For most small online stores, one discount-bearing exit popup is enough: shown only to first-time, non-subscribed visitors, on desktop exit with a scroll trigger on mobile, offering 10% or less. Other pages get non-discount prompts, and dedicated testing tools pay off around 10,000 to 15,000 monthly visitors.
A popup that wins the signup by giving away more margin than the customer was ever going to cost you is not a list growth tactic. It is a discount program with an email field attached.
Open ten small Shopify stores in a row and you will see the same thing on most of them: a popup asking for an email in exchange for 10 or 15% off, firing on a timer seconds after you land, and often firing again when you move to leave. Some stores run two or three of these at once from two or three different apps. The exit popup has become a reflex, and reflexes are rarely tuned.
That matters more now than it did a few years ago. Paid traffic costs more, a growing share of visitors arrive on phones where classic exit intent cannot fire, and every discount you hand out at the door comes straight off the margin of a first order that is already the hardest one to make profitable. The question is no longer which of thirty popup designs looks best. It is who should see one, what it should cost you, and whether your store has enough traffic to justify a dedicated tool at all.
This guide is written for stores from the first sale up to roughly $1M a year. If you are larger, the principles hold, but your answer is a testing program, and I will point to where that starts.
An exit intent popup fires when a desktop visitor moves the cursor toward the browser’s close button, tab bar or address bar, and it cannot fire that way on a phone because touch screens have no cursor. That single technical fact changes how you should set one up: check your own analytics, and if most of your sessions are on phones, a desktop-only exit popup is invisible to most of your visitors.
Shopify’s own theme guidance on removing intrusive interstitials is explicit on both points. It notes that exit intent triggers only work with a pointer, so touch sessions need a scroll trigger instead, and it advises against adding a timer as the touch fallback. A Klaviyo employee said the same thing in the company’s community forum back in 2021: traditional exit popups don’t work on touch devices, and the practical mobile option is a “show after scrolling X% of the page” rule.
The second reason to care is search. Google’s guidance on avoiding intrusive interstitials and dialogs says overlays that cover the page make content harder for search engines to understand and recommends small banners instead. A popup that only appears when someone is leaving, or after they have scrolled and engaged, does not block the content they came for. A full screen overlay three seconds after landing does.
So the working definition for a small store is this: on desktop, exit intent means cursor movement toward leaving; on mobile, it means a scroll depth trigger set deep enough (60 to 80% of a product page is a sensible starting point, not a benchmark) that the visitor has seen the product first. If your current setup uses a timer on mobile, that is the first thing to change, and it costs nothing.
A bigger discount usually wins fewer profitable customers than a smaller one, because steeper offers attract shoppers who came for the code rather than the product. The best data I have on this comes from episode 492 of the podcast, where Taras Talimonchuk, CMO of Claspo, shared what his company sees across billions of popup views.
His numbers, in his words: “So we have, we see that we have a 7% conversion for offers like from 1% to 10% and lower conversion 6% for offers from 11% to 15%.” That is the headline, and it is counterintuitive enough that most merchants never test it. The condition matters just as much. Taras also said conversion jumps to about 17% for offers in the 16 to 20% range, more than double the lower band, and he attributed part of that to gamified popups in the same dataset, where a wheel advertises prizes of up to 50% even though the average win is far lower. When Claspo looked past signups to revenue, the 1 to 10% band did better than the higher discounts. His summary was blunt: “So higher discount means that you got better conversion, but you got discount hunters.”
Two caveats keep this honest. First, this is one vendor’s aggregate data, not a controlled study of your store. Taras himself said that beyond the numbers, “all I have about this interpretation is guessing because I don’t know exactly why it is so.” Second, he did not argue against discounts. “We are not saying the 10% off pop-up is bad. It is not bad. It is a good tactic, but not for all traffic, not for all audience.”
What I take from it is a default and a test. The default for a small store is 10% or less, or an offer that costs nothing, such as free shipping above your current average order value. The test, once you have the traffic, is whether a smaller offer holds your signup rate while lifting revenue per visitor. If you do use a discount, issue it as a unique, single use code; our guide on when unique Shopify discount codes protect profit covers why generic welcome codes leak onto coupon sites.
Your exit popup should never show to existing subscribers, past customers, visitors arriving from your own emails, or anyone already in checkout. Each of those groups either has your offer already or was going to buy without it, so a popup to them is a discount you did not need to give.
