Website pop-ups: use them or not? What Google says and what we found on 28 UK online shops

A website pop-up is a window that appears over the page to ask the visitor to do something, usually leave an email address in exchange for a discount or a piece of content. It makes sense when the offer is worth the interruption and the window doesn't cover the page for people arriving from Google on a smartphone: since 10 January 2017 Google may rank exactly those pages lower, while it exempts banners that take up a reasonable part of the screen and windows required by law, such as cookie consent.

The real question isn't "pop-up or no pop-up".

It's which pop-up, for whom and at what moment.

Page one of Google UK for "website pop ups" is mostly galleries of examples published by pop-up software vendors (OptiMonk, Triggerbee, LeadsPanda), plus a Pinterest board and a 2023 piece from The Verge. They show what a pop-up can look like. We wanted to know two other things: what Google's rule says once you add the clarification Google itself published in 2023, and how many UK online shops actually load one.

The mistakes we see most often are three:

• the full-screen pop-up that fires as soon as the page opens, on top of the cookie banner;
• the promotional bar that pushes the content down and worsens CLS (the visual stability of the page, one of the Core Web Vitals);
• the same discount shown on every visit, even to people who have already signed up.

The pop-up sits exactly where rankings and conversions meet, and those are the two numbers we work on for clients.

The online shop of Scovaventi, an Italian food producer, had visits but wasn't selling: rebuilt on Shopify and pushed with campaigns, it closed twelve months with more than €77,000 in sales and an average order value of €190. With that order value, a 10% sign-up discount costs €19 an order. It's a decision to make with the numbers, not on a hunch.

Contesini Consulenze, on the other hand, sells electricity and gas consultancy in Italy: 299 enquiries at €7.20 each, with 68% turning into a sales opportunity. There, the work was done by a new landing page and proper tracking.

So do pop-ups work or not?

It depends on the site. Here you'll find:

• what Google actually says, in 2017, in 2023 and in today's documentation;
• the hidden cost to CLS, which hardly anyone factors in;
• our test on 28 UK online shops, with the figures by platform;
• how well they convert according to a vendor that measures more than a billion displays;
• the rule by type of site and the checklist for setting them up.

Let's start.

What a website pop-up is (and the forms it really takes)

A pop-up is an element that lays itself over the page without the visitor asking for it. Google splits them into interstitials, which cover the whole page, and dialogs, which cover only part of it: it isn't a textbook distinction, because it decides whether the window falls among the ones Google advises against.

On UK sites the recurring forms are six:

Form	What it covers	When it usually appears	Risk with Google on mobile
Full-screen modal	The whole page	On opening or after a few seconds	High, if it appears to people arriving from search
Interstitial on a separate page	It's a page of its own, before the content	On opening	High: Google asks you not to redirect to a separate page
Central modal	The central part, with the background dimmed	After a few seconds or after scrolling	Medium: it depends on how much of the phone screen it takes
Corner slide-in	A small box	After scrolling	Low
Top or bottom bar	A strip	On opening	Low for Google, but watch CLS
Exit intent	Variable	When the mouse leaves the window, on desktop	Low: it doesn't appear in the step from search to page

What about the cookie banner? That's a window too, but in another category: it has a section of its own further down.

Do pop-ups hurt SEO? (the 2017 rule and the 2023 clarification)

Yes, but in one specific case: when the window covers the main content for people arriving from a search result on mobile. Google announced it on 23 August 2016 and it has applied since 10 January 2017: pages where content isn't easily accessible on the step from mobile search "may not rank as highly".

In the same post Google listed what does NOT fall under the rule, if used responsibly:

• windows that respond to a legal obligation, such as cookie consent or age verification;
• login dialogs on sites with private or paywalled content;
• banners that use a reasonable amount of screen space.

And it wrote two things that usually get forgotten. The content under the pop-up stays present and indexable, so the problem isn't that Google can't read the page, but that it treats it as a worse experience. And that signal was "just one of hundreds", with the intent of the search still weighing more. It isn't a manual penalty: no message arrives in Search Console, unlike Google manual actions.

Then, in April 2023, came the clarification.

Google wrote that of the page experience signals (Core Web Vitals, mobile-friendliness, HTTPS and the absence of intrusive interstitials) "not all of these may be directly used to inform ranking". The page experience page, updated in September 2026, goes further: beyond Core Web Vitals, the other aspects "don't directly help your website rank higher".

So can I put up a full-screen pop-up and forget about it?

No. The current documentation on interstitials, updated in December 2025, still says "Don't obscure the entire page with interstitials" and adds that intrusive dialogs and interstitials "make it hard for Google and other search engines to understand your content".

Our reading: the direct risk to rankings has shrunk, the risk to how visitors behave hasn't. Someone who taps a result, finds a window covering everything and goes back is a lost visitor either way. It's one of the eight cases where SEO and UX really clash, and one of the points to check in mobile SEO.

The cost you don't see: CLS (and why a bar weighs more than a modal)

CLS (Cumulative Layout Shift) measures how much visible elements move while the page loads, and it's one of the three Core Web Vitals, the only page experience aspects Google says it uses in ranking. According to web.dev a good CLS is 0.1 or less; above 0.25 is poor.

