How to Use Google Ads Remarketing: Complete Guide

Remarketing with Google Ads, also known as retargeting, is the practice of showing ads to people who have already interacted with your website, your app or your business online. It is not about winning back those who did not buy: it is about distinguishing, within that audience, those who were ready to buy but were interrupted from those who were only exploring, and communicating with each in the appropriate way. Visitors who have already interacted with the site convert at rates 3–7 times higher than cold traffic, with cost per lead often 50–60% lower (Tready, 2026). In 2026, remarketing has undergone a structural transformation: with the end of third-party cookies on Chrome, completed in 2025, strategies based on generic pixels have lost effectiveness, and remarketing built on first-party data (customer lists, CRM, verified conversions) has become the dominant approach and the most resilient to regulatory changes on privacy.

The remarketing paradox: 97% of visitors do not convert on their first visit

The vast majority of visitors who reach a website, through any channel, organic or paid, do not convert on their first visit. Average e-commerce conversion rates stand between 2% and 4% (Groas, 2026): this means that 96–98% of the users you have acquired with effort and budget leave without buying. Remarketing is the tool that lets you keep that relationship, stay visible while the user is evaluating, and return with the right message at the moment they are most likely to come back and convert.

The logic of remarketing is not “pestering those who didn’t buy”: it is building a communication sequence that accompanies users along their decision-making journey. Someone who visited the homepage and left after 10 seconds has a different intention from someone who added a product to the basket and abandoned it before payment. Effective remarketing treats these two users in completely different ways, with messages, creative and offers calibrated to their position in the purchase funnel.

The 4 types of remarketing on Google Ads in 2026

The end of third-party cookies: the paradigm shift that is redrawing remarketing

With the end of third-party cookies on Chrome, completed in 2025, and iOS’s growing restrictions on tracking, traditional remarketing based on generic pixels has lost a significant share of its coverage. Companies that kept relying on the old methods saw the effectiveness of their campaigns collapse; those that adapted promptly maintained, and often improved, their performance (Tready, 2026).

Modern remarketing in 2026 falls into two categories: remarketing based on pixels and the Conversions API (site visitors tracked via JavaScript plus server-side tracking to offset signal loss) and remarketing based on first-party lists (existing customers, CRM leads, newsletter subscribers). The second category has become the more strategic: first-party data is owned by the company, does not depend on third-party cookies, is not subject to browser restrictions and produces audiences that are more precise and more resilient to regulatory change.

Conversions API: server-side tracking as a prerequisite

To keep pixel-based remarketing effective in 2026, the JavaScript tag on the site is no longer sufficient on its own. Google’s Conversions API (or Enhanced Conversions) sends conversion data directly from the company’s servers to Google, regardless of the user’s browser or cookie blocking. The result is more complete tracking that narrows the signal gap and allows Google’s AI to build more accurate remarketing audiences. For most companies, implementing the Conversions API requires technical support or integration with platforms such as Google Tag Manager with a server-side configuration.

The audience segmentation framework: the 5 levels of proximity to conversion

Audience segmentation is the factor that most distinguishes high-ROAS remarketing campaigns from mediocre ones (Tready, 2026). It is not enough to create a “site visitors in the last 30 days” list and show everyone the same ad. Effective remarketing builds distinct audiences based on user behaviour, and treats each one with appropriate messages, creative and offers.

1. Level 1, Abandoned basket (top priority): users who added products to their basket but did not complete payment. This is the segment with the highest conversion rate and the highest recovery value. The message must be direct, often with a specific incentive (free delivery, a time-limited discount), and the creative must show the abandoned products. Recommended list duration: 7–14 days.
2. Level 2, Visitors to specific product pages: users who visited product pages without adding to the basket. They have shown interest but not yet purchase intent. Dynamic remarketing, which automatically shows the products viewed, is the most effective solution. Recommended duration: 14–30 days.
3. Level 3, Site visitors with medium engagement: users who have browsed several pages or spent more than X seconds on the site, but have not visited specific product pages. They are in the exploration phase. The message should offer value (content, a guide, a consultation offer) rather than be directly commercial. Recommended duration: 30–60 days.
4. Level 4, Existing customers (Customer Match): those who have already purchased. The goal is not the first conversion but cross-selling, upselling or loyalty. The message must be radically different from the one for prospects, not “buy”, but “discover the new collection”, “access the customer-only offer”, “leave us a review”. This segment is also excellent for exclusion: in many cases, it makes sense to exclude existing customers from prospecting campaigns so as not to waste budget on those who already know the brand.
5. Level 5, General visitors (low priority): users who visited only the homepage or left the site within a few seconds. They include a lot of noise, bots, accidental visitors, irrelevant users. They belong in the general brand awareness campaign, not in high-pressure sales campaigns. Recommended maximum duration: 60–90 days, with low impression frequency.