This is the most common failure I see, and it is not limited to small stores. On episode 492 I described shopping at stores where I was logged in, had bought only men’s clothing, and still saw personalization that ignored what I had already bought. Taras’s point in reply was that every popup app offers targeting for new versus returning visitors, but identifying a returning visitor reliably is harder than it looks, and that email platforms, which he called frenemies rather than competitors, do it less thoroughly than dedicated popup apps. Krisztian Kiraly of OptiMonk made a related point on episode 469: traffic source alone should change what a visitor sees, and UTM parameters on your own email links tell you a lot about who came from your list.
In practice, the exclusion list for a small store has four lines. Exclude known subscribers (most email platform forms do this by default when the visitor is identified). Exclude visitors whose landing URL carries your email or SMS UTM tags. Exclude cart and checkout pages from discount offers, because a shopper at checkout who sees a new code will often leave to find a better one. And exclude anyone who has already dismissed the popup, for at least 30 days.
That last rule sounds minor, but repeated popups are what turn a single reasonable ask into the experience shoppers complain about. One appearance per visitor per month is a sane ceiling for a store at this stage.
The exit popup ideas worth testing are the ones that answer the specific reason a visitor is leaving, not the ones with the best design. A gallery of thirty examples is useful for layout inspiration, but the offer is what decides whether the popup earns its keep, and the offer should follow the page the visitor is on.
Notice that four of the six cost you no margin at all. That is deliberate. Small stores tend to reach for the discount first because it is the template every app ships with, but a visitor who leaves because shipping looked expensive is better served by a free shipping reminder than by 15% off, and a visitor reading a guide has not decided to shop yet.
Pick one idea per page type, not six at once. A store with a product page popup, a cart page popup and a blog popup has three tests to watch, which is already plenty for one person. If you want a process for deciding which of those tests runs first, our piece on why the “we should test that” list never gets tested lays out owners, decision dates and kill rules.
On eCommerce Fastlane, exit intent popups converted better than timed and scroll popups, which is evidence that exit timing works and not evidence that it will work the same on a store. eCommerce Fastlane is a media business rather than a store, but we run our own popups and exit intents for the newsletter and partner offers, on OptiMonk.
OptiMonk published a case study of our setup in September 2024 covering six months. Timed and scroll triggered popups (after 10 seconds or 15% scroll) converted at 6.02%, and exit intent popups shown on selected articles converted at 8.48%. Those are real numbers, and they point in the same direction as the advice above: asking at the moment someone is leaving beat asking them shortly after they arrived.
Here is what I would not let anyone conclude from them. The case study counts a conversion as a click on the popup, not a confirmed email or a sale, so the true signup rate is lower and unknown from the published figure. The exit popups ran on selected topic pages, so they were already more targeted than the site wide timed popup, which flatters the comparison. The timed popup also carried a discount offer, so offer and timing changed together. And our visitors are readers, closer to the “reading, not shopping” row in the table than to someone on a product page.
The call I would make from our own data is narrow: if you run one popup only, make it the exit popup rather than the timed one, and measure it on confirmed signups and the revenue those signups produce in their first 30 days, not on clicks. That is the change I would make to our own reporting too.
For transparency: Fastlane uses OptiMonk, both OptiMonk and Claspo have had team members on the podcast, and both offered listener discounts on their episodes.
For a typical small store running several overlapping popups, the right move is to remove all but one, move it to exit timing, and replace the biggest discount with a smaller or free offer. To show how that call gets made, here is a composite store built from patterns I have seen repeatedly, not a real merchant, and the figures are illustrative.
The composite: a candle and home fragrance brand doing about $25,000 a month, around 8,000 monthly visitors, roughly 70% on mobile. It runs a timed welcome popup offering 15% off at five seconds, a spin to win wheel from a second app, and an exit popup from its email platform that repeats the 15% offer. Average order value sits a little under its $60 free shipping threshold.
What I would look at first is overlap: three popups competing for one email address, two of them offering the same discount, and a five second timer firing on phones before anyone has seen a product. Then the offer: 15% off an order just under the free shipping line is the most expensive way to answer a shipping objection.
The call. Switch off the timed popup and remove the spin to win app entirely. Keep one popup in the email platform: exit intent on desktop, 70% scroll on mobile product pages, first-time non-subscribed visitors only, offering 10% as a unique code. On the cart page, replace the discount with a free shipping progress message. On blog pages, offer the scent guide instead of a code. Then wait 30 days and compare confirmed signups and first order margin against the previous 30. At 8,000 visitors a month that comparison will not reach statistical certainty, so treat the new setup as a margin protecting default you keep unless the numbers clearly turn against it, not as a proven winner.
