Remarketing in 2026: how it works and what really changed
Remarketing shows ads to people who already visited your site. Chrome did not drop third-party cookies: in Belgium, consent decides what still works
Summary
- Chrome did not drop third-party cookies: on 22 April 2025 Google decided to leave the choice with the user, and in October 2025 it retired most Privacy Sandbox technologies
- Safari has fully blocked third-party cookies since 2020, Firefox on desktop gives every site its own cookie jar
- In Belgium, consent decides: without the ad_personalization signal in Consent Mode v2, your remarketing lists in Google Ads receive no data
- Google Ads requires at least 100 active users in 30 days per list; Meta works with recommendations, a hard minimum only applies to lookalikes
- In a randomised experiment, retargeting lifted purchases by 10.5% (Journal of Marketing Research, 2017), but popular figures such as 70% more conversions cannot be traced to a study
You visit an online store, look at a pair of shoes and leave. A day later, those same shoes show up in your Instagram feed. That is remarketing, and it is one of the oldest tricks in online advertising. Two stories about remarketing in 2026 are doing the rounds: that cookies have disappeared and remarketing therefore no longer works, or that nothing has changed.
Neither is right. Below, you’ll find what remarketing is, what Chrome, Safari and Firefox actually decided, why in Belgium consent mostly determines what still works, and how to measure remarketing without relying on third-party cookies. With the sources included, and without the popular figures that have no study behind them.
What is remarketing and how does it work?
Remarketing means advertising again to people who already saw your website, app or ads. A tag or pixel on your site, such as the Google tag or the Meta Pixel, adds visitors to a list, for example everyone who viewed a product page. Google Ads, Meta Ads or LinkedIn Ads then show the people on that list targeted ads while they are on other sites, social media or YouTube.
Google uses the term remarketing in its documentation, many other sources speak of retargeting. The retargeting entry in the glossary treats both terms as synonyms.
The principle is simple. People who already know you are closer to a purchase or enquiry than a cold audience. The execution has become less simple, because browsers and privacy rules determine who you can still add to a list and for how long.
ClickForest uses remarketing as part of Meta Ads for online stores and SMEs in Mechelen, Antwerp and the rest of Flanders. Visitors who viewed a product or added it to their cart get a targeted ad, so that part of the traffic you already paid for comes back.
Has Chrome dropped third-party cookies?
No. On 22 April 2025 Google decided to keep its current approach: users choose in Chrome’s settings whether they allow third-party cookies, and there will be no separate choice prompt. Incognito mode does block them by default. In October 2025 Google also retired most Privacy Sandbox technologies.
“We’ve made the decision to maintain our current approach to offering users third-party cookie choice in Chrome, and will not be rolling out a new standalone prompt for third-party cookies.”
— Anthony Chavez, VP Privacy Sandbox at Google
That is in the announcement of 22 April 2025 on Google’s Privacy Sandbox site. On 17 October 2025, Topics, Protected Audience, Attribution Reporting and IP Protection, among others, were retired, “in light of their low levels of adoption” (Google, 2025). This is really happening: Related Website Sets has been removed in Chrome 153 (Chrome for Developers).
Anyone who reads that third-party cookies have disappeared from Chrome is reading about a plan Google itself has scrapped. For Chrome users, remarketing via third-party cookies technically still works as before.
What do Safari and Firefox do differently?
Safari and Firefox restrict tracking much more strictly than Chrome. Safari has blocked all third-party cookies since 24 March 2020, and since 2019 it has capped cookies set via JavaScript at 7 days at most. Firefox on desktop gives every website its own cookie jar and blocks cross-site tracking cookies from known trackers by default. On those browsers, a remarketing list fills up more slowly and less completely.
“Cookies for cross-site resources are now blocked by default across the board.”
— John Wilander, WebKit (Apple)
That sentence comes from the WebKit announcement of 2020. Since Intelligent Tracking Prevention 2.1 (2019), persistent cookies set via document.cookie have been capped at 7 days (WebKit, 2019). That the remarketing window on Safari is therefore shorter than on Chrome is something we infer from those rules; Apple does not say so itself. Safari 27 goes further and blocks network access to so-called identity resolution infrastructure, which links email addresses to a tracking ID (WebKit, 2026). Apple does not say which companies or platforms this covers.
