Retargeting explained: how ads follow your visitors (and why it works)

Last updated: June 2026.

You browse a PrestaShop store, look at a pair of shoes, and leave without buying. For the next two weeks those exact shoes follow you around the internet, on Facebook, on Instagram, in the corner of a news site, before a YouTube video. That is retargeting, and the reason store owners keep coming back to it is simple arithmetic: the people it talks to have already raised their hand. They visited. They looked. They didn't buy yet. Reaching them again is far cheaper than finding a brand-new stranger and convincing them you exist.

This guide is about the strategy of retargeting on a PrestaShop store: why warm visitors behave differently from cold ones, how to slice them into audiences by intent, how often to show your ads before "helpful reminder" turns into "creepy stalker," and how to measure whether any of it actually paid off. It is deliberately not a pixel-installation walkthrough. The tracking plumbing that makes retargeting possible (Facebook Pixel, TikTok Pixel, GA4, and the rest) has its own home, and we point you there at each step rather than rehashing it here.

Why retargeting works when cold advertising struggles

Typical e-commerce conversion rates sit in the low single digits, most merchants see somewhere around 1–3% of visitors place an order, though your real number depends entirely on your traffic quality and niche. Flip that figure over and the point becomes obvious: the overwhelming majority of people who reach your store leave without buying. But "left without buying" is not the same as "not interested." They clicked through, they read a product page, some of them filled a cart. They are warm prospects who got interrupted, by a price they wanted to think about, a delivery question, a crying child, a closing browser tab.

Cold advertising spends money introducing your brand to people who have never heard of it. Retargeting spends money on people who already know you and already showed intent. That single difference is why retargeted audiences reliably convert at a higher rate and at a lower cost per acquisition than cold campaigns, the exact multiple varies wildly by store, product price, and creative, so treat any "10x" headline you read elsewhere as marketing, not a forecast. What you can rely on is the direction: warm beats cold, and it usually beats it by a margin big enough to matter.

How retargeting works (the short version)

Mechanically, retargeting rides on a tracking pixel, a small piece of JavaScript that fires when someone visits your store, drops a cookie in their browser, and reports their behaviour back to an ad platform. When that same person later browses another site in the platform's network, the platform recognises them and serves your ad. The visitor never has to create an account or hand over an email; the pixel and cookie do the work invisibly. With dynamic product ads, the platform goes further and rebuilds the ad around the specific products that person viewed, with live prices and stock.

That is the whole mechanism in a paragraph, because on PrestaShop the setup is mostly "install the right pixel and let it fire the standard events." The events that build your audiences, PageView, ViewContent, AddToCart, Purchase, are emitted by the pixel module, not by anything you configure by hand. If you want the actual installation, event-mapping, and consent-gating details, we cover them properly in social tracking explained simply and, for the most common case, in Facebook Pixel for PrestaShop. The rest of this article assumes the pixel is already firing and asks the more valuable question: now what do you do with it?

Choosing where to run your retargeting

Most major ad platforms support retargeting through a pixel/tag, uploaded audiences, or platform engagement audiences, and the right one depends on where your audience already spends attention, there's no single winner. A quick orientation, with the deeper channel decisions linked out so this post stays focused on strategy rather than turning into a platform tour:

If you're still deciding which social channels deserve any spend at all before you layer retargeting on top, start one level up with which platforms actually drive sales. Retargeting amplifies whatever channel you choose, so it pays to choose well first.

Segment your audiences by intent. Not everyone is one click from buying

The single biggest mistake in retargeting is treating "everyone who visited" as one audience and showing them all the same ad. A person who skimmed a category page and a person who abandoned a full cart are not in the same place, and spending the same budget and message on both wastes money on one and under-serves the other. Because the pixel already tags each behaviour, you can build distinct audiences in the ad platform and treat them differently. Ranked by how close they are to buying:

Cart abandoners, highest intent

They added products and stopped short of completing checkout. This is the audience that returns the most per euro, because they were one screen away from ordering. Show them the exact items left in the cart, and consider a small, honest nudge, free shipping over a threshold, or a modest code. Retargeting recovers some of these orders, but it works best alongside an on-site fix: if your checkout itself is the reason they bled out, no ad will paper over it. That side of the problem lives in your on-store conversion work, and an abandoned-cart email sequence catches the ones who never see the ad.

Product viewers, medium intent

They opened specific product pages but never added to cart. Something stopped them, price, a sizing doubt, a distraction. Show them the product they looked at plus a couple of close alternatives, in case the original wasn't quite right. Dynamic product ads handle this almost automatically once the ViewContent event is firing.

Category browsers, lower intent

They explored a category but didn't commit to any one product. They're still discovering. Don't push a hard sell; show best-sellers from the categories they browsed and let them find the thing they were circling.

Past purchasers, retention, not acquisition

Your warmest audience of all, and the one most stores neglect in their retargeting. Don't sell them what they already own, show new arrivals, complementary products, or a replenishment reminder timed to the product's lifespan. Retargeting existing customers is some of the cheapest revenue you'll find, because the trust is already built.

Frequency and duration: where retargeting earns its "creepy" reputation

The difference between a welcome reminder and brand damage is almost entirely about restraint. A few controls keep you on the right side of it:

• Cap the frequency. Limit how many times the same person sees your ad in a given week, where the platform allows explicit caps; otherwise manage frequency through budget, audience duration, exclusions and creative rotation. Past a certain point, extra impressions stop converting and start annoying, and an annoyed prospect is worse than a forgotten one. Watch your frequency metric and tighten the cap the moment click-through starts falling while impressions keep climbing.
• Set a sensible duration window. For everyday products, intent fades fast, a window of roughly one to two weeks usually captures the bulk of the recoverable conversions. For considered, expensive purchases (furniture, electronics), people genuinely deliberate longer, so a longer window earns its place. Don't keep chasing a visitor for two months over a €20 order.
• Exclude people who already bought. The fastest way to look incompetent is to keep advertising a product to the person who just purchased it. Use the Purchase event your pixel fires to build an exclusion audience and remove recent buyers automatically. It's obvious and stores still forget it constantly.
• Sequence the message instead of repeating it. Show the product first; if they still haven't bought after a few days, switch the creative to social proof (reviews, ratings); only later introduce a gentle incentive. Repeating one identical ad twenty times is how you train people to ignore you.

