How Remarketing Lists Work in Google Ads (and Why Most eCommerce Stores Set Them Up Wrong)

Remarketing is one of the most talked-about tactics in eCommerce advertising and one of the most often misconfigured. A store owner creates an audience, attaches it to a campaign, sees 0 users and decides remarketing does not work. Usually the strategy is fine and the setup is the problem.

Here is how remarketing in Google Ads actually works, and how to build a setup that does.

How Remarketing Lists Are Built

A remarketing list is an audience of people who have visited your site or taken a particular action on it. Google builds these lists by recognising users across visits through cookies, advertising IDs and, with Enhanced Conversions, hashed email addresses.

There are two places to create them:

In Google Ads (tag-based lists): with the Google Ads tag (gtag.js) on your site, you can build lists based on page visits, events and tag parameters. They live in Google Ads Audience Manager.

In GA4 (audience export): you build audiences from GA4 event and parameter data and publish them to your linked Google Ads account, where they appear as imported audiences.

Both work. GA4 audiences are more powerful because GA4 holds richer behavioural data: you can build audiences from complex event sequences, parameter values and predictive signals that the Google Ads tag alone cannot match.

The Most Common Mistake: Minimum List Sizes

A remarketing list must reach a minimum size before it can be used for targeting:

• Search and Shopping (RLSA): 1,000 active visitors in the last 30 days
• Display: 100 active visitors
• YouTube: 1,000 active visitors

If a list shows 0 users or sits below the minimum, it cannot be used. It will appear in your campaign but never actually affect bids or delivery.

New and low traffic stores hit this wall all the time. A Shopify store with 300 sessions a month cannot build a 1,000 user Search remarketing list, however well it is set up. For these stores, Customer Match lists (built from email addresses rather than site visits) are a more practical starting point. They also need 1,000 matched users, but you can build them from existing customer emails instead of current traffic.

Build the Right Audience Tiers

Remarketing pays off when you treat visitors differently according to how close they are to buying. Not every visitor is equally worth retargeting.

A practical tier structure for eCommerce:

Tier 1: basket abandoners (highest intent)

People who added a product to their basket but did not buy. This group converts at the highest rate of any remarketing segment, because the intent is explicit: they wanted to buy and something stopped them.

In GA4: create an audience with event add_to_cart and not purchase, with a 30 day lookback.

In Google Ads Audience Manager: create a website visitors list with URL containing “/cart” and exclude URL containing “/checkout/thank-you” (or your own order confirmation URL).

Tier 2: product page viewers (high intent)

People who viewed specific products but did not add to basket. They showed product-level interest but did not reach the add to basket decision.

In GA4: create an audience with event view_item and not add_to_cart.

In Google Ads: create a list with URLs matching your product URL pattern (for example “/products/”) and exclude the basket abandoner list.

Tier 3: general site visitors (lower intent)

Everyone who visited within the lookback period, excluding buyers. This is the broadest audience, useful for upper funnel retargeting at lower bids.

Tier 4: past purchasers (exclusions and lifetime value)

People who have bought from you. Use this list twice: exclude it from prospecting campaigns (so acquisition budget is not spent on existing customers), and target it with repeat purchase or cross-sell campaigns.

RLSA in Shopping Campaigns

RLSA (remarketing lists for search ads) works differently in Shopping. You cannot show Shopping ads only to list members, because there are no keywords to target. In Shopping, RLSA works as a bid adjustment.

You add a remarketing list to a Shopping campaign or ad group and set an adjustment, for example +40% for basket abandoners, so Google bids 40% more when the search comes from someone on that list. (Bid adjustments apply when you bid manually; with Smart Bidding, Google uses the audience as a signal and sets bids itself.)

The effect: your products rank higher and show more often for people who have already shown interest. Budget still covers every eligible search, but you compete harder for the high intent subset.

To set it up in Standard Shopping, go to Audiences, add the list and set the adjustment. In PMax, audience signals do a similar but less directly controllable job: add your remarketing lists as signals to steer PMax towards people who resemble your past visitors.

Creating Remarketing Audiences in GA4

In GA4, go to Configure, Audiences and click New Audience.

For basket abandoners:

• Include: event name equals add_to_cart
• Exclude: event name equals purchase
• Membership duration: 30 days

For product page viewers:

• Include: event name equals view_item
• Exclude: event name equals add_to_cart
• Membership duration: 14 days

Once created, the audience is published to your linked Google Ads account automatically, provided the link has personalised advertising enabled.

Audiences take 24 to 48 hours to start populating in Google Ads. They start at 0 users and grow as matching visitors arrive. High engagement audiences such as basket abandoners grow faster than broad ones.

Check the size: after 48 hours, look in Google Ads Audience Manager. If the audience shows “< 1,000” or a number below the threshold, it is not yet usable for Search or Shopping RLSA, so keep monitoring. A store with 5,000 sessions a month should reach 1,000 basket abandoners within one to three weeks, depending on its add to basket rate.

Why Your Audience Shows 0 Users

If an audience still shows 0 users after 48 hours, the likely causes are:

GA4 is not tracking the event. If add_to_cart never fires in GA4, the basket abandoner audience will never fill. Check DebugView or the Events report to confirm the event arrives.

GA4 is not linked to Google Ads. Audiences only export to linked accounts. In GA4, go to Admin, Product links, Google Ads links and confirm the link is active.

The conditions are too strict. An audience with five conditions that must all be true at once may match almost nobody. Simplify the conditions and see whether it starts to fill.

Consent Mode is limiting cookie-based audiences. If many visitors decline consent, they are left out of cookie-based audiences, so your usable pool is smaller than your total traffic. That is expected, and the right response is Enhanced Conversions plus Customer Match for coverage beyond cookies.

Display Remarketing vs Search Remarketing

They do different jobs and should be judged separately.

Search remarketing (RLSA) reaches people already searching for products. The intent is there; you are just bidding harder for people more likely to buy from you specifically. It is very efficient because extra spend only goes on searches that already show buying intent.

Display remarketing shows banner ads to past visitors while they browse other sites. There is no active search, so you are keeping your brand visible and nudging people who have not bought yet. Conversion rates are lower than RLSA because the context is passive browsing rather than active searching.

Use different benchmarks for each. A 200% ROAS from Display remarketing and a 500% ROAS from RLSA can both be fine, depending on your category and margins. Comparing them directly and cutting Display because it “performs worse” ignores that they work in completely different contexts.

Frequency Caps for Display Remarketing

Without frequency limits, Display remarketing can show one person the same ad dozens of times a day, which annoys people and wastes budget.

Set an impression frequency cap at campaign level for Display. A reasonable starting point is three to five impressions per person per day and 10 to 15 per week, which keeps you visible without overexposure.

Display campaigns have frequency settings under Campaign Settings. In Performance Max, frequency is handled automatically by Google with little manual control.

The Exclusion Most Stores Skip

Always exclude past purchasers from prospecting campaigns. It is one of the most valuable uses of remarketing lists, and most stores do not do it.

Someone who bought last month is not a new customer to acquire. Spending acquisition budget to bring them back is re-engagement, not prospecting, and re-engagement is usually cheaper through email or Customer Match than through prospecting bids.

Add your purchaser list as an exclusion on every prospecting campaign, so prospecting budget stays focused on people who have never bought from you. That is where the real acquisition value lies.