The fix isn't removing the pop-up.
A pop-up that fires the instant the page loads converts at roughly 1.4% - and it raises bounce by 8-12% while it's at it. The site-wide average, well-timed ones included, sits between 3.5% and 5% (Wisepops' 2026 benchmark of one billion displays), and the top 10% of campaigns clear 20%.
Across the 1,284 pages Levri has analysed, 74% of sites running pop-ups fire them on page load - before the visitor has read a line. The gap between 1.4% and 20% isn't luck. It's six specific mistakes, and every one is fixable in an afternoon without deleting a single pop-up. (If a pop-up is fighting your real call to action for attention, that's a different problem - see the three-CTA problem.)
The six patterns that move pop-up conversion.
- On-load triggers - the pop-up ambushes the visitor before they've read anything.
- No segmentation - the same pop-up to new, returning, paid, and engaged visitors alike.
- No frequency cap - the same pop-up every visit until people close it on reflex.
- Desktop pop-ups on mobile - a 600px modal squeezed onto a 375px screen.
- Generic "subscribe" copy - no value exchange, no reason to trade an email.
- No urgency or micro-step - the refinements the top 10% use and most sites skip.
#1 - On-load triggers that ambush the visitor.
A pop-up firing the instant the page loads is the digital equivalent of a shop assistant blocking the door. Immediate on-load pop-ups average 1.42% conversion - the lowest of any trigger - and they push bounce up 8-12%. You're not just failing to convert; you're driving people away.
What we see on scans
- The overlay appears within 0-2 seconds of load.
- No scroll-depth or time-on-page trigger.
- Bounce on pop-up pages running 9-14% higher than pages without.
What to ship
- Switch to a scroll-depth trigger at 50% - these convert at 3.6-3.9% with no bounce penalty.
- If timed, set the delay to 11-15 seconds - the window that peaks at 6.45%.
- For exit-intent, require 8-12 seconds of dwell first to filter immediate bouncers (that alone lifts exit-intent 28-35%).
Typical lift: +150% to +350% vs on-load. Impact: high - usually doubles captures overnight.
#2 - Same pop-up, every visitor.
Most sites show one pop-up to everyone: new visitors, returning customers, people deep in a blog post. Same headline, same offer, same irrelevance. Pop-ups targeting new visitors only convert at 8.30% versus 4.60% shown to everyone - an 80% lift from showing first-timers a different message.
What we see on scans
- Identical markup served regardless of new vs returning.
- No UTM or referral-source targeting.
- The same discount shown to customers who already bought.
What to ship
- Segment at minimum into new vs returning - a welcome offer for one, a review or loyalty prompt for the other.
- Match copy to page intent: cart incentives on product pages, content upgrades on blog pages, demo offers on pricing.
- Contextual offers convert at 3.8% vs 1.4% for generic "subscribe" - a 171% lift from relevance.
- Suppress pop-ups entirely for visitors who already converted on them.
Typical lift: +40% to +80%. Impact: high - segmentation is a setting most tools already support and most teams never configure.
#3 - No frequency capping.
Showing the same pop-up every visit creates "pop-up blindness" - people learn to close it on muscle memory, and conversion drops with every repeat. Across two million impressions: every visit captures 1.2% (with 45% closing without reading); once per 7 days hits 3.2%, the optimal window. That's a 167% lift from one setting.
What we see on scans
- The same cookie or trigger firing every session.
- No suppression window for dismissed pop-ups.
- Existing subscribers still seeing email-capture pop-ups.
What to ship
- Cap to once per 7 days for non-subscribers - the empirical sweet spot.
- Cap exit-intent to once per session, once per day for returning visitors.
- Permanently suppress capture pop-ups for subscribers; show downstream prompts instead.
- If you must re-expose within 30 days, rotate the offer - the same creative twice is worse than none.
Typical lift: +100% to +170% vs uncapped. Impact: high - under five minutes in any major tool.
#4 - Desktop pop-ups crammed onto a phone.
