Google Ads Remarketing Best Practices

  • Google Ads

This page is updated every two months with current best practices for Google Ads remarketing and retargeting. Reconnecting with people who already know you is usually the cheapest conversion you will ever buy, but tightening privacy rules and the fading cookie are reshaping how remarketing audiences are built and used. Each update draws on our own experience plus authoritative industry sources and verified real-time research. Bookmark this page and check back for the latest remarketing best practices. Each update includes worked examples with the arithmetic shown.

Last updated: 25 August 2026

In This Guide

  1. Executive Summary
  2. Benchmarks & Numbers at a Glance
  3. How Remarketing Works
  4. Building Audiences
  5. List Types & Durations
  6. Display & Dynamic Remarketing
  7. RLSA
  8. YouTube & Demand Gen
  9. Funnel Segmentation
  10. Privacy & Cookieless
  11. Measurement & Optimisation
  12. Common Mistakes to Avoid
  13. What Changed Recently
  14. References

Reading this to build your own campaign? Skip the learning curve — we’ll build the whole thing for you for $129, delivered within 24 hours as a spreadsheet you own. See how it works →

1. Executive Summary

Google Ads remarketing in 2026 is a first-party-data discipline, not a cookie-dependent afterthought. The five principles that separate high-performing accounts from wasteful ones are:

  • Segment by intent, not just by visit. Build separate audience lists for each funnel stage — generic visitors, product viewers, cart abandoners, checkout starters, and post-converters — and apply distinct bids, creative, and membership durations to each. A single “all visitors” list is the single biggest structural mistake in remarketing.[1]
  • Set membership durations to match the buying cycle. The 540-day maximum is available but rarely appropriate. Short-cycle purchases (same-day or within a week) need windows of 7–14 days; considered purchases may justify 30–90 days; CRM reactivation campaigns are the only common use case for windows beyond 180 days.[1]
  • Control frequency deliberately. Google Ads does not enforce a universal frequency cap on remarketing. Left uncapped, display and video remarketing will exhaust small audiences quickly, driving up CPA and alienating users. Start at 3–5 impressions per user per day and tighten based on CTR decay and incremental conversion data.[15]
  • Exclude ruthlessly. Converters, employees, support-page visitors, and existing customers (where the campaign goal is acquisition) must be excluded before the campaign launches. Exclusions become more valuable as consent restrictions shrink observable audience pools.[15]
  • Build for the cookieless reality now. Consent Mode, a properly configured CMP, Customer Match, and GA4-based audiences are the durable infrastructure. Campaigns that depend on unconsented third-party cookies are already degraded in Australia and will degrade further. First-party list quality is the compounding asset.[11][3]

2. Benchmarks and Numbers at a Glance

Metric Typical range or threshold Applies when Source
Minimum active users for Display and YouTube audience eligibility 100 active users in the last 30 days All remarketing lists; lists below this threshold will not serve [1]
Minimum active users for Search Network remarketing eligibility 100 active users in the last 30 days Lists uploaded or refreshed after 1 February 2024 [1]
Default audience membership duration 30 days Applies at list creation unless manually overridden [1]
Maximum audience membership duration 540 days Search and Display remarketing lists; timer resets on re-visit [1]
Median CTR for Google Display remarketing 0.82% Vendor claim across Display remarketing placements; not a controlled study [63]
Median CPA for Google Display remarketing USD $42 (approx. AUD $65 at 1.55 exchange rate) Vendor claim; varies widely by industry and bid strategy [63]
Conversion rate uplift — RLSA vs standard search 161% higher conversion rate Published RLSA study; treat as directional, not a universal benchmark [62]
Conversion rate uplift — dynamic remarketing vs regular display ads 203% higher conversion rate Published dynamic remarketing study; single-account data, treat directionally [67]
Conversion rate uplift — dynamic remarketing vs standard remarketing 119% higher conversion rate Published dynamic remarketing study; single-account data, treat directionally [67]
View-through conversion rate — dynamic vs standard remarketing 3× more likely to generate a view-through conversion Published dynamic remarketing study; single-account data, treat directionally [67]
CPA reduction — RLSA segmented audiences vs broad search 36% lower CPA Published RLSA study; treat as directional [71]
CTR uplift — RLSA vs regular search traffic 130% higher CTR Published RLSA study; treat as directional [71]
Recommended starting frequency cap — Display remarketing 3–5 impressions per user per day Practitioner guidance; tighten if CTR declines or CPA rises [15]
Minimum budget for Demand Gen tCPA campaigns 15× target CPA per day Google vendor guidance for Demand Gen campaigns using Target CPA bidding [10]

3. How Remarketing Works

Remarketing is the practice of identifying users who have previously interacted with your brand — via your website, app, YouTube channel, or CRM data — and serving them targeted ads as they continue using Google’s properties. The mechanism differs by data source but follows a common pattern: a signal is collected, a user is added to an audience list, and that list is used to modify bids, restrict targeting, or personalise creative in active campaigns.[6]

When a user visits a tagged page, the Google tag (or GA4 tag with audience export enabled) writes a cookie or device identifier that associates that user with one or more audience lists. If the user consents and returns within the membership window, they are eligible to be served an ad in that list’s associated campaigns. When consent is denied, Google’s Consent Mode can send cookieless pings for modelling purposes, but those signals cannot be used to build remarketing lists or identify individual users.[3]

In Google’s current UI, the older term “remarketing” has largely been replaced by “your data” segments and “audience segments” in campaign reporting and audience management pages. The underlying mechanics are unchanged, but practitioners auditing accounts should use the updated terminology when navigating the Audiences tab.[1]

New audience lists can take 48 to 72 hours to fully populate after the tag is deployed, and lists must reach the minimum threshold of 100 active users in the last 30 days before they are eligible to serve on the Display Network or YouTube.[1][10] For Search Network RLSA, the same 100-user threshold applies to lists created or refreshed after 1 February 2024.[1]

