GA4 Audiences: Build, Export to Google Ads & Fix Them

GA4 Audiences: What They Actually Do and Why Yours Are Empty

Two weeks ago a lead-gen client sent me a Loom video. They had built six GA4 audiences, linked GA4 to Google Ads, created remarketing campaigns against those audiences, and waited. Fourteen days later, every audience showed zero users in Google Ads. No impressions, no spend, no delivery. Their remarketing budget sat idle for two weeks while competitors retargeted their visitors.

The cause: the toggle to enable personalized advertising was off. The GA4-to-Google Ads link existed, but audiences were never shared. That is a single checkbox. Two weeks of lost remarketing because a toggle defaulted to off and nobody checked.

This is the kind of thing I fix constantly. GA4 audiences are conceptually simple -- define a group of users based on behavior, push that group to Google Ads for targeting. In practice, the pipeline has at least five points of failure, and most of them are silent.

How to Create Audiences in GA4

Before fixing what breaks, here is how the setup works when everything goes right.

In GA4, go to Admin > Data display > Audiences and click "New audience." You can start from a template (Purchasers, Recently Active Users, etc.) or build a custom audience from scratch. A custom audience definition has three components:

Component	What it controls
Conditions	Which events, parameters, or user properties define membership
Scope	Whether the condition applies within a single event, a single session, or across all sessions
Membership duration	How long a user stays in the audience (1-540 days)

The scope setting is where most people get tripped up. "Across all sessions" means any user who visited /pricing at any point in the last N days qualifies. "Within the same session" means they must satisfy the condition in a single session. Different scope, completely different audience.

You can also layer sequences -- user did X, then did Y within a time window. This is useful for building GA4 remarketing audiences like "viewed product but did not purchase within 7 days." Sequences are powerful but fragile. If the events feeding the sequence are misconfigured, the sequence condition silently matches no one.

I covered the mechanics of event naming and parameter errors in GA4 Events: Parameters, Custom Events & Common Mistakes. If your audience conditions reference events that are not firing correctly, no amount of audience configuration will fix the output.

GA4 Segments vs Audiences

This trips people up more than it should. GA4 segments and GA4 audiences use similar condition builders, but they serve entirely different purposes.

Segments are retroactive filters applied inside Explorations. They slice historical data so you can analyze behavior. They do not create a persistent user list. They do not push data anywhere. They exist only inside the Exploration where you define them.

Audiences are prospective, continuously evaluated user lists. Once you create an audience in GA4, it begins accumulating users going forward. It does not backfill -- users who matched the condition before the audience was created are not included. And critically, audiences can be shared with linked Google Ads accounts for targeting.

If you build a segment in Explorations and like what you see, you still need to recreate that logic as an audience before you can act on it in Google Ads. GA4 does let you create an audience directly from a segment in Explorations, which saves time.

How to Import GA4 Audiences to Google Ads

Getting GA4 audiences into Google Ads requires three things to be true simultaneously. Miss any one and the audience list stays empty on the Google Ads side.

1. Link GA4 to Google Ads

In GA4, go to Admin > Product links > Google Ads linking and connect your Google Ads account. You need Edit access on the GA4 property and Admin access on the Google Ads account. This is the same linking step used for importing conversions, so if you have already linked for GA4 conversion tracking, this step is done.

2. Enable personalized advertising

This is the checkbox that cost my client two weeks. In GA4, under Admin > Data display > Data collection, confirm that Google signals data collection is enabled. Then under Admin > Product links > Google Ads linking, edit the link and make sure "Enable personalized advertising" is toggled on. Google's documentation on audience sharing requirements specifies this must be active for audiences to flow to Google Ads.

3. Verify in Google Ads

In Google Ads, navigate to Tools > Shared library > Audience manager. Your GA4 audiences should appear under "Your data segments" as "Analytics" sources. If they are not there, the link is either missing, the personalized-advertising toggle is off, or the audience was created before the link was established and has not been re-saved.

One detail people miss: GA4 audiences do not populate retroactively in Google Ads. The list starts accumulating users only after it is created and shared. Google Ads requires a minimum of 1,000 users before it can serve on the Search Network, and 100 for Display. Low-traffic sites may wait days or weeks before reaching usable size.

GA4 Audiences Not Populating in Google Ads: The Full Checklist

This is the section that matters most if you are reading this because something is already broken. When GA4 audiences are not populating in Google Ads, work through these causes in order. I have ranked them by frequency from my own audits.