How much to spend

Retargeting is usually a minority slice of total ad budget, a small share of spend that pulls a disproportionate share of ad-driven revenue, precisely because it's working warm audiences rather than cold ones. The sensible way in is to start narrow: fund your highest-intent audience (cart abandoners) first, prove it returns more than it costs, then widen to product viewers and beyond. A modest daily figure is enough to reach a warm audience on a small store; the goal early on isn't scale, it's confirming the channel pays before you pour money into it.

Privacy and consent, non-negotiable on PrestaShop

Retargeting runs on cookies, and in the EU cookies for marketing require explicit consent before they're set. On PrestaShop that means your tracking pixels must not fire until the visitor has accepted marketing cookies in your consent banner, firing a Meta or TikTok pixel on page load, before consent, is a GDPR problem, not a growth tactic. The MPR pixel integrations are built to respect a consent gate rather than fire blindly, which keeps the legal exposure off your desk while still capturing the audiences you're allowed to capture. The broader shift toward privacy, Apple's App Tracking Transparency, browsers phasing out third-party cookies. Has trimmed retargeting's reach, which is exactly why first-party data and server-side tracking are becoming the way to keep it working.

Measuring whether it actually worked

Retargeting is unusually easy to fool yourself about, because the people in the audience were likely to come back anyway, so a naive "look, they converted!" overstates the credit. Track the numbers that tell you the truth:

• Return on ad spend (ROAS), revenue per euro spent, watched per audience rather than as one blended figure.
• Conversion rate by audience. Cart abandoners should out-convert product viewers, who should out-convert category browsers. If that ladder is inverted, your segmentation or your creative is wrong.
• Frequency vs. click-through. Rising frequency with falling clicks is the early warning of ad fatigue; cut the cap before it costs you.
• View-through conversions, people who saw the ad and bought later without clicking it. This is the reminder effect, and ignoring it under-credits retargeting; over-weighting it over-credits it. Hold it in mind, don't build your whole case on it.

And give it room to breathe before you judge. A few days of data is noise. Let a campaign run long enough to clear the natural week-to-week swings before you declare it a winner or a dud.

Setting it up on your PrestaShop store

Pulled together, the path from zero to a working retargeting setup on PrestaShop is short, because most of it is the pixel doing its job:

• Install the relevant pixel module for the platform you chose, Facebook, TikTok, Pinterest, Snapchat, Reddit, Quora, AdRoll, Outbrain, Taboola or X, or GA4 / Google Tag Manager for the Google ecosystem. Each MPR pixel module installs and configures from the back office, fires the standard e-commerce events automatically, and respects your cookie-consent gate, so there's no theme editing and no developer invoice to get the data flowing.
• Confirm the events are firing. Page views, product views, add-to-cart and purchase. These are what populate your audiences; if they're silent, nothing downstream works.
• Build the intent audiences in the ad platform, cart abandoners, product viewers, category browsers, past purchasers, from that event data.
• Write a campaign per audience with messaging matched to where each one sits, not one ad blasted at everyone.
• Set frequency caps, duration windows and a buyer-exclusion before you turn it on, not after the complaints arrive.

Common questions about retargeting on PrestaShop

Do I have to run ads myself, or does the module do the retargeting? The module doesn't run ads, it fires the tracking pixel and the standard e-commerce events (PageView, ViewContent, AddToCart, Purchase) that build your audiences. The actual campaigns, budgets and creative live in the ad platform (Meta, Google, TikTok). The MPR pixel's job is getting clean event data flowing; what you do with that audience is the strategy half of this article.

Can I fire the Facebook or TikTok pixel on page load to capture everyone? Not in the EU. Marketing cookies require explicit consent before they're set, so a pixel that fires on page load ahead of consent is a GDPR problem, not a growth tactic. The MPR pixel integrations are built to wait for the visitor to accept marketing cookies in your consent banner, you capture the audiences you're allowed to capture, and the legal exposure stays off your desk.

Which audience should I fund first? Cart abandoners. They were one screen from ordering and return the most per euro. Start narrow: prove that audience pays, then widen to product viewers, category browsers and past purchasers. Don't blast one ad at "everyone who visited"; that wastes budget on the low-intent half and under-serves the high-intent half.

How do I stop retargeting from feeling creepy? Restraint, mostly. Cap how often the same person sees your ad in a week (or manage frequency through budget and audience duration where explicit caps aren't available), set a sensible window, roughly one to two weeks for everyday products, longer only for considered purchases, exclude people who already bought using the Purchase event, and sequence the creative instead of repeating one identical ad twenty times.

Will third-party cookie changes kill retargeting? They've trimmed its reach, Apple's App Tracking Transparency and browsers phasing out third-party cookies both cut into browser-based retargeting. That's precisely why first-party data and server-side tracking are becoming the way to keep it working, rather than a reason to abandon the channel.

Retargeting isn't magic and it isn't a growth hack, it's the unglamorous discipline of reaching people who already wanted what you sell, at the moment they're most likely to come back, without wearing out your welcome. Get the pixel firing cleanly, segment by intent, respect the frequency line, and measure honestly. Done that way, it runs quietly in the background recovering orders that would otherwise have walked out the door for good.