In this article
Remarketing still works. But if your ROAS has dropped since 2021 and your audience lists look smaller than they used to, the mechanics you relied on have changed. iOS 14.5 removed the signal. iOS 17 shortened the cookie window. Chrome is moving away from third-party cookies entirely. This guide explains what actually changed, what still works, and how Australian e-commerce brands are rebuilding retargeting with first-party data, server-side tracking, Performance Max, and Meta Advantage+ Shopping.
Contents
How remarketing works in 2026
The iOS effect on retargeting
Google remarketing in 2026
Meta remarketing in 2026
Server-side tagging and first-party data
Remarketing strategies that still work
Frequency caps and creative fatigue
TikTok and Pinterest remarketing
FAQ: remarketing in a privacy-first world
How remarketing works in 2026
Remarketing is the practice of showing targeted ads to people who have already interacted with your brand, using signals like website visits, app activity, purchase history, or email lists. In 2026, the defining shift is that those signals increasingly come from your own first-party data rather than third-party browser cookies.
The basic concept hasn’t changed. Someone visits your product page, doesn’t buy, and later sees your ad. What has changed is how you know they were there, and how reliably that information survives the journey from their browser to your ad platform.
Third-party cookies, which once powered most cross-site audience tracking, are now significantly constrained. Safari has blocked them since 2017. Firefox followed. Chrome, which holds the majority of browser market share, began restricting third-party cookies for a portion of users in 2024 under the Privacy Sandbox framework, with further changes continuing into 2026. The result: audience lists built entirely on browser cookies are smaller, less accurate, and less stable than they were three years ago.
The shift is toward signals you control. First-party data collected through your own website, CRM, email list, and point-of-sale system is now the foundation of effective remarketing. Combined with on-platform signals from Google and Meta, and supplemented by server-side tracking, this data produces audience lists that are more durable and more precise than what cookie pools ever delivered.
Sources: Baymard Institute (cart abandonment); Google/IAB industry benchmarks (CTR); AppsFlyer ATT benchmarks (opt-in rate). Figures represent general industry ranges.
The iOS effect on retargeting
Apple’s App Tracking Transparency (ATT) framework, introduced with iOS 14.5 in April 2021, required apps to ask permission before tracking users across other apps and websites. Only 30-40% of Australian users opt in. That decision cut the signal reaching Meta’s ad platform by more than half overnight, and the effects compounded with each subsequent iOS release.
Here’s what signal loss actually means in practice. Before ATT, when someone browsed your Shopify store and then opened Instagram, Meta’s system connected those events and could attribute a later purchase to the ad they saw. After ATT, without opt-in, that connection breaks. Meta’s algorithms see a smaller, biased data set: only the users who opted in, who skew toward older demographics and users who actively manage privacy settings.
iOS 17, released in September 2023, added another layer. Safari’s Intelligent Tracking Prevention (ITP) now caps first-party cookies set via JavaScript at 7 days. A user who visits your site and doesn’t convert within that window may not be cookied when they return. Combined with link-tracking protection, which strips UTM parameters from links opened in certain contexts, attribution for Safari users became significantly harder to measure accurately.
WHAT iOS SIGNAL LOSS ACTUALLY AFFECTS
- Audience list size: Website custom audiences in Meta Ads Manager are smaller. Users on iOS who declined ATT are excluded.
- Attribution windows: Meta’s default attribution is now click-based. View-through attribution is less reliable without pixel signal.
- ROAS reporting: Reported ROAS in Ads Manager under-counts conversions from iOS users. Actual results are often better than the dashboard shows.
- 7-day cookie cap (Safari): Return visitors on Safari may not be recognised if they don’t convert within one week of their first visit.
- Not affected: On-platform signals (Meta’s own engagement data, Instagram interactions, video views) are unaffected by ATT and remain usable for audience targeting.
The practical response for most Australian e-commerce brands is to accept that pixel-only tracking is incomplete and layer in additional measurement: Meta’s Conversions API (CAPI), Google’s enhanced conversions, and server-side event tracking. We cover each of these in the sections below.