Firefox made Total Cookie Protection the default for all desktop users worldwide in 2022 (Mozilla, 2022), and blocks the cross-site tracking cookies of known trackers by default, based on a list from Disconnect (Mozilla Support).
Why does consent matter more than the browser in Belgium?
Because Google requires advertisers to collect consent from users in the European Economic Area before their personal data is used for personalised ads and remarketing. Since March 2024, Google has asked advertisers to pass on that consent via Consent Mode v2. Without the ad_personalization signal, remarketing lists in Google Ads receive no data, whichever browser the visitor uses.
Consent Mode does not collect consent itself. A cookie banner or consent platform asks for the choice, and Consent Mode adjusts the behaviour of the Google tags accordingly. Google distinguishes four signals: ad_storage, analytics_storage, ad_user_data and ad_personalization (Google for Developers). In the basic implementation, the tags only load after a choice, and if the visitor declines, nothing goes to Google. In the advanced implementation, the tags load immediately and, without consent, send measurements without cookies.
“If you are collecting data for remarketing with Google’s advertising services, you must implement Consent Mode and set the ad_personalization flag.”
— Simo Ahava, Senior Data Advocate at Reaktor
Google itself puts it even more briefly: “Personalized advertising and remarketing will require affirmative end-user consent signals” (Google Analytics Help). For a Belgian advertiser, the cookie banner is therefore the real gateway for remarketing, much more than Google’s choice for Chrome. Whether your measurement handles that properly is covered in the guide to GA4 on Shopify.
How large does your audience need to be for remarketing?
Google Ads requires at least 100 active users or visitors in the last 30 days, for Display, Search and YouTube. Meta has no hard minimum for a website audience, but recommends waiting until there are several hundred people in it. For a lookalike audience, Meta does require at least 100 people in the source audience.
| Platform | Minimum for a list | Good to know |
|---|---|---|
| Google Ads (Display, Search, YouTube) | 100 active users in the last 30 days | Someone stays in a list for at most 540 days; older blogs still mention 1,000 users for Search |
| Meta, website audience | No hard limit, recommendation: several hundred people | Only counts visitors with a Meta account, excluding those who opted out of behavioural advertising |
| Meta, customer list | No hard limit, recommendation: at least 1,000 customers | A list that is too small can affect the delivery of your ad |
| Meta, lookalike | At least 100 people in the source audience | This is the only hard 100 at Meta |
Sources: Google Ads Help, Meta Business Help on website audiences and customer lists, and Meta for Developers on lookalikes. A website audience at Meta is also always smaller than your number of visitors. It only counts visitors with a Meta account, may miss some iOS 14 users and only starts filling from the moment you create it (Meta Business Help). So create your lists early, even if you only want to advertise later.
What is dynamic remarketing and what do you need for it?
Dynamic remarketing shows a visitor exactly the products they viewed, straight from your product feed. With Google, you need a feed in Merchant Center and a tag that passes on product views, add-to-cart actions and purchases. With Meta, you need a catalogue, a pixel with three fixed events and a connection between pixel and catalogue.
At Google, those events are called view_item, add_to_cart and purchase, and the product ID in the tag must match the ID in your Merchant Center feed (Google Ads Help). At Meta, they are ViewContent, AddToCart and Purchase, with product IDs that match your catalogue (Meta Business Help). Meta has renamed this ad format: what used to be called Advantage+ catalog ads is now called ads showing product media.
For online stores, ClickForest works with Advantage+ shopping campaigns and catalog sales based on the product feed. For MyEssentials, they deliver consistent revenue and a positive ROI. Whether you steer on revenue or on margin makes a big difference: the article on POAS versus ROAS explains why.
Does remarketing really work?
Yes, but more modestly than many marketing figures claim. In a randomised field experiment, retargeting lifted website visits by 17.2% and purchases by 10.5%. Another experiment found that dynamic retargeting worked less well on average than generic ads, except for visitors who already knew what they were looking for. Each study was run with a single advertiser, so treat them as indicative.
The first experiment comes from Johnson, Lewis and Nubbemeyer, published in the Journal of Marketing Research in 2017 (Crossref). It concerns one display campaign of one online retailer. The second, by Lambrecht and Tucker from 2013, was run at an online travel company:
“Surprisingly, the data suggest that dynamic retargeted ads are, on average, less effective than their generic equivalents.”