Worked example

Checking whether a new cart-abandoner list is ready to serve

  • Setup: An Australian home-goods e-commerce account spending $8,000 per month launches a cart-abandoner remarketing list on 1 September 2026. The Google tag is correctly deployed on the /cart and /checkout pages.
  • Numbers: The site receives approximately 4,200 sessions per month to the cart page. At a 4% add-to-cart rate that results in exits before purchase, the cart-abandoner pool is roughly 168 unique users per month, or approximately 5–6 new users per day. At that rate, the list reaches the 100-user minimum threshold in approximately 17–18 days — around 18 September 2026. Before that date, the list is ineligible to serve on Display or YouTube.[1]
  • Decision: Pause the cart-abandoner ad group until 19 September 2026, and set a calendar reminder to confirm list size in the Audiences tab on that date before enabling the ad group.
  • Why: Google requires 100 active users in the last 30 days before an audience list is eligible to serve; activating the ad group before this threshold is met results in zero impressions and wasted time.[1]

4. Building Remarketing Audiences

The quality of a remarketing programme is determined almost entirely by the quality and specificity of its audience lists. A single “all visitors” list is the lowest-value starting point; a properly segmented set of lists — each representing a distinct intent signal — enables precise bidding, relevant creative, and meaningful exclusions.[6][13]

Recommended data sources

  • Website visitors via Google tag or GA4: The most common and flexible source. GA4 audience export allows behavioural conditions (event sequences, page depth, session duration) that the native Google Ads tag cannot replicate without custom parameters.[13]
  • App users: App analytics or Google Play-linked audiences enable re-engagement of lapsed users, in-app purchasers, and users who have not completed key in-app actions.[6]
  • YouTube viewers: Audiences built from channel subscribers, video views (25%, 50%, 75%, 100% completion), and ad interactions provide warm remarketing pools for brands with active video content.[6]
  • Customer Match: Upload hashed email addresses, phone numbers, or postal addresses from your CRM to reach existing customers across Search, YouTube, Gmail, and Display. As third-party cookies diminish, Customer Match becomes progressively more important as a durable first-party signal.[11]

Recommended segmentation structure

  • Segment 1 — Generic site visitors: All users who visited any page, excluding those captured in deeper segments. Useful for awareness-stage messaging.
  • Segment 2 — Category or product viewers: Users who viewed a product page, category page, or service detail page without progressing further.
  • Segment 3 — High-intent abandoners: Users who added to cart, started a checkout, viewed a pricing page, or began a lead form but did not convert.
  • Segment 4 — Converters: Users who completed a purchase, submitted a lead form, or completed any defined macro-conversion. This list is primarily used for exclusion from acquisition campaigns and for inclusion in retention and cross-sell campaigns.[15]
  • Segment 5 — Lapsed customers (CRM-based): Customer Match uploads of users who purchased 90–365 days ago and have not returned. Use for reactivation campaigns with longer membership windows.[11]

Worked example

Building a five-tier audience structure for a B2B software account

  • Setup: A Melbourne-based B2B SaaS account with a 45-day average sales cycle and a monthly Google Ads budget of $12,000. The site has a pricing page, a free-trial sign-up form, and a contact-sales form.
  • Numbers: The account generates approximately 3,800 unique monthly sessions. Of these: 900 visit the pricing page (23.7%), 280 start the trial form (7.4%), 95 complete the trial sign-up (2.5%), and 40 submit the contact-sales form (1.1%). Each segment is therefore large enough to meet the 100-user minimum within 30 days for Display, and the pricing-page list (900/month) will be immediately eligible at launch.[1]
  • Decision: Create five GA4 audiences: (1) all users, 30-day window; (2) pricing-page visitors excluding sign-ups, 45-day window to match the sales cycle; (3) trial-form starters who did not complete, 14-day window; (4) trial completers and contact-sales submitters — exclusion list only; (5) CRM Customer Match upload of contacts inactive for 90+ days, 180-day window. Apply each as a separate audience in Google Ads with observation mode and distinct bid adjustments.[11][13]
  • Why: Segmenting by intent signal rather than using one catch-all list allows distinct bids and creative for each funnel stage, and keeps the 45-day sales cycle reflected in the membership windows rather than defaulting to the 30-day platform default.[1]

5. Audience List Types and Membership Durations

Setting membership duration is one of the most consequential and most frequently mishandled configuration decisions in remarketing. The default of 30 days is appropriate for some accounts and completely wrong for others. The maximum of 540 days should be reserved for specific reactivation use cases, not applied universally because it is the highest available number.[1]

Membership duration by funnel stage

Audience segment Recommended duration Rationale
Checkout or form abandoners 7–14 days Purchase intent is hot but decays rapidly; messaging beyond 14 days has diminishing returns for most categories
Product or pricing page visitors 14–30 days Consideration stage; intent is present but not urgent
General site visitors 30 days Matches platform default; appropriate for broad awareness follow-up
Category browsers (considered purchase) 30–90 days Long-cycle purchases such as vehicles, real estate, or enterprise software benefit from longer windows
YouTube video viewers (25–50% completion) 30 days Engagement signal is warm but not purchase-intent; avoid over-investing
Lapsed customers — CRM reactivation 90–180 days Longer window justified by relationship history; pair with Customer Match
High-value CRM segments — annual repurchase Up to 540 days Appropriate only when the repurchase cycle genuinely spans 12+ months

Note that the membership duration timer resets each time a user revisits the site or app and triggers the tag again, so a 30-day window for an active browser can extend indefinitely in practice.[1]

List type comparison

List type Best use case Minimum size (Display/YouTube) Primary data source
Website visitor list Funnel follow-up, cart recovery, RLSA 100 active users / 30 days[1] Google tag or GA4
App user list Re-engagement, churn prevention 100 active users / 30 days[1] App analytics / Google Play
YouTube engagement list Warm video audience follow-up 100 active users / 30 days[1] YouTube channel / video interactions
Customer Match list CRM retention, cross-sell, reactivation 100 active users / 30 days[1] CRM export (hashed email, phone, address)

Worked example

Choosing the right membership duration for a furniture retailer

  • Setup: A Brisbane furniture retailer with an average order value of $1,800 AUD and a typical research-to-purchase cycle of 42 days based on GA4 path analysis. The account spends $9,500 per month on Google Ads. The current remarketing setup uses one list — “all visitors” — at the 30-day default.
  • Numbers: GA4 data shows 68% of converters visited the site at least twice before purchasing, with an average of 26 days between first visit and conversion. Setting the product-page viewer list to 14 days (the current setting) means 26 − 14 = 12 days of the typical consideration window are not being covered. Extending to 45 days covers the full 42-day cycle plus a 3-day buffer. The site generates approximately 1,200 product-page views per month, so the list will remain above the 100-user eligibility threshold at all times.[1]
  • Decision: Change the product-page viewer list membership duration from 30 days to 45 days in the Audience Manager settings. Keep the checkout-abandoner list at 7 days. Create a new CRM-based Customer Match list of purchasers from 2025 and set it to 180 days for a reactivation campaign targeting a second purchase.
  • Why: Membership duration should match the actual buying cycle, not the platform default; a 42-day research window requires a minimum 42-day list to remain eligible to serve to users throughout their decision process.[1]