Google remarketing in 2026
Google remarketing in 2026 runs primarily through two formats: Performance Max campaigns, which use AI to serve ads across Search, Shopping, Display, YouTube, Gmail, and Maps from a single campaign, and traditional Remarketing Lists for Search Ads (RLSA), which layer audience intent onto keyword-based campaigns. Both benefit significantly from Customer Match lists built from your own customer data.
Performance Max replaced Smart Shopping campaigns in September 2022. Rather than managing separate Display or Shopping remarketing campaigns, Performance Max takes a unified approach: you provide asset groups (images, headlines, descriptions, logos, videos), audience signals (your customer lists, website visitors, high-intent audiences), and a budget. Google’s AI then allocates spend across channels in real time to hit your target ROAS or CPA.
For remarketing specifically, audience signals tell Performance Max who to prioritise. Past purchasers and cart abandoners can be uploaded as Customer Match lists. Previous site visitors feed in via the Google Ads tag. The campaign then uses these signals as starting points, finding similar users and optimising toward conversion. The key difference from the old Smart Shopping model: you don’t manually set audiences or exclude segments; the algorithm handles that distribution, learning from your conversion data continuously.
Remarketing Lists for Search Ads (RLSA) remain highly effective for high-intent queries. An RLSA campaign lets you show different ad copy, bid more aggressively, or expand to broader keywords when the searcher is someone who has already visited your site. A user who viewed your checkout page and didn’t complete the purchase is worth a much higher bid on a branded or category search than a first-time visitor.
Customer Match is Google’s first-party data remarketing tool. Upload hashed email addresses from your CRM or email list, and Google matches them to logged-in Google accounts. These lists power remarketing across Search, Shopping, YouTube, Gmail, and Display. Because they’re built from your own data, they’re unaffected by cookie restrictions. Customer Match lists updated regularly outperform cookie-based audiences on recency and match quality.
If you’re running Google Ads and you haven’t yet migrated from Smart Shopping to Performance Max, or you haven’t uploaded a Customer Match list, those are the two highest-priority changes to make to your Google remarketing setup in 2026.
Meta remarketing in 2026
Meta remarketing in 2026 centres on two things: getting accurate event data back into Meta’s system through the Conversions API (CAPI), and using Advantage+ Shopping Campaigns to let Meta’s AI handle audience discovery and retargeting within a single, AI-managed campaign. The Meta Pixel still plays a role, but it’s no longer sufficient on its own.
The Meta Pixel (previously known as the Facebook Pixel; the rebrand followed Meta’s 2021 corporate rename) is a browser-side JavaScript snippet that fires events when users take actions on your website: page views, add-to-cart, initiate checkout, purchase. These events feed Meta’s ad platform and form the basis of your website custom audiences. The problem is that browser-side events are blocked or degraded by iOS ATT restrictions, browser-based ad blockers, and Safari’s ITP. On modern traffic, pixel-only setups lose a significant share of events.
The Meta Conversions API (CAPI) is the server-side complement. Instead of firing events from the user’s browser, CAPI sends event data directly from your web server to Meta’s API. Because the data originates server-side, it bypasses browser restrictions entirely. The best configuration is running CAPI alongside the Meta Pixel, using event deduplication to avoid double-counting: browser events are sent where available, server events fill the gaps where browser signal is lost.
Advantage+ Shopping Campaigns, launched by Meta in 2022, are now the dominant format for e-commerce remarketing on Meta. They replace the old approach of building separate prospecting and retargeting campaigns. In an Advantage+ campaign, Meta’s AI automatically segments your audience across prospecting and retargeting, determining the optimal budget allocation between existing and new customers based on real-time conversion signals. You provide the creative, your product catalogue, and a budget. Meta handles the rest.
For brands that want to retain more control, custom audience retargeting (website visitors, video viewers, Instagram engagers, customer lists) still works within Meta’s standard campaign structure. But for most Australian e-commerce accounts with sufficient conversion volume, Advantage+ Shopping typically outperforms manually segmented retargeting campaigns because its optimisation runs across a larger, unified data set.
