GA4 audiences do their real work outside GA4. Every audience you define can flow automatically to Google Ads and Display & Video 360, where it becomes a remarketing pool, an exclusion list or a bidding signal. Ecommerce teams that treat the audience builder as a revenue tool — not a reporting curiosity — consistently squeeze more conversion out of the same traffic. These are the ten audiences worth building first, with the definitions that make them work.
1. Cart abandoners (high intent, unfinished)
Users with add_to_cart in the last 7 days and no purchase. This is the highest-yield remarketing pool in ecommerce. Keep the window short — seven days captures the decision cycle for most categories — and pair the campaign with a message that resolves hesitation: shipping cost clarity, returns policy, or a gentle reminder, not necessarily a discount.
2. Checkout abandoners (highest intent of all)
Users with begin_checkout and no purchase, 3-day window. These users cleared product doubt and stalled on friction — payment issues, forced account creation, surprise costs. They deserve their own campaign with tighter frequency and stronger urgency than cart abandoners, and their volume is a standing diagnostic: a growing checkout-abandoner audience is a checkout-UX bug report.
3. Predicted 28-day top spenders
GA4's predictive audience of users likely to spend the most in the next 28 days. Machine-learned from purchase probability and revenue signals, this is your Smart Bidding accelerant: layer it observationally onto Search and Shopping campaigns, then bid up once the data confirms its quality. Requires sufficient purchase volume for predictions to activate — most mid-size stores qualify.
4. Predicted churning purchasers
Past buyers GA4 predicts are unlikely to return. Catching a lapsing customer before the lapse costs a fraction of re-acquiring them cold. Route this audience to a win-back campaign and to your email platform via the GA4–BigQuery export if you run lifecycle messaging outside Google.
5. High-value repeat purchasers
Users with 2+ purchases or lifetime revenue above your threshold. Two uses: a seed for Google Ads' optimized targeting to find lookalike behavior, and an exclusion from prospecting campaigns so acquisition budget stops paying for customers you already own. The exclusion alone typically funds the effort.
6. Category browsers who never bought
Users with multiple view_item events in one product category, no purchase, 30-day window. Build one per flagship category. These power category-specific remarketing with matching creative — far higher click-through than generic dynamic ads — and reveal which categories attract interest they fail to convert.
7. Engaged non-converters
Sessions over two minutes or 5+ page views, zero commerce events, 30 days. These users like you but haven't entered the funnel. They are your cheapest conversion opportunity — often one merchandising nudge away — and the right pool for testing offers without training discount-hunting behavior in existing buyers.
8. First-time buyers in their honeymoon window
Exactly one purchase, within the last 30 days. The second purchase is the strongest predictor of long-term value in most ecommerce data. Target this audience with complementary-product campaigns timed to your category's natural repurchase interval, and exclude them from first-order discount messaging they no longer need.
9. Promo-responsive shoppers
Users whose purchases coincided with coupon events or promo-code landing pages. Activate them when you actually run promotions and exclude them when you don't — showing full-price campaigns to discount-conditioned buyers wastes impressions, and the audience keeps your promotion economics honest.
10. Recent purchasers (universal exclusion)
Purchasers in the last 14–30 days, excluded from acquisition and most remarketing. The least glamorous audience on this list and often the largest single efficiency gain: every impression not wasted on someone who just bought is budget reallocated to someone who might.
Making them work
Three disciplines turn this list into results. Define audiences from clean events — audiences inherit every flaw in your tagging, so validate add_to_cart and purchase first. Mind the membership windows — GA4 audiences are not retroactive, so build them before the campaign that needs them. And review performance by audience monthly in Google Ads: the segments that outperform earn bigger windows and budgets; the ones that don't get rebuilt with tighter definitions.
If your GA4 property lacks the event quality or purchase volume to power predictive audiences, that is a measurement-engineering problem worth solving — and a well-scoped project for a GA4 partner with ecommerce evidence.
Frequently Asked Questions
How do GA4 audiences get into Google Ads?
Link GA4 to Google Ads and enable personalized advertising signals; audiences then flow automatically into the shared library. New audiences populate going forward from creation — they are not retroactive — so build them before launching the campaigns that use them.
What volume do GA4 predictive audiences require?
For purchase-based predictions, GA4 generally needs at least 1,000 returning users who triggered the predictive condition and 1,000 who did not, within the last 28 days, with model quality sustained. Most established ecommerce sites qualify; very new or low-traffic stores may not.
Why are my GA4 remarketing audiences smaller than expected?
Common causes: consent denials removing users from ads personalization, membership windows shorter than intended, audience conditions built on mis-firing events, and the non-retroactive start. Check event volumes in GA4 first — audiences can only be as complete as the events beneath them.
Should I use GA4 audiences or Google Ads' own segments?
Both, for different jobs. GA4 audiences encode site-behavior logic (cart abandonment, category interest, predicted value) that Google Ads cannot see natively, while Ads' in-market and optimized targeting expand reach. The strongest accounts feed GA4 behavioral audiences into Ads as signals and exclusions.