What Is Ad Retargeting and How Does It Work?

In an ideal world, every visitor to an online store would complete a purchase, sign up for a newsletter, or otherwise convert. In practice, that rarely happens. Converting prospects into paying customers is a long, non-linear journey — and retargeting is one of the more practical tools for staying present throughout it.

Limited attention spans, banner blindness, ad overload — these are persistent features of the modern e-commerce landscape, making effective ad targeting increasingly difficult. Getting a user to land on a website is only half the battle. Until a sale is made, even a carefully constructed acquisition strategy can fall apart. Online shoppers are unpredictable: they leave for any number of reasons — a distraction, second thoughts, a price comparison, a temporary budget constraint — and once the browser tab closes, they're gone.

There's no reliable way to prevent users from leaving. There are, however, ways to re-engage them. That's the core premise of ad retargeting.

What Is Retargeting?

Retargeting (called remarketing in Google's nomenclature) is a technique for showing ads to users who have previously visited a website without completing a purchase or other desired action.

Most visitors to an online store aren't there to buy immediately. They're researching, comparing features, checking prices against competitors. The purchase decision forms gradually, and it often finalizes somewhere other than the original site. Retargeting gives advertisers a mechanism to be present at that moment of decision — wherever the user ends up across the web.

How Does Retargeting Work?

Technically, retargeting involves identifying anonymous visitors who have previously visited a website and showing them ads for the products or services they viewed. This works across multiple channels: display advertising, social media, and email.

To identify users across the web and serve them retargeted ads, a third-party cookie from the retargeting service must be stored in their browser. It's called third-party because it's stored under the retargeting service's domain, not the advertiser's.

Here's a video by Isaac Rudansky explaining the basics:

Retargeting is implemented by adding a JavaScript code snippet or a pixel — literally a 1×1 transparent image — to the website's pages, typically in the footer. The pixel is invisible to the user. Its purpose is to trigger a request to the retargeting service's server when the page loads, which then drops a cookie in the user's browser. Since every pixel contains a unique identifier, the cookie allows the retargeting service to recognize that user on other sites.

Why a Pixel?

The simplest way to implement retargeting is to load a file from the retargeting service's domain when a page loads — which triggers the server request needed to set a cookie. Since the service has no reason to display an actual image, it uses the smallest possible image: a 1×1 transparent pixel that is never visible to the user.

This is why a retargeting JavaScript tag is commonly called a retargeting pixel — the mechanism and the terminology are interchangeable in practice.

Here is a visual representation of how the retargeting process works:

Step-by-step breakdown of the process above:

• A visitor lands on shoeholics.com and views a pair of shoes.
• The retargeting service's code, placed within the <footer></footer> tags, sends a request for a 1×1 pixel.
• The retargeting service returns the pixel and assigns a cookie to the user under its own domain (e.g., ads.retargetser.com), storing information about the visitor's behaviour — including the product they viewed.
• The visitor leaves shoeholics.com, navigates to news.com, and sees an ad for the exact same pair of shoes.

How did the retargeting service recognize the visitor on news.com as the same one from shoeholics.com?

It comes down to the third-party cookie assigned on shoeholics.com. The ad displayed on news.com is loaded from the retargeting service's domain, and the cookie is included in the ad request — so the retargeting service recognizes the visitor.

In a real-world implementation, that retargeting service would be an AdTech platform — most commonly a demand-side platform (DSP). Once the DSP has assigned a cookie to a visitor, it can bid on ad impressions in an ad exchange that the same visitor would see on other websites.

However, there's a complication: an ad exchange can only read cookies it created itself. So for the DSP to identify that visitor within the exchange, the two platforms need to swap and match their respective cookies through a process called cookie syncing.

Cookie syncing allows the DSP and ad exchange to exchange and reconcile their cookies, enabling both platforms to recognize the same online visitor as they move across websites. In effect, it allows the platforms to share behavioural data about visitors to power accurate targeting.

Benefits of Retargeting

Speed and automation. Once set up, retargeting runs automatically with no ongoing human intervention. Visitors can start seeing retargeted ads as soon as they leave a site — though the timing is configurable.

Targeting efficiency. Retargeting focuses spend on users who have already demonstrated some level of interest in a product or service — either implicitly (by browsing) or directly (by adding to a cart). That makes it more cost-effective than broad, untargeted display advertising.