Mobile is now 50%+ of traffic, and mobile pop-ups designed properly convert 36% better than desktop (4.98% vs 3.67%). But "designed properly" carries the sentence - most sites shrink a 600px modal onto a 375px screen and call it responsive, producing tiny targets, accidental closes, and the risk of a Google interstitial penalty.
What we see on scans
- Pop-up width exceeds the mobile viewport, forcing scroll or clipping.
- Close button under 44x44px.
- The CTA in the upper half, outside the thumb zone.
- Multiple form fields on a mobile pop-up.
What to ship
- Design mobile pop-ups separately - full-screen with a visible 44x44px close, top-right.
- Move the primary CTA to the lower third where thumbs reach.
- One field on mobile - email only; collect the rest in the welcome sequence.
- Use bottom-bar or slide-in formats - they convert comparably without tripping Google's interstitial guidelines.
Typical lift: +20% to +36% on mobile. Impact: medium-high - scales with your mobile share, north of 50% for most.
#5 - Generic "subscribe" copy with no value exchange.
"Subscribe to our newsletter" is the worst-performing headline there is - 1.4% conversion. Swap in a specific, contextual offer and you hit 3.8%. People won't trade an email for vague "updates"; they need to see concrete value within three seconds. The form inside the pop-up matters too - the same field discipline from form field optimisation applies here.
What we see on scans
- Headlines built on "subscribe", "newsletter", "stay updated" with no offer.
- No discount amount, resource, or concrete benefit in the copy.
- A "Submit" or "Sign up" button - the two lowest-converting labels.
What to ship
- Lead with quantified value: "Get $20 off your first order" converts 62% better than no discount.
- For content sites, offer a specific resource - content-upgrade exit pop-ups convert far above generic capture.
- Replace "Submit" with action copy: "Get my free guide" beats "Submit" by ~48% (KlientBoost).
- Add an image - pop-ups with one convert 5.71% vs 3.98% without.
Typical lift: +60% to +170%. Impact: high - copy and offer changes are zero-cost and deploy in minutes.
#6 - No urgency, no micro-commitment.
Two features separate top-10% campaigns from average: countdown timers and multi-step flows. Timers lift conversion from 4.6% to 7.3% (a 58% increase); a single qualifying question before the offer lifts conversion to 5.17% from 4.62% and improves lead quality. Most sites use neither.
What we see on scans
- A discount pop-up with no expiry - the offer feels permanent, so there's no reason to act.
- A single-step form with no qualifying interaction.
- No personalisation by behaviour or page context.
What to ship
- Add a 24-hour countdown to discount pop-ups - real urgency, not fake.
- Use a micro-commitment first step: one yes/no question ("Looking for [category]?") before the form.
- For ecommerce exit-intent, lead with the strongest incentive for your vertical (discount codes ~37%, free shipping ~28%).
- Add a page-delay trigger - pop-ups after a two-page delay convert 9.84% vs 4.16% immediate.
Typical lift: +12% to +58%. Impact: medium - refinements, best applied after triggers, targeting, and copy.
How Levri spots all six in 60 seconds.
Paste a URL. Levri flags pop-up timing, mobile overlay problems, missing segmentation signals, and weak copy patterns, then ranks each by estimated revenue impact so you know what to fix first.
You paste the URL, and you get your fixes - a ranked list, each issue priced in $/mo, with a written hypothesis, a variant-B suggestion, and an expected lift range. No install, no tracking script, no "connect your analytics."
Fix these first.
Where 80% of the lift lives, in order:
- Move off on-load triggers to scroll-depth or a timed delay (biggest single lift).
- Add a 7-day frequency cap (five minutes, immediate compounding).
- Segment new vs returning (doubles relevance).
- Rewrite the headline with a specific, quantified offer (zero-cost).
- Design mobile pop-ups separately (captures the 36% mobile gap).
- Add countdown timers and a micro-step (refinement layer).
A pop-up should support the page's main action, not fight it (the three-CTA problem) and shouldn't bury your above-fold message (why your hero section is killing your CTR). Ship the first three and measure by Friday; if captures don't move, you're fixing the wrong settings, which is exactly what Levri is built to diagnose.