— Anja Lambrecht and Catherine Tucker, Journal of Marketing Research (2013)
According to the same study, visitors who had already looked at review sites did not respond worse to dynamic ads (LBS Research). The practical lesson: show generic ads to people who are still exploring their options, and specific products to people who are already searching in a targeted way. How to split those stages in your campaigns is covered in the article on the advertising strategy per stage of the funnel.
Be wary of the well-known figures. Statistics such as “retargeting increases conversion by up to 150%” or “retargeted visitors are 70% more likely to convert” appear on many marketing sites, but if you trace them back, you won’t find a study behind them.
How do you measure remarketing without relying on third-party cookies?
With server-side measurement alongside the pixel in the browser. Meta’s Conversions API sends purchases and leads directly from your server to Meta, and Google’s Enhanced Conversions matches hashed customer data to Google accounts. Both make your measurement more reliable. They are not a way around consent: in the EEA, it is still required.
Meta itself writes: “Data from the Conversions API is less impacted than the Meta Pixel by browser loading errors, connectivity issues and ad blockers” (Meta Business Help). With Enhanced Conversions, Google hashes your own customer data with a one-way hash (SHA256) and matches it to Google accounts that were signed in when someone engaged with your ad (Google Ads Help). At Google, the ad_user_data signal determines whether that data may be sent.
For Shopify stores, ClickForest sets up that server-side tracking by default, for other websites on request. If you are unsure whether your pixel, your Conversions API and Consent Mode work together correctly, an ads audit by ClickForest for SMEs in Mechelen, Antwerp and across Flanders checks whether the conversion events are set up correctly in Google Ads, Meta Ads and LinkedIn Ads, and whether server-side tracking is active.
Which channel should you choose for remarketing: Google, Meta or LinkedIn?
That depends on what you sell. Google reaches visitors while they search, on YouTube and on websites. Meta works well for products you can show and for catalogue ads. LinkedIn suits B2B, where you want to re-engage decision-makers who viewed your site or your posts. In all three cases, the landing page after the click helps determine whether the click delivers anything.
| Channel | Strong for | Watch out for |
|---|---|---|
| Google Ads | Reaching visitors again while they search, on YouTube and on websites | Consent Mode v2 and at least 100 active users per list |
| Meta Ads | Online stores, products you can show, catalogue ads | A website audience only counts visitors with a Meta account |
| LinkedIn Ads | B2B: website visitors and people who engaged with your posts | Only shows ads to people who use LinkedIn |
For Bold Padel, ClickForest built a new Meta Ads structure for the Spanish padel market, with retargeting up to 180 days after the first visit. For B2B, ClickForest uses LinkedIn retargeting on website visitors and engagement audiences, to turn cold prospects into warm conversations. The article on LinkedIn marketing for B2B leads goes deeper into that, and the comparison of Google Ads and Facebook Ads for Shopify helps with the choice for an online store.
How ClickForest combines Google Ads, Meta Ads and LinkedIn Ads is explained under performance marketing. Someone who clicks back has to recognise what they saw before. What needs to be on the page for that is covered in the article on a good landing page.
Conclusion: remarketing still works, with consent as the gateway
Remarketing is not dead. Chrome keeps third-party cookies on, Safari blocks them, Firefox blocks those of known trackers, and in Belgium the cookie banner mostly decides who ends up in your lists. That makes remarketing smaller and less automatic than before. If you have the basics in place, you can still see part of the traffic you already paid for come back.
Start with the basics. Set up Consent Mode v2 correctly, create your lists early, measure with both the pixel and the Conversions API, and send visitors to a page that matches the ad. Expect a modest effect, and measure it yourself.
ClickForest uses remarketing for online stores, SMEs and B2B companies in Mechelen, Antwerp and across Flanders, via Meta Ads and LinkedIn Ads. Want to know how your remarketing is doing? Get in touch.
More results from every marketing budget, less waste
Want to get more from your ad budget with a strategy that translates into real revenue? Discover our performance marketing approach
Discuss your challenge directly with Frederiek: Book a free strategy call or send us a message
Prefer email? Send your question to frederiek@clickforest.com or call +32 473 84 66 27
Strategy without action remains theory. Let's take your next step together.
Frequently asked questions
There is no difference in substance: both words mean advertising again to people who already saw your website, app or ads. Google uses the term remarketing in its documentation, many other sources speak of retargeting. The ClickForest glossary treats them as synonyms.
