Remarketing Lists for Search Ads (RLSA) let you adjust how your search ads behave when the searcher is someone who has previously visited your website or interacted with your app. In keyword-based campaigns, RLSA worked via bid modifiers: +20% when the searcher is a past visitor, -100% to exclude them entirely, or +bid-only to only show ads to previous visitors.
In AI Max for Search, RLSA functions differently. Understanding the change matters because the old RLSA workflow (explicit bid modifier stacking) doesn't apply.
How RLSA Works in AI Max for Search
AI Max for Search treats remarketing lists as audience signals — inputs to the AI's prediction model rather than explicit bid adjustments. When you add a Google Ads remarketing list (past visitors, past converters, cart abandoners, etc.) to an AI Max for Search campaign, the AI uses that signal to:
- Identify patterns in past user behavior that correlate with conversion
- Weight queries, devices, and times differently when the searcher matches the list
- Prioritize users who resemble your past visitors and converters even if they're not on the list
The AI doesn't multiply your Target CPA or ROAS bid by a +20% modifier. Instead, it adjusts its entire bid calculation based on the likelihood this particular user converts — and past site visit behavior is one of several inputs to that calculation.
What This Means Practically
In legacy RLSA + Smart Bidding, you stacked: Smart Bidding base bid × RLSA bid modifier. The modifier was an explicit override the algorithm applied.
In AI Max: the remarketing list membership is one input among many. A user who visited your product page yesterday, is searching from a high-conversion device type, during a high-conversion time, with a high-intent query — the AI combines all these signals holistically. The remarketing signal doesn't simply add a fixed percentage; it contributes to a contextual prediction.
Effect: AI Max with strong remarketing audience signals often out-performs legacy RLSA + Smart Bidding for remarketing campaigns, because the AI can combine the remarketing signal with other real-time signals simultaneously. But the AI's decisions aren't transparent in the same way explicit bid modifiers were.
Audience List Types for AI Max RLSA
All Google Ads audience list types work as signals in AI Max for Search:
Google Ads website visitor lists: Created via the Google Ads tag. Pages visited, time on site, and specific URL patterns. These feed AI Max as behavioral signals.
GA4 audience lists: Shared from GA4 to Google Ads. More sophisticated than Google Ads native lists — can include multi-step behavioral sequences, predictive audiences (users predicted to purchase in the next 7 days), and event-based segmentation.
Customer Match: Uploaded email lists matched to Google accounts. Your highest-quality first-party data signal for AI Max.
App user lists: For advertisers with apps; events from the app (install, in-app purchase, etc.) available as audience segments.
For AI Max for Search, upload all available audience types and let the AI determine which segments are most predictive of conversion. Don't limit to one list type.
The June 2026 Deletion and Audience List Sizes
The June 2026 data deletion (https://yositeup.com/blog/google-ads-reporting-data-deleted-june-2026) affected data within Google Ads, but website-based remarketing lists are governed by cookie/user matching data — separate from the reporting data that was deleted.
However: remarketing lists have membership windows. If your standard site visitor list has a 30-day window, users from June 15-July 15 who are now past the window have already cycled off. The deletion period may have overlapped with a window expiry for some lists.
Check your remarketing list sizes (in Shared Library > Audiences) for any unexpected drops. If lists look smaller than expected for lists with 30-60 day windows, the deletion period may have created a gap in new list additions. No direct fix — the lists will repopulate as new visitors arrive.
DSA Migration and RLSA Setup
For DSA campaigns migrating to AI Max for Search (https://yositeup.com/blog/google-ads-dsa-ai-max-migration-february-2027), check whether your DSA campaigns had RLSA audiences configured:
- RLSA audiences on DSA used bid modifiers
- On AI Max for Search, these same audiences become signals (no modifier needed)
- Add the audiences to the AI Max campaign as signals in the Audience section of campaign setup
- Don't add bid modifiers — AI Max ignores them for audiences
If DSA ran with "target" audience setting (only showed ads to RLSA members), review whether the AI Max campaign should also be restricted to remarketing audiences. In most cases, remove the restriction in AI Max for Search and let the AI expand beyond the list using the list as a signal for look-alike targeting.
AI Max for Shopping and RLSA/Remarketing
For AI Max for Shopping (https://yositeup.com/blog/google-ai-max-shopping-replacing-performance-max-2026), Shopping campaigns support audience signals similarly: past purchasers, cart abandoners, and product page viewers as signals.
Shopping campaigns also support dynamic remarketing: serving Shopping ads featuring the exact product a user viewed on your site. This requires:
- Google Ads dynamic remarketing tag on product pages, passing product IDs
- Product IDs that match your Merchant Center feed
- Dynamic remarketing audience enabled in the AI Max for Shopping campaign
Under the July 2026 ToS (https://yositeup.com/blog/google-ads-tos-july-2026-ai-automation-what-changed), dynamic remarketing data use for AI Max is covered under the standard AI Max data use provisions. Consent Mode v2 compliance is required for EU traffic — users who haven't consented cannot be added to remarketing lists or served dynamic remarketing ads.
Top comments (0)