RLSA: remarketing in Search campaigns

RLSA (Remarketing Lists for Search Ads) are one of the most powerful and underused features of Google Ads in 2026 (Groas, 2026). They allow you to tailor Search campaigns, bids, ads, keywords, according to whether the user has already visited the site. The strategic advantage is precise: with RLSA, you can target keywords you would normally avoid for cold traffic, because they are too generic, too expensive or too low in intent, exclusively for users who have already shown interest in your brand.

Example: a company selling project management software would not normally target the keyword “project management” for cold traffic, as the CPC is high and the intent too generic. With RLSA, however, it can target that keyword exclusively for users who have already visited the site: these users already know the brand, and that generic keyword becomes a high-value touchpoint rather than a waste of budget (Groas, 2026). RLSA can be set up in two modes:

• Observation: the campaign reaches all users, but collects separate performance data for those on the remarketing list. This shows how much better site visitors perform than cold traffic, and allows bid adjustments to be applied to that segment.
• Targeting: the campaign shows ads exclusively to users on the remarketing list. Suited to campaigns dedicated to recovering high-intent visitors.

Customer Match: your existing customers as a strategic asset

Customer Match lets you upload existing contact lists, emails, phone numbers, addresses, to Google Ads in order to reach these people on Search, Display, YouTube and Gmail when they are active on Google platforms. It is the most precise form of remarketing: it relies not on browsing behaviour (which can be noisy) but on real identity data.

Since 7 April 2025, Google has imposed a maximum duration of 540 days on Customer Match lists: users added or updated more than 540 days ago are automatically removed from the list (ALMCorp, 2026). To keep a list active and eligible for targeting, at least 100 members updated within this time window are required. This makes regular refreshing of Customer Match lists an operational necessity, not an optional best practice: an outdated CRM produces lists that gradually empty without the practitioner noticing.

The three strategic uses of Customer Match in 2026:

1. Cross-selling and upselling to existing customers: show complementary or higher-tier products to those who have already purchased. The conversion rate is structurally higher than with prospects because trust in the brand already exists.
2. Exclusion from prospecting: exclude existing customers from prospecting campaigns so as not to waste budget on people who already know the brand and do not need to be acquired.
3. Audience signal for Performance Max: upload customer lists as an audience signal in Performance Max campaigns to speed up the learning phase. With quality Customer Match lists, the learning phase shrinks from 4–6 weeks to 10–14 days (Tready, 2026).

The hidden risk: overexposure that burns the brand

Remarketing carries a specific risk that almost no technical guide names with due seriousness: overexposure. Showing the same ads to the same users too often does not increase conversions; it reduces them. And it has a side effect that goes beyond campaign performance: it damages brand perception. A user who sees your display ad 40 times in a week does not become more inclined to buy; they become irritated, and associate that annoyance with the brand.

Frequency management is a brand governance decision, not merely a matter of campaign optimisation. Recommended frequency caps vary by campaign type and audience segment:

• Display remarketing to cold audiences (level 5): 3–5 impressions a week per user. Enough to maintain visibility; not enough to be intrusive.
• Display remarketing to warm audiences (levels 1–3): 7–15 impressions a week in the first phase (1–7 days after the interaction), then a progressive reduction if the user does not convert.
• List duration: lists that are too long (90+ days for generic audiences) include users who are no longer relevant and waste budget. Shorter lists with frequent refreshes are better than long, outdated ones.

The 5 Ws of remarketing with Google Ads

• Who: Any company with website traffic, from SMEs to enterprises. Remarketing is particularly effective for those who have already invested in acquisition campaigns: it amplifies the return on budget already spent, recovering a share of users who would otherwise be lost for good.
• What: A system of campaigns that shows personalised ads to users who have already interacted with the brand, tailoring the message to behaviour, intent and position in the purchase funnel.
• When: As soon as the site has enough traffic to build meaningful audience lists (as a rough guide, at least 100 users per list for Display; 1,000 for RLSA). For Customer Match, the list is built from the CRM regardless of traffic.
• Where: Across the entire Google network, Display, Search (RLSA), YouTube, Gmail, Shopping, with messages adapted to the format and context of each channel.
• Why: Because the cost of acquiring the traffic has already been paid, and without remarketing, 96–98% of that budget produces visitors who will never return. Remarketing turns sunk acquisition spend into an investment that pays back over time.