Here's something counter-intuitive.

The central modal, the one everyone hates, usually doesn't move anything: it sits on top of the page, outside the document flow.

The promotional bar at the top, which looks harmless, often pushes all the content down when it appears. web.dev gives it as a typical example: banners and forms that pop in at the top or bottom and "shift the rest of the page's content". The fix it suggests is to reserve the space before the bar loads, or to lay it over the content.

It isn't a theoretical problem. When we ran Lighthouse on the nine pages on page one of Google UK for "ux seo" for our guide to UX and SEO, none went above 0.012. On the same test in Italy, the page in fifth place had a CLS of 0.353, three and a half times the threshold. If your mobile CLS is above 0.1 and you have a bar that appears after loading, start there.

How many UK online shops use pop-ups (our test on 28 pages)

Nearly two in three. On 10 October 2026 we downloaded the code of 28 pages from UK-facing online shops that appear in the top ten results of Google UK on a smartphone for four buying searches: 18 out of 28, 64%, load a tool or module capable of showing a sign-up pop-up.

The method, because a figure is only as good as the way it's collected:

• searches: "coffee beans online", "womens swimwear online", "natural face moisturiser buy online" and "designer table lamps online shop", collected with DataForSEO on Google UK, mobile;
• newspapers, blogs, videos and marketplaces left out: that leaves 31 shop pages (mostly category pages, plus a few home pages and one product page), 3 of which couldn't be read (two answered with a 403 error, one returned only an empty JavaScript shell);
• in the HTML, downloaded as if from a smartphone, we looked for the scripts of email capture tools (Klaviyo, Attentive, Mailchimp, Wisepops, Getsitecontrol, Drip, OptinMonster, Omnisend, Privy and others) and for the themes' own newsletter pop-up modules;
• the test sees what's in the code, not what appears on screen: a module can have display rules (delay, scroll, exit), or be installed and switched off.

Platform	Pages analysed	With a pop-up tool or module	Share
Shopify	18	17	94%
Other (WooCommerce, custom builds, enterprise platforms)	10	1	10%
Total	28	18	64%

Across the 18 shops, this is how the tools break down: Klaviyo on 11, all of them on Shopify, Attentive on 4, a theme module on 4, Mailchimp on 2, and Wisepops, Getsitecontrol and Drip on one each. Several shops load more than one.

Two details struck us.

The first: on Shopify the pop-up is practically the norm, 17 shops out of 18. Email marketing apps make it an option you switch on in a few clicks, and on five pages the code already contained the offer, such as "Get 10% off your first order" or "Sign up for 15% off". The other settings that matter on this platform are in our guide to Shopify SEO.

The second: on one Shopify theme the newsletter pop-up section is in the code but marked as hidden. Having the module doesn't mean showing it, and that's why our count stops at the code. The form Google leaves alone, a small dialog, is a setting on almost all of these tools. You just have to choose it.

And the other 10? They don't load any of the tools we looked for. That doesn't mean they don't collect email addresses: they can do it with a form in the footer or within the page, with no window at all.

How well pop-ups convert (the figures going round and the ones to use)

On average, 2.1% of people who see a pop-up leave their email address. That's the figure Omnisend published for 2025, calculated on 1.24 billion displays and 26.4 million emails collected by the shops using its platform: an international sample, made up mostly of ecommerce, and down from 2.3% in 2024.

The most useful figure in the same Omnisend report is about when the window appears:

Delay before it appears	Average conversion rate 2025
0-1 seconds	1.9%
1-5 seconds	2.3%
6-10 seconds	2.4%
11-15 seconds	2.1%
16-30 seconds	1.9%
31-60 seconds	1.8%
Over 60 seconds	1.9%

Three more numbers from the same report: pop-ups designed for mobile convert at 2.2%, desktop-only ones at 1.4%; those shown only on selected pages convert at 2.4% against 2.0% for those shown everywhere; the spin-to-win wheel reaches 3.5% or more.

So should I make it appear after 6-10 seconds on smartphones too?

Not if it covers the page. In its 2016 post Google lists among the risky techniques a pop-up that covers the content "while they are looking through the page", that is, while someone is reading. On mobile the delay changes the conversion rate, not the risk: the size is what changes the risk.

But didn't I read that pop-ups convert at 12%?

Yes, and we quoted it ourselves in our call to action examples: First Page Sage gives 12.1% for full-page pop-ups with an offer or a discount. They're different measures, on different bases, and the gap between them is almost six times. The practical conclusion is just one: don't decide on an industry average, decide on your own data, with the test you'll find further down. For the rest of the channel's numbers there's our collection of email marketing statistics.

On the other side of the scales is annoyance. In 2017 Nielsen Norman Group asked 452 US adults to rate 23 ad formats from 1 to 7: modals ended up among the most hated on both desktop and mobile, and the average annoyance on mobile (5.45) was higher than on desktop (5.09). It's an American sample and it's about ads, but the mechanism is the same: a window that blocks reading. The comments also showed that people tolerate more what they can close or ignore.

