GA4 Audiences for Google Ads: Build, Link and Target

GA4 audiences are auto-updating groups of users who meet conditions you define in Google Analytics 4, such as "added to cart but did not purchase". When you link your GA4 property to Google Ads, those audiences flow into the ad platform and serve two jobs: direct targeting as remarketing lists, and extra signal for Smart Bidding. This guide covers the full path with the 2026 thresholds, from building the audience correctly to using it well in Google Ads.

What is a GA4 audience and why does it matter for Google Ads?

A GA4 audience is a dynamic user group defined by conditions built on event and user data: users enter when they qualify and drop out when membership expires or conditions no longer apply. It matters for Google Ads because, in an era of weakened third-party cookies, your own site's behavioral data is the most reliable targeting source, and GA4 is the main bridge that carries it into the ad system.

• Audience vs segment: A segment is a temporary filter used for analysis in Explorations; an audience is persistent, accumulates users and can be exported to Google Ads.
• Dynamic membership: Users join as they qualify and leave when the membership duration (default 30 days, maximum 540) expires.
• Automatic flow: Once the link exists, audiences land in the Google Ads shared library on their own; no per-list export is needed.
• Dual use: The same audience works as a remarketing target and as a signal for Smart Bidding and Performance Max.

How do you create an audience in GA4?

Audience creation lives in the Admin panel: you define conditions, pick a membership duration and save. The audience starts populating from that moment; in practice, expect the first 24-48 hours before it reaches a meaningful size. The most common build mistake is an overly broad single-event condition; narrowing the event with a second condition such as page path, repeat count or channel visibly raises the intent quality of the list. The steps:

1. Go to Admin > Audiences > New audience; pick a suggested template or build from scratch.
2. Define conditions: add filters based on events (add_to_cart, view_item), page paths, channels or user properties, combined with AND/OR logic.
3. Add exclusions: for example, remove users who triggered purchase in the last 30 days from the cart audience, so budget is not spent on users who already converted.
4. Pick the membership duration: 7-14 days suits short decision cycles, 90-180 days long ones; the ceiling is 540 days.
5. Name it and save: names like "Cart abandon 14d US" that encode condition and duration reduce selection mistakes on the Google Ads side.

How does the GA4 audience export to Google Ads work?

The export relies on a link between the GA4 property and the Google Ads account, with "enable personalized advertising" switched on in the link settings. Once that is done, all existing and future audiences flow automatically into the shared library under Tools in Google Ads. All of the following preconditions must hold:

• Account link: GA4 Admin > Product links > Google Ads; the user creating the link needs sufficient access on both sides.
• Personalized ads signal: Ads personalization must be on both at property level and in the link; if it is off, lists appear empty.
• Consent management: For European Economic Area traffic, users are not added to ad audiences without Consent Mode v2 (ad_user_data and ad_personalization).
• Size threshold: As of 2026 Google lowered the usable threshold to roughly 100 active users across Search, Display and YouTube; the old 1,000-user rule for Search is history.

Which GA4 audiences improve Google Ads performance?

The audiences that pay off most are groups carrying purchase-intent signals without having converted: cart abandoners, users who browsed product pages without requesting an offer, pricing-page visitors. In practice, conversion rates typically climb the closer the group sits to the bottom of the funnel; the five recipes below are a solid starter set for most accounts.

• Cart abandoners: add_to_cart present, purchase absent; 7-14 day membership. The highest-intent remarketing audience.
• High-intent browsers: Users who viewed a specific product or pricing page 2+ times; target with mid-funnel messaging.
• Engaged non-converters: High session duration or page depth, no conversion event; a broad upper-funnel re-engagement pool.
• Past purchasers: A target for cross-sell and repeat-purchase campaigns, an exclusion list for new-customer campaigns.
• Lapsing high-value users: Former frequent buyers inactive for 60-90 days; reach them with win-back messaging.

Which message and frequency to use per audience is its own topic; we cover the playbook step by step in our remarketing strategy guide. What matters here is that every audience name encodes its condition and duration, and that purchasers are consistently excluded from new-customer campaigns.

When do predictive audiences come into play?

A predictive audience is one GA4 builds by forecasting future behavior with its machine learning model: "users likely to purchase within 7 days" or "active users likely to churn within 7 days". The purchase probability metric scores the chance that a user active in the last 28 days will trigger the purchase event in the next 7 days.

The bar is high: per platform documentation, the model needs at least 1,000 users who triggered the positive example (such as purchasing) and at least 1,000 who did not, within the last 28 days, and the purchase event must be measured correctly; you can verify the setup with our GA4 purchase event checklist. On accounts with enough traffic, predictive audiences surface users classic remarketing lists miss: those close to converting who have not signaled it yet.

How do you use GA4 audiences as a bidding signal?

Adding an audience to a campaign in "observation" rather than "targeting" mode does not narrow reach; Google Ads reports that audience's performance separately and Smart Bidding uses the data as an extra signal in bid decisions. This is the low-risk entry: observe first, then move to targeting or bid adjustments once the audience proves it converts better.

• Observation on Search campaigns: Add the audience in observation mode; with manual bidding, apply positive bid adjustments to audiences that convert well.
• Audience signals in Performance Max: GA4 audiences act as starting hints, not strict targets; the system expands toward similar users.
• Exclusion as a signal: Excluding existing customers and low-value traffic steers budget toward new, high-intent users.

What should you check when an audience does not appear or populate?

The most common causes sit in the link and consent layer: if personalized ads are off, the consent signal is missing or the audience is under the 100-user threshold, the list shows as ineligible in Google Ads. Debug in this order:

1. Verify the link: Is the GA4 Admin > Google Ads link active, and is personalized advertising enabled?
2. Check the size: How many users does the audience show in GA4? Below 100, widen the condition or extend the membership duration.
3. Test the consent signal: On EEA traffic, do the Consent Mode v2 parameters (ad_user_data, ad_personalization) return granted?
4. Allow time: A new audience starts filling within 24-48 hours; an empty first day is normal.
5. Separate data gaps: GA4 and Google Ads numbers never match one to one; we explain how much divergence is normal in the conversion discrepancy guide.

Once audiences are live, the real work is measurement: which list drives conversions in which campaign, and which one burns budget? Ads Sensor unifies your GA4 traffic and conversion breakdown with your Google Ads, Meta and TikTok data in one panel; its AI analysis turns risks and opportunities into reasoned actions. Join the beta and try it on your own account.