Bliss Agency case studies: remarketing as a component of integrated performance

At Bliss Agency, remarketing is not a campaign separate from strategy: it is the component that turns acquisition budget into an ongoing relationship with the audience. The Profumum Roma case, a luxury heritage perfume house, documents how, on Google Ads ecommerce, the integrated campaign system (acquisition + remarketing + Customer Match on existing customers) produced a ROAS of 17,1. In a luxury segment with a long purchase cycle, remarketing to audiences segmented by purchasing behaviour, those who visited specific olfactory families, those who viewed limited-edition pages, those who bought once and have not yet repurchased, is the main driver of the difference between an average ROAS and an exceptional one.

The Doreca case, an HO.RE.CA. beverage distributor, shows how the same logic applies in a B2B context: RLSA on Search campaigns, with more aggressive bids for users who had already visited the site, segmented by product type, contributed to a CTR of 13,36% on 51.599 clicks, significantly above the sector average. For operational details: Bliss Agency case studies.

2026 trends: where remarketing is heading

Demand Gen: the new remarketing on YouTube and Discovery

Demand Gen campaigns, the evolution of Discovery campaigns, make it possible to reach remarketing audiences on YouTube Shorts, YouTube In-stream, Gmail and Discover with visual ads (images and video). This is Google’s answer to the demand for video formats in remarketing: not just display banners, but personalised videos for users who have already interacted with the brand. In 2026, Demand Gen has become the preferred format for remarketing to warm audiences with brand content, storytelling, social proof and product comparisons, which hold up better in video than in display banners.

AI and lookalike audiences: from remarketing to expansion

In 2026, high-quality remarketing lists, especially Customer Match lists of customers who have converted, have become the starting point for automated audience expansion. Google’s AI uses these lists as a signal to find users with similar characteristics (similar audiences) among cold traffic. It is no longer necessary to build interest or demographic segments manually for prospecting: you start from your best customers and let the AI find those who resemble them.

The right measure: incrementality, not attribution

Remarketing suffers from a structural measurement problem: attribution models tend to overestimate its contribution. A user who would have converted anyway, because they had already decided to buy, is counted as a remarketing conversion when they see a banner before returning to the site. In 2026, the most advanced companies measure remarketing not through standard attribution but through incrementality tests: a control group is created (users who do not see the remarketing ads) and the difference in conversion rate between the two groups is measured. Only this difference represents the true incremental contribution of remarketing (Groas, 2026).

New Connections (FAQ)

What is remarketing on Google Ads?

Remarketing is the practice of showing ads to people who have already visited your website, used your app or interacted with your business online. On Google Ads, it works through four main formats: Display (banners on sites in the Google network), RLSA (Search ads tailored to site visitors), Dynamic Remarketing (ads featuring the specific products the user viewed) and Customer Match (targeting based on lists of existing customers uploaded by the company). Users already exposed to the brand convert at rates 3–7 times higher than cold traffic (Tready, 2026).

How do you create a remarketing list on Google Ads?

Remarketing lists are created in Google Ads under Tools → Audience manager → Data segments. The prerequisite is that the Google Ads tag (or GA4 linked to the Ads account) is correctly installed on every page of the site. You define the inclusion rule (e.g. “visited the /carrello page” or “spent more than 60 seconds on the site”) and the list membership duration (from 1 to 540 days). Lists become available for targeting once they reach at least 100 active users (Display) or 1,000 users (Search/RLSA).

What is RLSA and how does it work?

RLSA (Remarketing Lists for Search Ads) let you tailor Search campaigns according to whether the user has already visited the site. You can raise bids for site visitors (who are more likely to convert), show different ads, or target keywords that would otherwise be too generic or too costly, exclusively for people who already know the brand. They are set up in Observation mode (all users, with separate data for those on the list) or Targeting mode (only users on the list). It is one of the most underused tools in Google Ads, with one of the highest potential impacts on Search campaign performance.

What is Customer Match and how is it used?

Customer Match lets you upload lists of existing customers (emails, phone numbers, addresses) to Google Ads to reach them on Search, Display, YouTube and Gmail. Since 7 April 2025, Customer Match lists have a maximum duration of 540 days: users not refreshed within that window are removed automatically (Google, 2025). It is used for cross-selling and upselling to existing customers, for excluding them from prospecting campaigns, and as an audience signal in Performance Max to speed up the learning phase (from 4–6 weeks to 10–14 days with quality lists).

How many times should a user see my ad in remarketing?

The optimal frequency depends on the segment and the objective. For generic audiences (short visits, low proximity to conversion): 3–5 impressions a week. For warm audiences (abandoned basket, product pages): 7–15 impressions in the first week, reduced progressively if the user does not convert. Excessive frequency does not increase conversions, it reduces them, and associates a sense of irritation with the brand. Frequency management is a brand governance decision even before it is a campaign optimisation one.

BLISS®
© 2026. ALL RIGHTS RESERVED