6. Display and Dynamic Remarketing

Standard Display remarketing serves image or responsive ads to previous site visitors as they browse the Google Display Network. Dynamic remarketing extends this by automatically pulling product or service details from a data feed and displaying the specific item a user previously viewed, generating personalised ad creative at scale.[9][23]

Standard Display remarketing — configuration checklist

  • Segment campaigns by audience tier: do not combine high-intent abandoners with generic site visitors in a single campaign or ad group.[13]
  • Use responsive display ads with at least 5 image assets, 5 headline variations, and 5 description variations so the system can optimise creative combinations across placements.[13]
  • Set frequency caps at the campaign level: begin at 3 impressions per user per day and adjust based on CTR and CPA trends after 14 days of data.[15]
  • Exclude converters from all acquisition-focused campaigns before launch, not as an afterthought.[15]
  • Use Target CPA bidding for conversion-focused Display remarketing campaigns once at least 30 conversions have accumulated in the past 30 days in the campaign or account.[13]

Dynamic remarketing — additional requirements

  • A product or service feed must be linked in Google Merchant Center (for retail) or directly in Google Ads (for other verticals including travel, real estate, and education).[9][23]
  • The Google tag must fire product ID, page type, and value parameters on all relevant pages so users are matched correctly to feed items.[9]
  • Dynamic remarketing consistently outperforms standard remarketing at the conversion stage: published data shows dynamic remarketing achieving a 203% higher conversion rate than regular display ads and a 119% higher conversion rate than standard remarketing.[67] Note these figures come from a single published study and should be treated as directional rather than universal benchmarks.
  • Dynamic remarketing also generates 3 times more view-through conversions than standard remarketing in the same published data set.[67]

Worked example

Launching dynamic remarketing for a sporting goods retailer

  • Setup: A Sydney sporting goods e-commerce retailer with 4,500 active SKUs, spending $15,000 per month on Google Ads across Search and Display. The existing Display remarketing campaign uses a single “all visitors” list with generic creative and is generating a CPA of AUD $95 against a target of AUD $70.
  • Numbers: The site receives 18,000 sessions per month. Of these, 6,300 (35%) view a product page without purchasing. At a median Display remarketing CTR of 0.82%[63], the dynamic remarketing campaign at the same impression volume as the current campaign would be expected to generate materially higher click volume due to personalised creative relevance. If the account’s current Display CPA of $95 is reduced toward the published benchmark of AUD $65 (converted from the USD $42 median[63]), the monthly saving on 158 conversions (calculated as $15,000 ÷ $95) would be approximately ($95 − $65) × 158 = $4,740 per month in CPA reduction, assuming flat conversion volume.
  • Decision: Enable dynamic remarketing by linking the existing Google Merchant Center feed to Google Ads, verify that the Google tag fires retailer_id, pagetype, and ecomm_prodid parameters on all product, cart, and purchase pages, then create a separate dynamic remarketing campaign with a Target CPA of AUD $70, a 7-day membership window for cart abandoners, and a 30-day window for product-page viewers. Set a frequency cap of 3 impressions per user per day.
  • Why: Dynamic remarketing showing the exact product a user viewed delivers a 119% higher conversion rate than standard remarketing[67], making it the correct tool for a product-heavy e-commerce account where the generic creative is suppressing performance.

7. Remarketing Lists for Search Ads (RLSA)

RLSA allows you to modify search bids, ads, or keyword eligibility based on whether a user searching on Google is already on one of your remarketing lists. Unlike Display remarketing, RLSA does not interrupt the user — it adjusts your presence in auctions the user is already participating in, making it one of the most efficient remarketing tools available.[1]

Two operating modes

  • Observation (recommended starting point): The campaign serves to all eligible searchers, but bid adjustments are applied to users who are also on your remarketing lists. You gain performance data without restricting reach.[11]
  • Targeting: The campaign serves only to searchers who are also on your remarketing lists. Use this only for dedicated RLSA campaigns where the intent is explicitly to reach returning users.[11]

RLSA best practices

  • Apply RLSA lists to competitive, expensive keywords where return visitors convert at a materially higher rate than cold searchers. Published data shows RLSA audiences achieving a 130% higher CTR and 36% lower CPA than standard search traffic for the same terms.[71]
  • Start all lists in Observation mode. After accumulating at least 100 conversions across the observed audience (or 30 days of data, whichever is longer), apply bid adjustments of +15% to +40% for high-intent segments and negative adjustments of −20% to −40% for low-value segments.[11]
  • Align RLSA bid increases with audience recency: a user who abandoned checkout 2 days ago should attract a higher bid increase than a user who visited the homepage 25 days ago.[1]
  • RLSA conversion rate uplift of 161% vs standard search campaigns has been published[62], but this figure comes from a single published study and should be used directionally when setting bid adjustment hypotheses, not as a guaranteed outcome.

Worked example

Applying RLSA bid adjustments to a high-CPC legal services campaign

  • Setup: A Melbourne personal injury law firm spending $22,000 per month on Search, with an average CPC of $38 AUD on core terms such as “compensation claim lawyer” and “personal injury solicitor Melbourne”. The account currently has no RLSA configuration. The website receives 1,400 unique monthly sessions, of which 310 visit the “How it works” or “Claim calculator” pages — a strong intent signal.
  • Numbers: The 310 “how it works” visitors represent 22.1% of all sessions. At 1,400 sessions per month, this list will reach the 100-user minimum in approximately 10 days (310 ÷ 31 days × 10 days ≈ 100 users).[1] Current account-wide search conversion rate is 4.2%. If the RLSA segment converts at a rate consistent with the published 161% uplift[62], the expected conversion rate for this segment would be 4.2% × 2.61 = 10.96%. At a CPC of $38, each conversion from the RLSA segment would cost $38 ÷ 10.96% = AUD $347 per conversion, versus the current $38 ÷ 4.2% = AUD $905 per conversion from cold traffic — a potential CPA reduction of $558 per conversion at equivalent CPC.
  • Decision: Add the “how it works / claim calculator page visitors” list to the core Search campaign in Observation mode with a +30% bid adjustment. Review after 30 days or 50 RLSA-attributed clicks (whichever comes first) and adjust the bid modifier based on observed conversion rate relative to cold traffic.
  • Why: RLSA lists on high-CPC legal terms concentrate budget on returning, higher-intent users who have already demonstrated research behaviour, consistent with published data showing RLSA audiences delivering 130% higher CTR and 36% lower CPA than standard search traffic.[71]