Shout Digital manages Meta Ads for growth-stage and established brands across retail, apparel, and consumer goods. If your Meta ROAS has been declining since 2022 and you’re still running traditional retargeting campaigns without CAPI, that’s typically the first place to investigate.
Server-side tagging and first-party data
Server-side tagging moves event tracking from the user’s browser to your own web server, where browser restrictions, ad blockers, and iOS privacy settings have no effect. For brands running paid remarketing at meaningful scale, it’s the single most important infrastructure upgrade available in 2026.
Here’s how it works in practice. With standard browser-side tracking, your website fires a JavaScript tag (the Meta Pixel, a Google Ads tag, a GA4 tag) that sends data directly from the user’s browser to the ad platform’s server. If the user has an ad blocker, is on iOS with ATT declined, or is using Safari with ITP enabled, that tag may fail to fire, or fire with degraded data.
Server-side tagging uses Google Tag Manager’s server-side container (GTM Server-Side), hosted on your own infrastructure. Your website sends events to your server first. Your server then forwards those events to each ad platform’s API: Meta CAPI, Google Ads Enhanced Conversions, GA4, and so on. Because the data never originates in the user’s browser, browser-level restrictions can’t intercept it. First-party cookies set by your server also survive Safari’s ITP restrictions, because server-set cookies are subject to a longer retention window than JavaScript-set cookies.
FIRST-PARTY DATA INFRASTRUCTURE: WHERE TO START
The investment in server-side infrastructure pays back directly in campaign performance: more complete conversion data means better algorithmic optimisation, more accurate audience lists, and more reliable ROAS measurement. Brands that implemented CAPI alongside their pixel typically reported meaningful recovery in attributed conversions within the first reporting cycle after implementation.
Remarketing strategies that still work
The fundamentals of good remarketing haven’t changed: reach the right person with the right message at the right moment in their buying process. What’s changed is how you define “the right person” when your audience signals are less complete. Segmentation, creative relevance, and timing matter more than they did when cookie pools were large enough to absorb imprecision.
Cart abandoners sit at approximately 70% of e-commerce traffic globally (Baymard Institute). They were close enough to buy that they put the item in their cart. A well-timed ad showing the exact product they abandoned, ideally with a time-limited offer or a reminder of free shipping, has consistently been the highest-converting remarketing segment. Suppress this audience from your broader prospecting campaigns so you’re not paying to reach someone already in your cart funnel.
A homepage visitor and someone who spent three minutes on a product page are not the same audience. Build separate segments for each: homepage and blog visitors get brand-awareness creative; product page visitors see the specific product they viewed; checkout abandoners see urgency-led creative. The more precisely your ad reflects where the person stopped, the higher your conversion rate on that segment.
Verified customer reviews embedded in ad creative address the hesitation that stopped someone from converting the first time. A product-page visitor who saw your item but didn’t buy is often sceptical, not indifferent. An ad showing a 4.8-star review from a real customer carries more weight than a promotional headline. This approach works across Meta, Google Display, and TikTok. Source reviews from Trustpilot, Google, or your own platform, and check usage rights before embedding.
Customers who bought from you during a previous sale period or seasonal peak are your highest-quality audience for the equivalent period next year. A customer who bought gardening tools in October 2024 is a stronger seasonal target in October 2025 than any lookalike. Upload these segments as Customer Match lists and serve them ahead of the season, 2 to 4 weeks before your major campaign launches, so they’re already re-engaged before your main promotional push begins.
Past purchasers, and especially high-spend customers, have already demonstrated trust in your brand. Repeat customers convert at a higher rate and lower CPC than cold audiences. Segment your past purchasers by order value or product category and serve them relevant follow-on products, replenishment reminders (if your product has a typical usage cycle), or new product launches. Suppress these audiences from prospecting campaigns to keep messaging relevant and avoid creative fatigue.
For a deeper look at how remarketing fits into a broader e-commerce acquisition strategy, see our guide to improving eCommerce conversion rates with digital marketing.
