How do GA4 audiences work for remarketing?
GA4 audiences are defined user lists built from event and parameter data, automatically updated as users qualify or expire based on your definition. Publish them to Google Ads (via the GA4-Google Ads link) and they become targetable audiences for Display, YouTube, Search (RLSA), Shopping, and Performance Max campaigns.
The typical audience size lag: users are added within 24–48 hours of qualifying events; audiences must reach minimum sizes (1,000 users for Display/YouTube, 1,000 for Search RLSA) before they can be targeted. Most high-value remarketing audiences for mid-market properties reach minimum size within 7–14 days of being created.
The highest-converting remarketing audiences share a common trait: they capture users who demonstrated specific intent signals (viewed a product, started a checkout, used the search function) rather than vague behavioural patterns.
GA4 audience types
Standard audiences (manual definition)
You define the conditions. Users who meet the conditions are added. Users who expire the membership duration are removed.
Configuration:
Admin → Audiences → + New Audience → Create a custom audience
Key settings:
- Conditions: event names, parameter values, user properties, or combinations
- Membership duration: how long users stay in the audience after qualifying (1 to 540 days)
- Audience trigger: optionally fire a GA4 event when a user first joins this audience
Suggested audiences (pre-built templates)
GA4 provides template audiences for common use cases (cart abandoners, recent purchasers, etc.). These are starting points — review and customise before publishing to Google Ads. Template audience definitions are sometimes broader than optimal.
Predictive audiences
GA4's machine learning-based audiences that predict future user behaviour. Available types:
- Likely 7-day purchasers: users predicted to purchase within 7 days
- Likely 7-day churning users: users who are predicted to become inactive
- Predicted 28-day top spenders: users in the top 10th percentile of predicted revenue
Minimum thresholds required for predictive audiences:
- At least 1,000 users who triggered the purchase (or churn) event in the past 28 days
- At least 1,000 users who did NOT trigger the event in the past 28 days
- The model requires this data split to train
Most SMB properties don't meet the 1,000-purchaser threshold. Mid-market and enterprise e-commerce typically qualify. Check: Admin → Predictive capabilities — if this section shows "More data needed," your property doesn't yet qualify.
The 8 highest-converting remarketing audiences
Audience 1 — Cart abandoners (3-day window)
Definition: Users who triggered add_to_cart AND did NOT trigger purchase within the last 3 days.
Membership duration: 7 days
Why 3 days? Cart abandonment intent decays quickly. Users who abandoned 14 days ago have much lower purchase intent than users who abandoned yesterday or 3 days ago.
Google Ads use: Display and YouTube campaigns with high urgency creative ("Still thinking about [product]?"). RLSA bid boost on generic product searches.
Typical conversion lift vs non-remarketing: 3–8x higher conversion rate.
Audience 2 — Product page viewers, no cart add (7-day window)
Definition: Users who triggered view_item AND did NOT trigger add_to_cart within the last 7 days.
Membership duration: 14 days
Why this matters: These users are earlier in the funnel than cart abandoners — they saw a product but weren't motivated to add it. Creative should focus on benefits and social proof, not urgency.
Audience 3 — High-intent site searchers
Definition: Users who triggered search event (requires site search tracking implementation).
Want to see whether purchase, revenue, or item-level tracking is drifting in your property?
Membership duration: 30 days
Why this matters: Users who search your site have specific product intent beyond casual browsing. Even without a cart add, they're demonstrably higher intent than average visitors.
Audience 4 — Recent purchasers (repeat purchase window)
Definition: Users who triggered purchase within the last 30 days.
Membership duration: 90 days (or align with your typical repeat purchase cycle)
Use: Cross-sell and upsell campaigns. Show complementary products or refill campaigns.
Exclude from: New customer acquisition campaigns (bidding on your own existing customers for first-purchase offers wastes spend).
Audience 5 — Lapsed purchasers
Definition: Users who triggered purchase more than 90 days ago AND have NOT returned (no session_start) in the last 60 days.
Implementation note: This requires combining two temporal conditions in GA4's audience builder. Create it as:
- Include: event
purchaseoccurred more than 90 days ago (use "All sessions" time window → filter by event count > 0 with time condition) - Exclude: users who triggered
session_startin the last 60 days
Use: Win-back campaigns. Higher bid adjustments + stronger incentive messaging.
Audience 6 — Email leads who haven't purchased
Definition: Users who triggered generate_lead (lead form submission) AND did NOT trigger purchase.
Membership duration: 90 days
Use: Lead nurture display campaigns. Particularly effective for B2B and high-consideration purchase categories.
Audience 7 — Checkout starters
Definition: Users who triggered begin_checkout AND did NOT trigger purchase.
Membership duration: 7 days
Note: This is a subset of cart abandoners (those who went further in the funnel). These users deserve separate, higher-urgency campaigns from general cart abandoners — they were much closer to purchasing.
Typical conversion lift: 5–12x vs cold audiences.
Audience 8 — High-value look-alike seed audience
Definition: Users who triggered purchase at least 2 times (repeat purchasers, higher LTV signal).
Membership duration: 540 days (maximum — you want to keep your highest-value customers in this seed list)
Use: Not for direct remarketing — use as the seed list for Google Ads similar audiences (reach users similar to your best customers for prospecting campaigns).
Four audience types that underperform
All site visitors (30-day window): Too broad. Most visitors have no purchase intent. Click-through rates are very low, impression costs are high, and ROAS is typically below break-even. Use more specific intent-based audiences instead.
Audiences based on page views alone (without intent signals): "Users who viewed the homepage" or "users who viewed the blog" don't indicate purchase intent. Blog readers especially tend to be low-intent research visitors.
Very long membership windows without refreshing: A 540-day cart abandoner audience includes users who may have purchased from competitors, moved house, or changed circumstances entirely 18 months ago. Use 30-day maximum for urgency-based creative.
Audiences built on unverified custom events: If the add_to_cart event fires incorrectly (e.g., firing on wishlist adds or product image clicks), the audience is polluted with non-intent users. Verify event firing before building audiences from custom events.
FAQ: GA4 Audiences and Remarketing: Building Lists That Actually Convert
What is the first thing to verify when ga4 audiences and remarketing: building lists that actually convert affects revenue?
Should I compare GA4 only to the ecommerce platform total?
How do I keep this from breaking after the next release?
Related guides for GA4 Audiences and Remarketing: Building Lists That Actually Convert
Shopify GA4 Setup: Web Pixel vs theme.liquid in 2026
The 2026 best practice for Shopify GA4 is Web Pixel via Customer Events API — Shopify's sandboxed pixel system that runs GA4 in isolation, supports the Customer Privacy API for consent, fires standard e-commerce events automatically, and works on all Shopify plans (including Basic)…
Item Array Integrity: What Stops Items Reporting in GA4 (2026)
Items fail to appear in GA4 e-commerce reports when the items array is missing, malformed, or inconsistent across the funnel. Eight common bugs: (1) missing item_id (the only required field — items without it are dropped), (2) item_id mismatched between view_item and purchase (same product reports as different items)…
Audit GA4 Audiences and Remarketing: Building Lists That Actually Convert before revenue reporting drifts further
Run a free GA4 audit to catch purchase, refund, item-array, and attribution issues before they distort ecommerce decision-making.