8. YouTube and Demand Gen Remarketing

YouTube remarketing and Demand Gen campaigns represent the highest-reach remarketing surface in Google’s ecosystem, covering YouTube (in-stream, in-feed, Shorts), Gmail, and Google Discover. The audience rules are the same — 100 active users in the last 30 days — but the creative, bidding, and structural requirements differ meaningfully from Display and Search.[10]

YouTube remarketing audiences

YouTube engagement audiences are built from: video views at specified completion thresholds (25%, 50%, 75%, 100%), channel subscriptions, likes, and ad interactions. These audiences are warm — users have already engaged with brand content — making them appropriate for consideration-stage or conversion-stage messaging with a clear next step.[6]

Demand Gen campaign structure for remarketing

  • Consolidate similar audience themes into the same ad groups rather than fragmenting by minor audience variations. Fragmentation limits the system’s ability to learn efficiently.[6]
  • Provide a mix of horizontal video (16:9), vertical video (9:16), and image assets in each ad group so the campaign can serve across all available placements.[6][7]
  • Follow ABCD creative principles: Attention in the first 5 seconds, Brand integration early and persistently, Connection via voiceover and on-screen text, Direction with a clear CTA reinforced in both audio and visual channels.[2]
  • For Demand Gen campaigns targeting conversions, start with Maximise Conversions bidding. Transition to Target CPA only after sufficient conversion volume is available — Google’s guidance recommends a minimum of 4–6 weeks of learning before evaluating tCPA performance.[15]
  • Set a daily budget of at least 15× your Target CPA for Demand Gen campaigns using tCPA bidding. For a tCPA of AUD $80, this means a minimum daily budget of AUD $1,200.[10]
  • Avoid making frequent bid changes during the learning period. Changes reset the learning phase and extend the time before stable performance is achievable.[6]

Worked example

Setting up a Demand Gen remarketing campaign for a national education provider

  • Setup: A national online education provider offering postgraduate courses at AUD $4,500 average enrolment value, spending $30,000 per month on Google Ads. The account has an existing YouTube channel with 12,000 subscribers and 85,000 monthly video views. The target CPA for a course enquiry form submission is AUD $90.
  • Numbers: Applying the 15× budget rule: minimum daily budget = AUD $90 × 15 = AUD $1,350 per day, or AUD $40,500 per month.[10] The existing $30,000 monthly budget is below this threshold by $10,500. At the current budget, tCPA bidding is not recommended; the account should run Maximise Conversions bidding until monthly spend is at least $40,500, or the tCPA target is reduced. Reducing the tCPA to AUD $65 would bring the minimum daily budget to $65 × 15 = $975/day = $29,250/month, which is within the $30,000 budget. The YouTube 75%-completion viewer list has approximately 8,500 active users per month — well above the 100-user minimum.[1]
  • Decision: Launch the Demand Gen remarketing campaign targeting 75%-completion YouTube video viewers with Maximise Conversions bidding and a daily budget of $1,000 (AUD). After 6 weeks and a minimum of 50 form-submission conversions attributed to the campaign, evaluate transition to tCPA at AUD $90, contingent on increasing the daily budget to AUD $1,350.
  • Why: Google’s guidance requires a daily budget of at least 15× tCPA for Demand Gen tCPA campaigns[10]; running tCPA at an underfunded daily budget prevents the algorithm from exiting the learning phase and produces unstable CPA outcomes.

9. Funnel Segmentation and Frequency

Funnel segmentation and frequency management are the two levers that determine whether remarketing adds incremental value or simply annoys the same people repeatedly until they convert through another channel and claim the credit. Both require deliberate decisions, not platform defaults.[15][1]

Segmenting by funnel position

  • Top of funnel (awareness follow-up): Generic site visitors, content readers, YouTube viewers at 25% completion. Use light frequency (1–2 impressions/day), broad creative, and long membership windows (30 days). Bid conservatively — these users are discovering, not deciding.
  • Mid-funnel (consideration): Product viewers, category browsers, repeat visitors, video viewers at 75% completion. Use moderate frequency (2–3 impressions/day), benefit-focused creative, and 14–30 day windows. Bid at market rate or with a modest upward adjustment based on observed CVR data.
  • Bottom of funnel (high intent): Pricing-page visitors, cart abandoners, checkout starters, form starters. Use tight frequency (3–5 impressions/day maximum), specific and urgent creative, and 7–14 day windows. Bid aggressively — published RLSA data shows this segment can convert at 161% above standard search rates.[62]
  • Post-conversion: Exclude from acquisition campaigns immediately. Activate in dedicated cross-sell or retention campaigns with separate messaging and a distinct budget.[15]

Frequency management principles

  • Google Ads does not enforce a universal frequency cap for remarketing campaigns. Caps must be set manually at the campaign or ad group level.[15]
  • The recommended starting point is 3–5 impressions per user per day for Display remarketing.[15] YouTube and Demand Gen frequency should be set separately, typically lower, given the higher per-impression cost and impact of video ad fatigue.
  • Monitor CTR, conversion rate, and reach on a weekly basis. When CTR declines by more than 20% relative to the campaign’s first two weeks while frequency is stable or rising, reduce the daily cap by 1 impression and test for 7 days.
  • Small remarketing pools (100–300 active users) exhaust quickly at high frequency caps. For audiences under 300 users, cap at 2 impressions per user per day to preserve reach duration across the membership window.