Frequency caps and creative fatigue
Showing an ad too often is as damaging as not showing it enough. Creative fatigue (where a user has seen the same ad so many times that it produces irritation rather than interest) is one of the most common reasons remarketing campaigns plateau. The right frequency ceiling depends on the platform, the creative format, and how much time has passed since the user’s last interaction with your brand.
On Meta, the platform’s AI delivery systems now manage frequency to some extent automatically, particularly within Advantage+ Shopping campaigns. Meta’s delivery algorithm deprioritises serving ads to users who have already converted or who have seen the same creative repeatedly without engaging. This doesn’t eliminate the need for frequency oversight, but it does mean the system self-corrects more than earlier campaign structures did.
For standard campaigns where you set frequency manually, Meta recommends monitoring frequency score alongside CPM and click-through rate. A rising CPM combined with a falling CTR typically signals that your audience is exhausted and your creative needs rotation. As a practical reference point: a frequency above 3-4 per week for a warm remarketing audience often produces diminishing returns, though the right number varies by product category and creative quality.
On Google, frequency caps for Display and YouTube remarketing can be set at the campaign level in impressions per day, week, or month. Performance Max doesn’t expose direct frequency controls, but you can influence delivery by tightening audience signals and ensuring your creative asset groups are varied enough for Google’s system to rotate rather than repeat the same combination.
The practical solution to creative fatigue is a content calendar for remarketing creative, not just prospecting. Plan quarterly refreshes of your remarketing ad sets: new images, new social proof angles, new offers. A cart-abandoner who has seen your “10% off if you come back” ad six times this week will convert from the seventh impression far less often than from a fresh piece of creative that leads with a new testimonial or a product feature they haven’t seen before.
SIGNS YOUR REMARKETING CREATIVE HAS FATIGUED
- Rising CPM, falling CTR: Your audience has seen the ad before and is ignoring it. Time to rotate creative.
- Frequency above 3-4/week: For warm audiences on Meta, sustained high frequency with no CTR recovery signals exhaustion, not reinforcement.
- Conversion rate declining despite stable reach: The audience is still there, but the message has stopped working. New creative or a new angle is needed.
- Performance Max plateau: If delivery is narrowing to the same asset combinations, add new asset groups to force rotation.
TikTok and Pinterest remarketing
TikTok and Pinterest both support remarketing, and for Australian brands selling to the right demographics, they offer lower competition for warm audiences than Meta and Google. Neither platform has the same scale of first-party data as Meta, but both have matured significantly since 2022 and now provide viable retargeting capabilities.
TikTok remarketing uses the TikTok Pixel (browser-side) and the TikTok Events API (server-side, equivalent to Meta CAPI) to track website events and build custom audiences. Website visitor audiences, product view audiences, and cart abandoner lists are all buildable in TikTok Ads Manager. For brands selling products that suit short-form video (fashion, beauty, consumer electronics, fitness, food), TikTok remarketing generates strong engagement from warm audiences, particularly 18-35 demographics who are heavy TikTok users. Creative format matters acutely here: TikTok remarketing ads that feel native to the feed (UGC-style, no hard sell, genuine product use) outperform polished brand creative by a wide margin.
Pinterest remarketing suits brands in home, lifestyle, fashion, and gifting categories, where Pinterest’s search and save behaviour creates unusually high-intent audiences. A user who saves your product pin is signalling active interest, and Pinterest’s Actalike audiences (equivalent to lookalikes) built from pinners who saved your content are among the most purchase-ready cold audiences on the platform. Remarketing to past website visitors on Pinterest uses the Pinterest Tag and is straightforward to set up. Competition for warm audiences on Pinterest is lower than Meta and Google, which often produces better CPAs for brands in the right categories.
FAQ: remarketing in a privacy-first world
Common questions from marketing managers running paid social and search campaigns in 2026, after the iOS privacy changes and the shift toward AI-managed campaign types.
Updated June 2026. Shout Digital is a Melbourne-based digital marketing agency specialising in Google Ads, Meta Ads, and TikTok Ads for growth-stage and established brands across Australia. For related reading, see our guide to eCommerce conversion rate optimisation and our Remarketing 101 introduction.
We are digital marketing experts