The numbers support it. Evidence consistently shows that retargeted ads outperform standard display ads on measurable dimensions:

• An estimated two-thirds of surveyed users reported noticing a retargeted ad specifically because they had previously researched that product. Standard, non-retargeted ads tend to go unnoticed — a consequence of banner blindness.
• Around 90 percent of surveyed users reported a positive or neutral reaction to retargeted ads. The higher relevance of retargeted ads — compared to spray-and-pray formats with no targeting parameters — explains much of that goodwill.
• The average click-through rate for retargeted ads sits at approximately 0.7 percent, compared to roughly 0.07 percent for standard display — roughly ten times higher. Dynamic retargeting ads, which serve personalized creative based on the specific products viewed, can push CTR even higher.
• Retargeting has been shown to generate a 1,046 percent lift in search behaviour compared to other targeting strategies.

Challenges of Retargeting

Frequency and perception. Retargeting is effective, but overuse undermines it quickly. Bombarding a user with the same ad for weeks because they visited a site once can register as intrusive — or even as stalking. Frequency caps and thoughtful campaign design are essential to avoid burning goodwill.

Cookie limitations. Retargeting is entirely dependent on cookies, which creates several structural vulnerabilities:

• Users browsing in incognito mode have their cookies deleted when they close the browser, making retargeting impossible for those sessions.
• Users can explicitly block third-party cookies through browser settings, cutting off retargeting entirely.
• Cross-device tracking is inherently difficult. Desktop and mobile devices use different technologies for storing and reading cookies, meaning a user browsing on a laptop and then switching to a phone appears as two separate, unconnected entities without additional matching techniques.
• Safari on iOS blocks third-party cookies by default, making Apple device users substantially harder to retarget.
• The iOS 11 cookie policy introduced further constraints: first-party cookies on Apple devices expire within 24 hours of being stored. This significantly compresses the available retargeting window on the Apple ecosystem and limits the ability to accumulate user-behaviour data over time.

Device tracking vs. person tracking. Because retargeting is cookie-based, it tracks devices — not people. There is no technical mechanism to determine whether the same PC is being used by different members of a household. This frequently results in irrelevant ads being served to the wrong person, wasting advertiser budget. Compounding the issue, many users operate multiple devices simultaneously, making attribution and consistent targeting across that device ecosystem a persistent challenge.

Platforms that require users to log in across devices — Google, Amazon, Apple, eBay, and Yahoo, among others — can target more precisely because they leverage personally identifiable information tied to authenticated accounts. For everyone else, deterministic and probabilistic matching are the practical workarounds.

Deterministic matching ties devices to a known identity (e.g., an email address used to log in). Probabilistic matching is less precise, but infers device-to-user relationships based on shared signals such as IP address, location, age, gender, and interests that remain consistent across devices. As cookie-based tracking becomes increasingly constrained by browser policies and regulation, probabilistic methods are growing in relevance as a way to track user identities rather than just devices.

Segmentation and strategy. Retargeting isn't a set-it-and-forget-it tactic — it requires deliberate user segmentation. Dividing site visitors into meaningful segments (by product category viewed, cart abandonment stage, time since last visit, etc.) and tailoring ad creative to each segment significantly improves outcomes. Using custom parameters in retargeting tags at the product or category page level is the standard approach for achieving this granularity.

Relevance of destination. A retargeted ad should always link to the specific product page or landing page the user actually viewed — not the homepage. Sending retargeted users to a homepage is considered poor practice because it discards the behavioural context that made the retargeting relevant in the first place.

Privacy concerns. Retargeting inherently involves collecting and acting on data about users' browsing behaviour. Even when implemented responsibly, this raises legitimate privacy concerns among users and contributes to broader distrust of digital advertising. Advertisers using retargeting need to be mindful of applicable privacy regulations and transparent about data use.

Measuring Retargeting Effectiveness

The efficiency of a retargeting campaign can be tracked through several concrete metrics: website clicks, form submissions, and cost per lead (CPL) are among the most useful. These give advertisers levers to adjust targeting parameters, creative, and segmentation over time.

Even when a retargeting campaign doesn't produce immediate conversions, it can generate value through extended exposure — reinforcing brand awareness among users who are still in the consideration phase. Conversion and brand lift are both legitimate goals, and a well-structured retargeting strategy can serve either.