Worked example

Diagnosing frequency fatigue in a cart-abandoner campaign

  • Setup: A Perth fashion retailer running a Display remarketing campaign targeting cart abandoners. The audience has 420 active users. The campaign has no frequency cap set and is running a daily budget of AUD $250. The campaign launched on 1 October 2026.
  • Numbers: By 14 October 2026, the campaign had delivered 8,820 impressions to the 420-user audience — an average of 8,820 ÷ 420 ÷ 14 days = 1.5 impressions per user per day. CTR in week 1 was 1.1% (97 clicks). CTR in week 2 dropped to 0.6% (53 clicks) with the same impression volume — a 45% CTR decline. The conversion rate on landing-page visits remained stable at 3.8%, confirming that creative fatigue rather than landing-page quality is driving the decline. At 0.6% CTR and AUD $1.20 average CPC, the campaign is delivering 53 clicks × 3.8% = 2.01 conversions per week at a cost of 53 × $1.20 = $63.60 per week — a CPA of $31.64 per conversion. In week 1 at 1.1% CTR, the CPA was 97 × $1.20 ÷ (97 × 3.8%) = $31.58 — nearly identical CPA but substantially lower incremental reach efficiency as the pool saturates.
  • Decision: Set a frequency cap of 3 impressions per user per day at the campaign level, introduce 3 new responsive display ad creative sets with different headlines and images, and rotate creative weekly to slow audience saturation. Review CTR after 7 days.
  • Why: A 45% CTR decline from week 1 to week 2 with stable conversion rate signals creative fatigue from a small audience being over-served; the recommended corrective action is capping at 3–5 impressions per user per day and rotating creative.[15]

Worked example

Structuring a three-tier funnel remarketing campaign for a travel agency

  • Setup: A Gold Coast travel agency selling international holiday packages at an average booking value of AUD $6,200, with a consideration cycle of 60–90 days. Monthly Google Ads budget is $18,000.
  • Numbers: The site generates 5,500 monthly sessions. Segment breakdown: (1) all visitors = 5,500, membership window 30 days, frequency cap 2/day; (2) destination-page viewers (not enquired) = 1,800 users/month, window 60 days, frequency cap 3/day; (3) enquiry-form starters who did not submit = 340 users/month, window 14 days, frequency cap 5/day. All three segments exceed the 100-user minimum.[1] Allocate budget as follows: 20% ($3,600/month) to Tier 1, 40% ($7,200/month) to Tier 2, 40% ($7,200/month) to Tier 3. Tier 3 receives disproportionate budget because enquiry-form starters represent 6.2% of sessions but the highest purchase intent.
  • Decision: Create three separate Display remarketing campaigns — one per tier — with distinct creative (awareness messaging for Tier 1, destination-specific imagery for Tier 2, urgency and limited-availability messaging for Tier 3), separate frequency caps as above, and distinct membership windows. Exclude Tier 3 completers (enquiry submitters) from Tiers 1 and 2.
  • Why: Allocating budget proportional to purchase intent rather than audience size maximises the return on the high-CPA consideration segment, and separate campaign structures prevent the algorithm from conflating low-intent and high-intent user signals.[1][15]

10. Privacy, Consent and the Cookieless Future

Privacy regulation and Google’s evolving consent framework are not peripheral concerns for remarketing practitioners — they directly determine how large your observable audience pools are, which signals you can legally collect, and which measurement methods produce reliable data. In Australia, the Privacy Act 1988 and its current reform trajectory require advertisers to treat consent as a genuine user choice, not a dark-pattern checkbox.[3]

Google Consent Mode: how it affects remarketing

Google’s Consent Mode framework allows tags to send signals to Google based on the consent status the user has provided via a CMP. When a user grants ad personalisation consent, standard cookie-based remarketing operates normally. When consent is denied, Google’s tags send cookieless pings — anonymous, aggregated signals used for conversion modelling — but these pings are explicitly not used to build remarketing lists, track users across sites, or generate individual user profiles.[3]

The practical consequence is that in markets with high consent-denial rates (typically 30–50% of users in regulated markets), observable remarketing pools are materially smaller than total site traffic suggests. An account with 10,000 monthly sessions and a 40% consent-denial rate has an effective observable remarketing pool of approximately 6,000 users — and some of those may still fall below the 100-user minimum for individual list segments.[1][3]

Consent Mode implementation checklist

  • Deploy a compliant CMP (Consent Management Platform) that is integrated with Google’s Consent Mode API.[3]
  • Configure the CMP to send ad_storage and ad_user_data signals to Google tags in real time based on user choice.[3]
  • Default to denied consent where required by Australian or applicable international privacy law, then update to granted only after affirmative user action.[3]
  • Test both consent-granted and consent-denied tag states after any site or tag change using a tag validation tool. Verify that cookieless pings fire in denied state and that no remarketing cookies are written.[3]
  • For Customer Match: collect email addresses and other personal data only where a clear, documented consent to use that data for advertising personalisation exists. This is a legal requirement, not a preference.[11]

Planning for smaller observable audiences

  • Grow first-party list quality: invest in CRM capture, email sign-ups, lead magnets, and loyalty programmes so Customer Match lists expand as cookie-based lists shrink.[11]
  • Prioritise GA4 audience export over native Google Ads tag audiences where possible, as GA4’s modelled measurement can partially compensate for consent gaps in reporting even when it cannot recover individual user-level targeting.[13]
  • Use aggregate audience performance data (segment-level CVR, ROAS, CPA) rather than individual-level attribution when consented pools are small. Model-based reporting in GA4 and Google Ads will be the primary measurement lens for a growing proportion of traffic.[3]

Worked example

Estimating the impact of consent-denial rate on a remarketing audience

  • Setup: An Adelaide health and wellness e-commerce account with 7,200 monthly sessions and a remarketing programme built entirely on website visitor lists. The account recently deployed a compliant CMP in July 2026, which has resulted in a measured consent-denial rate of 38%.
  • Numbers: Pre-CMP, the “pricing and product page visitors” list contained approximately 2,100 active users per month (29.2% of 7,200 sessions visiting those pages). Post-CMP, with 38% consent denied: 2,100 × (1 − 0.38) = 1,302 observable users per month. The list remains above the 100-user minimum[1] so it continues to serve, but the addressable pool has shrunk by 798 users (38%). At the previous median Display remarketing CTR of 0.82%[63], the expected monthly clicks from this list fall from 2,100 × 0.82% = 17.2 to 1,302 × 0.82% = 10.7 — a reduction of approximately 6.5 clicks per month from this segment. The checkout-abandoner list (previously 310 users) falls to 310 × 0.62 = 192 users — still above the 100-user minimum but with reduced margin.
  • Decision: Increase Customer Match list size by exporting the last 12 months of purchaser emails (2,400 addresses) from the CRM, hashing and uploading via the Audience Manager, and activating a Customer Match-based retention campaign as a parallel audience layer. This provides a consent-independent first-party audience for returning purchasers that does not depend on cookie acceptance.[11]
  • Why: As consent-denial rates reduce observable cookie-based audiences, Customer Match uploads of consented CRM data are the primary mechanism for maintaining first-party remarketing reach; the 38% consent-denial rate makes this shift from reactive to proactive a current priority rather than a future consideration.[11][3]

