 4.5 - 7.8% | 2.9 - 5.1% |
| Free shipping threshold | 2.4 - 4.0% | 3.1 - 5.4% | 2.0 - 3.6% |
| Email-only capture (no offer) | 1.6 - 2.8% | 1.9 - 3.2% | 1.2 - 2.4% |
| Reassurance / social proof | 2.1 - 3.5% | 2.4 - 4.1% | 1.8 - 3.0% |
| Quiz or product finder | 4.2 - 7.0% | 5.8 - 9.4% | 3.0 - 5.5% |
Quiz overlays consistently top the table because they convert the exit moment into engagement instead of a transaction ask. A beauty store offering a 60-second skin quiz routinely captures 6-9% of exiting sessions and feeds richer first-party data into segmentation than a flat email field ever would.
Design rules that protect brand trust
The fastest way to wreck an overlay is to make it hard to dismiss. A close button under 32px, hidden in low-contrast grey, or delayed for three seconds reads as a dark pattern and costs more in brand trust than the marginal capture is worth. Make the close affordance obvious, instant, and as prominent as the primary CTA.
Match the offer to the context. A first-time visitor on a €40 hoodie page doesn't need a 25% discount — 10% with free shipping over €60 converts almost as well and protects margin. Reserve aggressive discounts for high-AOV carts where the recovered transaction justifies the hit, and never stack overlay discounts on top of automatic site-wide promotions.
Three rules that separate good overlays from bad
1) Frequency cap: once per 14 days per visitor, suppressed for recent buyers. 2) Mobile parity: never show desktop-only overlays that break the mobile layout — design native mobile variants. 3) Real value: an overlay that promises a discount, captures the email, and never sends the code destroys trust permanently. Wire it to your ESP before launch and test the full flow.
Frequently asked questions
Exit overlays themselves aren't penalised by Google because they trigger after the page has loaded and the user is leaving. Entry interstitials are a different story — those can trigger the intrusive interstitial penalty on mobile. Keep first-load experiences clean and reserve overlays for the exit moment.
A standard email capture fires on a timer or scroll depth regardless of intent. An exit-intent overlay fires only when behaviour signals departure — cursor velocity on desktop, idle or back-button on mobile. Exit-intent generally has higher conversion per impression but fewer impressions overall.
No. Mobile triggers are less reliable and screen real estate is tighter, so mobile overlays should be shorter, single-field, and use a bottom-sheet pattern rather than a modal. Many stores see better mobile results from a sticky bar than a full overlay.
Cap at once every 14 days per visitor as a default, and suppress entirely for visitors with an order in the last 60 days or who are logged in. Frequency caps live in your overlay tool's session storage or your customer data platform.
A well-built overlay script adds 15-40KB and runs after page interactive, so Core Web Vitals impact is minimal. The risk is stacking three tools — heatmap, A/B test, popup — each with their own bundle. Consolidating to a single script that handles all three is usually a 200-400ms LCP win.
Exit overlays are one tactic inside the broader bounce rate reduction toolkit. Bounce reduction also covers page speed, content matching, internal linking, and CTA clarity — fixes that prevent the bounce upstream. Overlays only catch what's already escaping.
Start at 10% for first-order overlays in apparel and beauty, 5-7% for higher-margin verticals like supplements, and free shipping over a threshold for low-AOV stores. Test downward, not upward — it's easier to defend a smaller discount than to claw back trained behaviour.
Yes, and you should. Run the overlay against a holdout that sees nothing, measure incremental conversion and 30-day attributable revenue, not just overlay opt-in rate. Many overlays look great on opt-in metrics and break even on actual incremental revenue.
Across all contexts, 2-5% of triggered impressions is the typical band. Cart and checkout exit overlays push 8-12%, while homepage first-visit overlays sit closer to 1-2%. Anything claiming a 20%+ overlay conversion rate is measuring opt-ins on a tiny qualified subset, not real recovery.
Yes, they complement each other. The overlay captures the email so the abandoned-cart flow has something to send to, and recovers visitors who weren't logged in. Without the overlay, anonymous cart abandoners simply leave with no follow-up channel.
Get an AI expert review of your site
Paste your URL — Metricuno's AI runs the same heuristic checks a senior CRO consultant would, scoring your page and prioritising the fixes that'll move conversion fastest.