Where you could reasonably disagree: removing the wheel may cost some signups in the short term, because gamified popups do capture more emails. Claspo’s Taras also argued on episode 492 that a game leaves a better first memory of the brand than a bare discount, and that its script loads asynchronously without hurting Core Web Vitals; those are vendor claims, but fair ones to weigh. At 8,000 visitors a month I would still remove the wheel, because the store cannot yet tell which signups become profitable customers, and a second app is one more tool to pay for and maintain. At 20,000 or more visitors a month with a clean measurement setup, I would bring a gamified variant back as a proper A/B test.
A dedicated popup and testing tool earns its cost once a store has enough traffic to run real tests, which in my view starts around 10,000 to 15,000 monthly visitors; below that, the forms built into Klaviyo or Omnisend are usually enough. Native forms already handle exit triggers, scroll triggers and subscriber exclusions, and they keep every signup in the same system that sends your welcome flow.
Krisztian Kiraly gave the clearest threshold I have heard on episode 469. For OptiMonk’s AI driven optimization, he said a store needs “at least 10 or 15,000 unique visitors per month, this is I would say the minimum level where it makes sense to switch on the AI powered CRO tools,” and that below that level classic popup tools are still worth trying for subscriber capture and cart abandonment. Both vendors offer free entry tiers, so the real constraint for a small store is testing power and attention rather than price. Taras described a different kind of ceiling on episode 492: a marketer who notices they are stuck at around 3% opt-in on native forms, and wants gamification, segmentation and personalization their email platform cannot do, is the one who goes looking for a dedicated app.
The other side of this decision is removal. During my years at Shopify, one of the first things I did when I took on a new brand was an app audit, and forgotten apps were a regular finding: tools nobody on the current team could explain, still billing and still adding scripts to every page load. If your store has more than one popup or form app installed, the audit is the first step, before any new tool. Our roundup of Shopify Black Friday apps and when to skip each one makes the same argument for the busy season: check that forms and popups are not hiding the add to cart button or slowing the mobile page before BFCM traffic arrives.
If you do move to a dedicated tool, judge it on the same measure as the composite store above: confirmed signups that turn into first orders at a healthy margin, not popup clicks. A tool that raises signups by handing out bigger codes has not grown your list. It has moved margin into your email platform.
Exit intent popups do not work on mobile in the classic sense, because they rely on cursor movement and phones have no cursor. Shopify’s theme guidance says touch sessions should use a scroll trigger instead, and advises against using a timer as the mobile fallback. A practical setup is exit intent on desktop and a scroll depth trigger of around 60 to 80% on mobile product pages, so the visitor has seen the product before the ask appears. Preview the form on a real phone after you set it up, so you know the mobile version actually appears.
Offer 10% or less, or a non-discount reason to stay, as your default exit intent offer. On episode 492 of the eCommerce Fastlane podcast, Claspo’s CMO shared aggregate data showing roughly 7% conversion for offers of 1 to 10% versus 6% for 11 to 15%, with the lower band also producing better revenue per visitor. Offers above 15% converted higher in that data, partly because gamified wheels advertising large prizes were included, and they tended to attract discount hunters. Use a unique, single use code so the offer cannot spread to coupon sites, and test a smaller offer before assuming a bigger one is needed.
Exclude existing subscribers, past customers, visitors arriving from your own email or SMS links, and anyone on the cart or checkout pages from discount based exit popups. These visitors either already have your offer or were likely to buy without it, so showing them a new code gives away margin for nothing. Most email platform forms can hide from identified subscribers, and UTM parameters on your own campaign links let you exclude list traffic. Also suppress the popup for at least 30 days after someone dismisses it, so a single reasonable ask does not become a repeated interruption.
Klaviyo or Omnisend forms are enough for most stores under roughly 10,000 to 15,000 monthly visitors, the traffic level where real A/B testing starts to become practical. They support exit and scroll triggers, can exclude known subscribers, and keep every signup inside the platform that sends your welcome emails. A dedicated popup app such as OptiMonk or Claspo starts to earn its cost when you have the traffic to run meaningful A/B tests and want finer targeting, gamification or product page personalization. Before adding any popup app, audit what is already installed, because stores often run two or three overlapping popup tools that slow pages and compete for the same email address.
Measure an exit intent popup on confirmed signups and on the revenue and margin those signups generate in their first 30 days, not on popup clicks or views. Click rates flatter popups, because a click is not an email address and an email address is not a customer. Compare a 30 day period before and after any change, keep the welcome email flow the same while you test, and track how many signups redeem the offer and at what order value. If signups rise but first order margin falls, the popup is buying subscribers with discounts rather than growing a list that buys.