11. Measurement and Optimisation

Measuring remarketing performance correctly requires separating the contribution of remarketing from the natural behaviour of returning users who would have converted regardless of seeing an ad. Standard last-click attribution dramatically overstates remarketing value; data-driven attribution (DDA) in Google Ads is the minimum acceptable measurement approach for accounts with sufficient conversion volume.[13]

Attribution and reporting

  • Use data-driven attribution in Google Ads for all remarketing campaigns where the account generates at least 300 conversions per month. DDA distributes credit across touchpoints based on observed incremental contribution rather than assigning 100% credit to the last click.[13]
  • Review performance in the Audiences tab (Google’s current terminology: “your data segments”) at the audience segment level, not just at the campaign or ad group level. Segment-level reporting reveals which recency windows and audience tiers are generating incremental conversions.[1]
  • Use view-through conversion windows conservatively: a 1-day view-through window is appropriate for most Display remarketing. Longer view-through windows inflate credited conversions by capturing organic behaviour.[13]
  • For GA4-integrated reporting, use the Advertising workspace — Attribution path analysis to see whether remarketing campaigns are appearing as first touch, assist, or last touch across the conversion path.[13]

Optimisation workflow

  • Review audience segment performance every 14 days minimum. Compare CVR, CPA, and ROAS by audience list, not just by campaign.
  • Apply bid adjustments based on observed CVR differences between audience segments: if the checkout-abandoner segment converts at 8.4% and the generic visitor segment converts at 1.2%, the checkout abandoner justifies a bid adjustment of up to +600% relative to the generic visitor in observation mode.[11][9]
  • Pause or exclude audience segments that have accumulated at least 500 impressions and 50 clicks without a single conversion. Do not let underperforming segments consume budget indefinitely.
  • Monitor frequency vs incremental conversions weekly. If frequency increases by 1 impression/user/day without a corresponding increase in weekly conversion volume, reduce the cap.

Worked example

Calculating a bid adjustment from segment-level CVR data

  • Setup: A Canberra online accounting software account using observation mode RLSA on its core brand and non-brand search campaigns. After 30 days, the Audiences tab shows the following data: generic site visitors (30-day window) — 1,840 clicks, 47 conversions, CVR 2.55%; pricing-page visitors (14-day window) — 312 clicks, 38 conversions, CVR 12.18%.
  • Numbers: The pricing-page visitor segment converts at 12.18% vs 2.55% for generic visitors — a ratio of 12.18 ÷ 2.55 = 4.78×. A proportional bid adjustment would be +378% (4.78 − 1 = 3.78 × 100). However, Google Ads caps bid adjustments at +900%, and applying the full 378% increase may push CPCs above the account’s target CPA of AUD $55. At a current average CPC of AUD $4.20 for the pricing-page segment and a CVR of 12.18%, the current CPA is $4.20 ÷ 12.18% = AUD $34.48. Applying a +100% bid adjustment raises the effective CPC to AUD $8.40, producing a CPA of $8.40 ÷ 12.18% = AUD $68.97 — above the $55 target. The maximum bid adjustment that keeps CPA at or below $55 is: ($55 × 12.18%) − $4.20 = $6.70 − $4.20 = $2.50 additional CPC, representing a +59.5% bid adjustment, rounded to +60%.
  • Decision: Set a +60% bid adjustment for the pricing-page visitors RLSA segment in the campaign audience settings. Review after 14 days and 30+ conversions from this segment.
  • Why: The bid adjustment is calculated from the segment’s observed CVR and the account’s AUD $55 target CPA, not from a generic rule of thumb; applying the maximum permissible bid increase that keeps CPA on target maximises volume without breaching the efficiency threshold.[11][9]

12. Common Mistakes to Avoid

The following errors are consistently observed across remarketing accounts and are responsible for the majority of wasted spend and missed conversion opportunity in this channel.

Structural mistakes

  • Using a single “all visitors” audience list. This combines users with fundamentally different intent levels into one targeting pool, preventing differentiated bidding and messaging. Build a minimum of three funnel tiers.[6][13]
  • Setting membership durations to the platform maximum (540 days) by default. A 540-day window makes sense for annual repurchase categories only. Applying it to checkout abandoners means serving urgent-tone ads to users who made a purchase decision 11 months ago.[1]
  • Failing to exclude converters from acquisition campaigns. This is the most common and most expensive error: spending budget to re-target users who already completed the goal, while simultaneously inflating attributed conversion numbers via view-through credit.[15]
  • Launching a remarketing campaign before the audience list reaches 100 active users. The campaign will receive zero impressions but consume management time and generate misleading “no data” signals that prompt unnecessary changes.[1]

Bidding and budget mistakes

  • Running Demand Gen tCPA campaigns without meeting the 15× daily budget requirement. A tCPA of AUD $80 requires a daily budget of at least AUD $1,200. Underfunding extends the learning phase indefinitely and produces an unstable CPA.[10]
  • Applying blanket bid increases to all remarketing users equally. A generic +30% bid adjustment for “all site visitors” ignores the 10-fold CVR difference that typically exists between generic visitors and checkout abandoners.[1]
  • Making frequent bid changes during the learning phase. Each significant bid change resets the learning phase for automated bidding strategies, compounding the instability.[6]

Creative and frequency mistakes

  • Serving the same creative to all remarketing audiences regardless of funnel stage. A generic brand awareness ad served to a checkout abandoner 24 hours after they left the cart is a missed conversion opportunity.[2]
  • Not setting a frequency cap. Without a cap, small remarketing pools (under 500 users) will be over-served within days, producing CTR decay, wasted spend, and potential brand damage.[15]
  • Ignoring creative fatigue signals. A week-on-week CTR decline of 20% or more while frequency holds steady is a signal that creative rotation is needed, not that the audience is exhausted.[15]

