What Is Retargeting? How It Works and How to Set It Up

In the run-up to Black Friday, a shopper lands on an online store, compares prices on a product page, adds an item to the basket and leaves before reaching checkout. They now know the brand and have shown clear interest in the product, but they have not made a decision yet. Retargeting is the advertising approach built to turn that interest into a conversion. This guide explains what retargeting is, how it works and how to set it up, from tagging and audiences through to measurement, with a data-led perspective.

What Is Retargeting?

Retargeting is the practice of showing ads to people who have already visited a website or app, or otherwise interacted with a brand, as they browse other websites, watch YouTube or scroll through social media. Its purpose is to win back users who have shown interest but have not yet converted.

What sets it apart from prospecting is that the audience has already seen the brand and the product. The message therefore focuses less on introduction and more on reminding, persuading and shortening the time to a decision.

How Does Retargeting Work?

Retargeting relies on small pieces of code added to a website, known as tags or pixels. When someone visits the site, these tags record which pages they viewed, whether they added a product to the basket and whether they submitted a form. These signals are used to build audience lists, and ad platforms then show ads to the people on those lists.

Every audience list has a membership duration, which the advertiser sets to control how long a user stays on the list. Beyond on-site behaviour, businesses can also build audiences by uploading their own customer data, such as email addresses or phone numbers. Because browsers such as Safari and Firefox restrict third-party cookies by default, first-party data collected by the brand itself has become the most valuable input for retargeting.

What Are the Main Types of Retargeting?

• Site visitor retargeting: People who visited the website see brand or category ads based on the pages or categories they browsed.
• Dynamic retargeting: A product catalogue (for example, product data in Google Merchant Center) is connected to the ad platform, so users see the exact products they viewed or similar items. It is particularly well suited to online stores with large product ranges.
• Customer list retargeting: Email or phone data from the CRM is matched through tools such as Google Customer Match or Meta Custom Audiences. It works well for cross-selling and repeat purchase campaigns.
• Engagement-based retargeting: People who watched the brand’s YouTube videos or engaged with its Instagram or Facebook profile are targeted. This helps warm up an audience that has interacted with the brand but has not yet visited the site.

How to Set Up Retargeting in 5 Steps

1. Define Your Goal and Conversions

Decide at the outset which outcome the campaign should optimise for, whether that is a sale, a form submission or a demo request. Audiences built without a clear goal are fed the wrong signals. Defining conversions correctly in Google Analytics 4 and Google Ads is therefore the first step.

2. Install Tags With Consent Management

Deploying the Google tag and the Meta Pixel through Google Tag Manager makes them manageable without editing the site’s code for every change. To reduce the impact of browser restrictions, server-side options such as the Meta Conversions API are also worth considering. Under privacy laws such as the GDPR or Turkey’s KVKK, non-essential marketing cookies require the user’s explicit consent, so tags must work together with the consent management platform.

3. Segment Audiences by Intent and Recency

“All visitors” on its own is far too broad an audience. Splitting users by intent signal and time since their last visit makes messages far more relevant. Cart abandoners from the last 1–7 days are the warmest audience and respond fastest to a product reminder and a small incentive. Product page viewers from the last 8–30 days can be shown similar products and reviews, while blog readers are better served with content that points to a next step rather than a hard sell.

Segment	Time Window	Suggested Message
Cart abandoners	Last 1–7 days	Product reminder, plus an incentive such as free delivery if needed
Product page viewers	Last 8–30 days	Similar products, customer reviews, trust signals
Blog and guide readers	Last 30–90 days	A next step such as a product category, guide or newsletter
Purchasers	After purchase	Exclude from related campaigns; complementary products only

 

Excluding people who have already completed a purchase is just as important. Otherwise, customers keep seeing ads for something they have already bought, which wastes budget and damages the customer experience.

4. Match Channels and Messages

On the Google side, the Display Network and YouTube make it possible to reach users across a large part of the web. In Performance Max campaigns, however, website visitor and customer lists act as audience signals that guide the algorithm rather than as strict targeting, and the campaign can also show ads to people outside those signals. On Meta, Facebook and Instagram are strong channels, particularly for dynamic product ads.

5. Set Frequency Caps and Budget

Showing the same ad to the same person without limit turns a reminder into an irritation. In Google Ads, frequency caps can be set for Display and Video campaigns, but the setting is not available in Demand Gen campaigns. Where supported, apply frequency caps, refresh creatives regularly and allocate budget according to each segment’s conversion potential.

How Should B2B and Lead Generation Retargeting Work?

In B2B, buying decisions often take weeks and involve several decision-makers. Rather than a single “buy now” message, this calls for a staged approach. Visitors who viewed the pricing or demo page without submitting a form can be shown case studies and product videos, while blog readers can be offered guides and webinars.

How Do You Measure Retargeting Performance?

The core metrics are ROAS (return on ad spend), CPA (cost per acquisition) and conversion rate. Retargeting measurement has an important trap, however: because the campaign targets people who are already close to buying, sales that would have happened anyway may be credited to it. View-through conversions in particular can make performance look stronger than it is.

Report click-through and view-through conversions separately for this reason. Where possible, measure the true incremental impact with holdout tests that exclude part of the audience from ads, or with Conversion Lift studies that compare people who saw the ads with people who did not.