When to use a pop-up and when not to (the rule by type of site)

A pop-up pays off when it offers something the visitor wants straight away and that's worth more than the time you're making them lose. By type of site, this is the rule we apply.

Ecommerce: yes, with a real discount and not on the first tap

This is where Omnisend's numbers make sense. A discount on the first order or free delivery are concrete reasons to leave an email address. On an average order of €190, like Scovaventi's, 10% is worth €19 an order: a cost to build into your margin, not a gift. Show it after the visitor has seen at least one product and, on a smartphone, in a box that leaves the page visible.

B2B or lead generation site: almost always no on entry

Someone looking for a supplier or a quote has come to do exactly that. A pop-up asking for a newsletter sign-up competes with the contact form, which is the only conversion that counts. If you want a window, use it on exit on desktop for a downloadable piece of content, and measure whether it takes enquiries away.

And before that, fix the measurement. When we took over Valore Solare, the Google Ads account was counting 15,267 "conversions" in a month from 5,467 clicks: they were technical site events, not quote requests. A badly tracked pop-up has the same effect. The other reasons enquiries don't arrive are in our guide on why your website isn't generating leads.

Blogs and content sites: a side box after reading

Someone reading an article came for the article. A small slide-in that appears halfway down the page, with a resource tied to what they're reading, asks something of a person who has already shown interest, and covers nothing.

Local businesses and services: no

Someone looking for a plumber or a restaurant wants the phone number or the address. Every window before that information is one more step towards the next result on the page.

Pop-ups and cookie banners (what the ICO asks and why two windows are too many)

The cookie banner is a mandatory window, so Google exempts it. In the UK the rules are in PECR, and the ICO guidance on storage and access technologies, finalised in April 2026, says how to handle them: if someone ignores the banner and carries on browsing, you must not use the technologies that require consent, and if someone has refused, you shouldn't keep asking. As a general guideline the ICO suggests six months before requesting consent again.

The problem starts when the promotional pop-up lands straight on top of the banner. Someone arriving from Google on a smartphone finds two windows before a single line of content, and the second isn't exempt from anything. The ICO itself, in the same guidance, asks you to think about the effect of banners and pop-ups on the user experience.

The rule we use: the promotional pop-up only fires after the visitor has closed the cookie banner and seen at least one page. And if it collects email addresses it must say what they're for and link to the privacy notice, because Article 13 of the UK GDPR requires you to inform people at the moment you collect their data.

How to set up a pop-up that costs you neither rankings nor sales (the 8-point checklist)

1. Never full screen on mobile on the first page viewed. That's exactly the case covered by the 2017 rule.
2. A trigger tied to behaviour. Second page viewed, half the page read or, on desktop, exit: Omnisend sees immediate pop-ups (1.9%) convert less than those with a few seconds' wait.
3. On smartphones, a small box or a bar. The main content must stay readable.
4. No separate page before the content. Google asks you not to redirect the visitor to a separate page for consent or sign-up.
5. A close button that's visible and big enough for a thumb. Annoyance drops when visitors feel they can close it.
6. Reserved space or an overlay for bars, so CLS stays under 0.1.
7. No discount for people already signed up, and a frequency cap. Someone who closed the window shouldn't see it again on the next page.
8. Sign-ups measured as a separate event in GA4, apart from purchases and enquiries, so you don't repeat the inflated-conversions mistake.

How to tell whether the pop-up is making you money (the test almost nobody runs)

You can only tell by switching it off for some of your visitors. The sign-up rate tells you how many emails you collect, not how many purchases or enquiries you lose from people who close the page.

The test we recommend:

• control group: for a fixed share of visitors, say one in five, the pop-up doesn't appear, using the tool's A/B test or a rule in Google Tag Manager;
• metric: revenue or enquiries per visitor in the two groups, for at least four weeks;
• reading: if the group without the pop-up buys as much as the other or more, the emails you collect are costing you sales.

For traffic from Google, look in Search Console at the mobile clicks and position of the pages where you switched the pop-up on, four weeks before and four after. It isn't proof of cause and effect, but a drop on mobile only is a signal not to ignore. If, on the other hand, the problem is that the site doesn't sell even without windows, the starting point is a different one: why your website isn't converting.

To sum up

Full-screen pop-ups for people arriving from Google on mobile: to avoid since 2017.

Cookies, age verification and small banners: exempt.

CLS is spoilt by bars that push the page down, not by modals.

Almost two UK online shops in three have a pop-up ready in the code: 18 out of 28.

Average conversion: 2.1% across 1.24 billion displays.

The decision is made with a control group, not with the industry average.

The industry talks a lot about how to make pop-ups less annoying. It says little about the fact that a pop-up on entry is often a patch on a page that doesn't ask for anything on its own: if the product page or the landing page had a clear offer in the right place, the window would be needed less. Before choosing the plugin, ask what the page is missing. It's the question we start from when we work on clients' SEO and lead generation.

Until next time!

Frequently asked questions about pop-ups