Privacy and compliance mistakes

  • Uploading Customer Match data without verified consent. Uploading personal data for ad targeting without documented consent to advertising use is both a Google Ads policy violation and a potential breach of the Australian Privacy Act.[11]
  • Not testing Consent Mode in the denied state. A CMP that fires Google tags in the denied state — writing remarketing cookies without consent — is a compliance failure that may not be visible without explicit QA testing.[3]

Worked example

Diagnosing and correcting a converter-exclusion failure

  • Setup: A Brisbane subscription software account running Display remarketing with a single “all visitors” list, 30-day membership window, and no exclusions. The account reports 220 Display remarketing conversions in September 2026 at an average CPA of AUD $28. The account uses a 7-day view-through conversion window.
  • Numbers: The account generates 180 actual subscription sign-ups per month tracked via the Thank You page. Yet the Display remarketing campaign reports 220 conversions — 40 more than the actual sign-up count. At AUD $28 CPA, the 220 attributed conversions represent $6,160 in reported spend justification. However, if 40 of those conversions are view-through conversions attributed to users who converted via Search or Direct after simply seeing (not clicking) a Display ad, the actual CPA for genuine Display-influenced conversions is $6,160 ÷ 180 = AUD $34.22 — 22.2% higher than reported. Additionally, converters remaining in the “all visitors” list are being re-targeted with acquisition messaging for 30 days post-conversion, consuming an estimated (220 attributed conversions ÷ 5,500 total remarketing users) = 4% of the audience pool in converter re-targeting.
  • Decision: Create a “September–October 2026 Purchasers” exclusion list based on the Thank You page URL, apply it as an audience exclusion to all Display remarketing campaigns, reduce the view-through conversion window from 7 days to 1 day, and rebuild the audience structure with three distinct lists as described in Section 9.
  • Why: Excluding converters and tightening the view-through window prevents inflated attribution and stops budget being consumed re-targeting users who have already completed the conversion goal.[15][13]

13. What Changed Recently (Last 30 Days)

The following changes and clarifications have been confirmed in official Google Ads documentation as of August 2026. Third-party commentary suggesting broader policy changes beyond these verified items has not been included, as those claims could not be confirmed in official sources.[1]

Confirmed recent changes

  • Audience terminology updated across the UI: Google Ads now uses “your data” in place of “remarketing” and “audience segments” in place of the older “audience types” language throughout campaign reporting and the Audience Manager interface. Practitioners auditing campaigns should navigate to the Audiences tab using the new terminology; the underlying mechanics are unchanged.[1]
  • Improved Audiences reporting page: The Audiences page within Campaigns now provides detailed segment-level reporting covering demographics, individual segments, and exclusions, and allows audience management to be performed directly from the reporting interface without navigating to a separate Audience Manager screen.[1]
  • 100-user minimum threshold clarification: Google has clarified that the 100-active-user requirement for the Search Network applies specifically to lists uploaded or refreshed after 1 February 2024. Lists created before that date are subject to the prior threshold. This distinction matters for accounts auditing legacy lists that were built before the threshold change.[1]
  • Change history window remains 30 days: Google Ads Change History and the Ads API ChangeEvent resource both expose a maximum of 30 days of historical change data. Accounts that do not maintain their own change logs or exports have no visibility into changes made more than 30 days ago within the platform.[1]

What has not changed

  • Maximum membership duration remains 540 days for both Search and Display remarketing lists.[1]
  • The Display and YouTube minimum audience size remains 100 active users in the last 30 days.[1]
  • New lists still require 48 to 72 hours to populate after tag deployment before they begin accumulating users.[10]
  • Consent Mode mechanics, Customer Match consent requirements, and the prohibition on using cookieless pings for remarketing list building are unchanged from prior documentation.[3][11]

What to do now

  • Audit your account’s Audiences tab using the updated “your data segments” terminology to ensure all lists are correctly labelled, sized, and applied to the right campaigns.[1]
  • Confirm that any lists created or refreshed after 1 February 2024 meet the 100-user minimum and are eligible to serve.[1]
  • Export your change history log now if you need a record beyond the next 30 days; the platform will not retain it.[1]