The GA4-Google Ads link is missing or broken

Sounds obvious. Still the most common cause. Go to Admin > Product links > Google Ads linking and confirm the link is active. If you recently migrated GA4 properties or changed Google Ads account structure, the link may need to be re-established.

Personalized advertising is disabled

The second most common cause and the most frustrating because there is no error message. GA4 does not warn you that audiences are not being shared. Check the toggle under your Google Ads link settings.

Membership duration is too short

If you set a 7-day membership duration and your remarketing cycle is longer, users fall out faster than new ones enter. I recommend 30-day membership as a baseline.

Consent Mode blocks audience signals

If you run Consent Mode v2 and a significant share of users deny consent, GA4 cannot associate those users with an audience for remarketing. Consent-denied users are excluded from audience lists by design -- Google does not model audience membership the way it models conversions.

Google signals is not enabled

Audiences built on cross-device behavior require Google signals. Without it, GA4 can only match users on the device and browser where the cookie is set. Enable it under Admin > Data collection > Google signals data collection. Note that enabling Google signals subjects your data to thresholding when user counts are small.

Your traffic volume is too low

Google Ads will not serve against an audience list with fewer than 1,000 users (Search) or 100 users (Display). If your site gets 500 sessions a day and your audience condition filters for a specific behavior, the list might be under threshold for weeks. No fix except broadening the condition or increasing traffic.

If you have walked through all of these and your audiences still are not populating, the issue is almost always in the event data feeding the audience conditions. Garbage in, empty audience out. I would start with a full GA4 tracking audit to verify the underlying data before rebuilding audience definitions.

Or, if you want someone to do this for you: I offer a tracking audit that covers exactly this.

GA4 Predictive Audiences

GA4 offers predictive audiences based on machine-learning models that estimate future behavior. Three predictions are available:

Predictive metric	What it estimates
Purchase probability	Likelihood a user will purchase in the next 7 days
Churn probability	Likelihood an active user will not visit in the next 7 days
Predicted revenue	Expected revenue from a user in the next 28 days

The catch: GA4 predictive audiences require minimum data thresholds to activate. You need at least 1,000 returning users who triggered the relevant event (purchase or churn) and 1,000 who did not, within a 28-day window. Most small-to-midsize sites do not hit this threshold. I have seen it work well for ecommerce brands with over 50,000 monthly sessions and consistent purchase volume. Below that, the model does not have enough signal and GA4 simply will not generate the prediction.

When the thresholds are met, predictive audiences are genuinely useful. A "Likely 7-day purchasers" audience pushed to Google Ads lets you bid more aggressively on users the model predicts will convert. Combine this with value-based bidding and you get a feedback loop where high-value predicted users receive higher bids automatically.

Audience Strategy: What to Actually Build

The audiences I find most useful in client accounts reflect stages of intent, each mapped to a different campaign strategy:

• Bottom-funnel remarketing. Users who started checkout but did not purchase, or submitted a partial lead form. Membership duration of 7 to 14 days. Highest intent, smallest list. Target with specific messaging about what they left behind.
• Mid-funnel engagement. Users who viewed a product page, pricing page, or key service page but did not convert. Membership of 14 to 30 days. Broader list, lower intent. Match with educational content or social proof.
• Exclusion audiences. Purchasers or converted leads, used as exclusions in prospecting campaigns so you stop paying to acquire people you already converted. This is the audience most accounts are missing.

For each tier, the underlying event data has to be correct. If your begin_checkout event is not firing, your cart-abandoner audience collects nobody. I covered event implementation details in GA4 Recommended Events: Which Ones to Implement.

Common Mistakes I See in GA4 Audience Configurations

Recurring errors from audits -- I see each of these at least once a month:

• Overly narrow conditions. An audience requiring three simultaneous conditions on a site with 2,000 monthly sessions will never reach the 1,000-user threshold for Search remarketing.
• Wrong scope. Setting "within the same event" when you mean "within the same session." A user who viewed a product and then added to cart did those in separate events. Event-level scope will not match.
• Membership duration mismatch. A 540-day membership for a remarketing audience means you are targeting people who forgot you exist 400 days ago. Match membership to your actual sales cycle.
• Building audiences before fixing events. This is the most expensive mistake. Half the "GA4 audiences not populating" complaints I investigate trace back to broken event tracking, not broken audience configuration. Fix the data first.

FAQ

Not sure your GA4 audiences are actually reaching Google Ads? Let me audit your setup -- I will tell you exactly what is broken and how to fix it.