Not for personalised ads. Google asks advertisers to collect consent from users in the European Economic Area for remarketing, and without the ad_personalization signal remarketing lists receive no data. The banner or consent platform collects that choice, Consent Mode passes it on to the tags.
No. Chrome keeps third-party cookies on by default, Safari blocks them, and Firefox blocks those of known trackers and isolates the rest per site. What changes is that remarketing relies more on first-party data, consent and server-side measurement. ClickForest therefore sets up server-side tracking alongside the pixel for online stores and SMEs in Flanders: by default for Shopify, on request for other websites.
Start with the basics: is Consent Mode v2 set up correctly, are your lists large enough, do the pixel and the Conversions API measure the same purchases or leads, and does the visitor land on a page that matches the ad? ClickForest helps SMEs in Mechelen, Antwerp and the rest of Flanders with that setup, and checks it in an ads audit.
Sources and references
Browsers and cookies:
- Google Privacy Sandbox: "Next steps for Privacy Sandbox and tracking protections in Chrome" (2025) · https://privacysandbox.google.com/blog/privacy-sandbox-next-steps
- Google Privacy Sandbox: "Update on plans for Privacy Sandbox technologies" (2025) · https://privacysandbox.google.com/blog/update-on-plans-for-privacy-sandbox-technologies
- WebKit: "Full third-party cookie blocking and more" (2020) · https://webkit.org/blog/10218/full-third-party-cookie-blocking-and-more/
- WebKit: "Intelligent Tracking Prevention 2.1" (2019) · https://webkit.org/blog/8613/intelligent-tracking-prevention-2-1/
- WebKit: "Blocking identity resolution infrastructure" (2026) · https://webkit.org/blog/18438/blocking-identity-resolution-infrastructure/
- Mozilla: "Firefox rolls out Total Cookie Protection by default to all users worldwide" (2022) · https://blog.mozilla.org/en/mozilla/firefox-rolls-out-total-cookie-protection-by-default-to-all-users-worldwide/
- Mozilla Support: "Enhanced Tracking Protection in Firefox for desktop" · https://support.mozilla.org/en-US/kb/enhanced-tracking-protection-firefox-desktop
- Chrome for Developers: "Chrome 153 beta" (2026) · https://developer.chrome.com/blog/chrome-153-beta
Google and Meta documentation:
- Google for Developers: "Consent mode" · https://developers.google.com/tag-platform/security/concepts/consent-mode
- Google Ads Help: "Updates to consent mode for traffic in European Economic Area" · https://support.google.com/google-ads/answer/13695607?hl=en
- Google Analytics Help: "Consent mode reference" · https://support.google.com/analytics/answer/13802165?hl=en
- Google Analytics Help: "EU user consent and consent mode" · https://support.google.com/analytics/answer/14275483?hl=en
- Google Ads Help: "How your data segments work" · https://support.google.com/google-ads/answer/2472738?hl=en
- Google Ads Help: "About dynamic remarketing" · https://support.google.com/google-ads/answer/3124536?hl=en
- Google Ads Help: "About enhanced conversions" · https://support.google.com/google-ads/answer/9888656?hl=en
- Meta Business Help: "Troubleshoot website custom audience size" · https://www.facebook.com/business/help/1523083297982009
- Meta Business Help: "How to use custom or lookalike audiences" · https://www.facebook.com/business/help/572787736078838
- Meta Business Help: "How to create a website custom audience in Meta Ads Manager" · https://www.facebook.com/business/help/1474662202748341
- Meta for Developers: "Lookalike Audiences" · https://developers.facebook.com/documentation/ads-commerce/marketing-api/audiences/guides/lookalike-audiences
- Meta Business Help: "Required Meta Pixel events and parameters for ads showing product media" · https://www.facebook.com/business/help/606577526529702
- Meta Business Help: "About Conversions API" · https://www.facebook.com/business/help/2041148702652965
Research and experts:
- Johnson, Lewis & Nubbemeyer: "Ghost ads: improving the economics of measuring online ad effectiveness", Journal of Marketing Research (2017) · https://api.crossref.org/works/10.1509/jmr.15.0297
- Lambrecht & Tucker: "When does retargeting work? Information specificity in online advertising", Journal of Marketing Research (2013) · https://lbsresearch.london.edu/id/eprint/388/
- Simo Ahava: "Consent mode V2 for Google tags" (2024) · https://www.simoahava.com/analytics/consent-mode-v2-google-tags/