References

  1. [1] https://support.google.com/google-ads/answer/2472738?hl=en support.google.com
  2. [2] https://pressvisibility.com/google-ads-2026-master-remarketing-and-cpa/ pressvisibility.com
  3. [3] https://www.groas.com/post/google-ads-remarketing-display-retargeting-2026-complete-guide-… www.groas.com
  4. [4] https://roa-marketing.com/blog/google-ads-remarketing-audiences-setup-guide-2026/ roa-marketing.com
  5. [5] https://adstralis.agency/en/blog/remarketing-customer-match-google-ads-australia/ adstralis.agency
  6. [6] https://support.google.com/google-ads/answer/2453998?hl=en support.google.com
  7. [7] https://developers.google.com/ads-data-hub/guides/audience-activation developers.google.com
  8. [8] https://support.google.com/google-ads/topic/3122874?hl=en-AU support.google.com
  9. [9] https://marketingagency.sg/ga4-audiences-segments-guide/ marketingagency.sg
  10. [10] https://support.google.com/google-ads/answer/7558048?hl=en support.google.com
  11. [11] https://support.google.com/google-ads/answer/7538811?hl=en support.google.com
  12. [12] https://gitnux.org/remarketing-statistics/ gitnux.org
  13. [13] https://support.google.com/analytics/answer/9313634?hl=en support.google.com
  14. [14] https://www.datafeedwatch.com/blog/google-remarketing-campaigns www.datafeedwatch.com
  15. [15] https://almcorp.com/blog/google-ads-retargeting-data-segments-guide/ almcorp.com
  16. [16] https://agencemoire.com/remarketing-google-ads-le-guide-2026/ agencemoire.com
  17. [17] https://support.google.com/google-ads/answer/14733311?hl=en support.google.com
  18. [18] https://business.google.com/us/ad-solutions/demand-generation/ business.google.com
  19. [19] https://business.google.com/us/accelerate/resources/articles/google-ads-ai-essentials-vide… business.google.com
  20. [20] https://support.google.com/displayvideo/answer/16062960?hl=en support.google.com
  21. [21] https://support.google.com/google-ads/answer/13704860?hl=en support.google.com
  22. [22] https://services.google.com/fh/files/misc/demand_gen_line_items_in_display_video_360_imple… services.google.com
  23. [23] https://support.google.com/google-ads/answer/6287125?hl=en support.google.com
  24. [24] https://support.google.com/google-ads/answer/13695777?hl=en-GB support.google.com
  25. [25] https://searchengineland.com/google-demand-gen-upgrade-strategies-453242 searchengineland.com
  26. [26] https://developers.google.com/google-ads/api/docs/remarketing/overview developers.google.com
  27. [27] https://support.google.com/google-ads/answer/7688468?hl=en support.google.com
  28. [28] https://www.reddit.com/r/PPC/comments/17looss/what_is_the_best_and_most_versatile_remarket… www.reddit.com
  29. [29] https://support.google.com/google-ads/answer/12301500?hl=en support.google.com
  30. [30] https://support.google.com/sa360/answer/14698523?hl=en-gb support.google.com
  31. [31] https://support.google.com/google-ads/answer/10000067?hl=en support.google.com
  32. [32] https://elementor.com/blog/google-consent/ elementor.com
  33. [33] https://www.northcountrygrowth.com/blog/how-to-use-audience-targeting-in-google-ads/ www.northcountrygrowth.com
  34. [34] https://www.auditsocials.com/blog/google-consent-mode-v2-enforcement-2026-eea-advertisers-… www.auditsocials.com
  35. [35] https://services.google.com/fh/files/misc/audience_product_experts_academy_lesson_four.pdf services.google.com
  36. [36] https://services.google.com/fh/files/misc/consent_mode_overview.pdf services.google.com
  37. [37] https://www.fullrun.app/learn/audience-targeting www.fullrun.app
  38. [38] https://support.google.com/google-ads/answer/2732132?hl=en support.google.com
  39. [39] https://www.whitesharkmedia.com/blog/google-ads/supercharge-your-adwords-campaigns-with-re… www.whitesharkmedia.com
  40. [40] https://support.google.com/google-ads/answer/17061251?hl=en support.google.com
  41. [41] https://www.surowiecki.org/en/consent-mode-comparison/ www.surowiecki.org
  42. [42] https://www.bigeyeagency.com/insights/google-ads-bid-adjustments-in-2026-what-still-works-… www.bigeyeagency.com
  43. [43] https://www.fullrun.app/learn/remarketing www.fullrun.app
  44. [44] https://support.google.com/google-ads/answer/19888?hl=en support.google.com
  45. [45] https://climentmedia.com/learn/google-ads-change-history/ climentmedia.com
  46. [46] https://developers.google.com/google-ads/api/fields/v22/change_event developers.google.com
  47. [47] https://support.google.com/google-ads/thread/11494182/google-ads-closing-remarketing-lists… support.google.com
  48. [48] https://distl.com.au/insights/googles-audience-minimum-just-dropped-to-100-users-why-small… distl.com.au
  49. [49] https://help.optmyzr.com/en/articles/4984177-change-history-widget help.optmyzr.com
  50. [50] https://www.webtonic.io/blog/google-ads-audience-targeting www.webtonic.io
  51. [51] https://developers.google.com/google-ads/api/docs/change-event developers.google.com
  52. [52] https://developers.google.com/google-ads/api/docs/query/date-ranges developers.google.com
  53. [53] https://developers.google.com/google-ads/api/reference/rpc/v25/ChangeEvent developers.google.com
  54. [54] https://support.google.com/google-ads/answer/7287822?hl=en-GB support.google.com
  55. [55] https://www.linkedin.com/posts/rakshitbshetty_googleads-ppc-remarketing-activity-741144429… www.linkedin.com
  56. [56] https://digitelia.co/blog/remarketing-google-ads-audiences-bidding-creative/ digitelia.co
  57. [57] https://groups.google.com/g/adwords-api/c/Iss3ZQNW2h8 groups.google.com
  58. [58] https://lineardesign.com/blog/google-ads-remarketing/ lineardesign.com
  59. [59] https://growthmindedmarketing.com/blog/google-ads-remarketing/ growthmindedmarketing.com
  60. [60] https://www.reddit.com/r/PPC/comments/1qbkec8/how_much_time_does_it_take_for/ www.reddit.com
  61. [61] https://adwisely.com/glossary/retargeting-remarketing-google-ads/ adwisely.com
  62. [62] https://services.google.com/fh/files/misc/casestudy-rlsa-tirendo.pdf services.google.com
  63. [63] https://www.benchmarketing.org/benchmarks/google-ads/display/remarketing www.benchmarketing.org
  64. [64] https://www.firebrand.marketing/case-studies/lead-generation-with-remarketing-ads-case-stu… www.firebrand.marketing
  65. [65] https://www.seerinteractive.com/work/case-studies/ga360-remarketing-leads-to-higher-roi-re… www.seerinteractive.com
  66. [66] https://www.webfx.com/blog/ppc/google-ads-benchmarks/ www.webfx.com
  67. [67] https://www.thinkwithgoogle.com/_qs/documents/1225/dynamic-remarketing-delivers-storkie-ex… www.thinkwithgoogle.com
  68. [68] https://www.growzai.com/case-studies/edtech-google-meta-ads-leads www.growzai.com
  69. [69] https://yeezypay.io/blog/guide-to-remarketing-in-google-ads yeezypay.io
  70. [70] https://www.groas.com/post/google-display-retargeting-remarketing-l www.groas.com
  71. [71] https://www.robinprinsen.nl/case-studies/webprint-case-study-rlsa.pdf www.robinprinsen.nl
  72. [72] https://www.marketinggraham.com/wp-content/uploads/2020/10/Case-Study-Google-Ads-Remarketi… www.marketinggraham.com

Don’t want to do this yourself?

Get a complete, ready-to-launch Google Ads campaign — keywords validated against real search data, copy written to the best practices on this page, and a proper negative-keyword list — built for your business and delivered within 24 hours.

Build my campaign — $129

One-off price · Human-reviewed · No subscription · No access to your Google Ads account

This page is maintained by Sean Cooney at Omologist.com. Content is refreshed every two months using real-time research from authoritative Google Ads sources. Worked examples are illustrative scenarios calculated from published benchmarks, not client results.

Share the Post:

Related Posts