The simplest AI Max bidding configuration — Maximize Conversions with no target — treats all conversions as equal. A $5 newsletter signup and a $5,000 enterprise deal count the same toward the AI's optimization goal. Value-based bidding fixes this by giving the AI a monetary signal: optimize toward total conversion value, not just conversion count.
What Value-Based Bidding Means in AI Max
Value-based bidding in AI Max uses either:
Maximize Conversion Value (no tROAS): The AI allocates budget to maximize the total monetary value of all conversions, regardless of cost efficiency.
Target ROAS (tROAS): The AI tries to achieve a specific return on ad spend (total conversion value / total cost). If you set tROAS to 4x, the AI targets $4 in conversion value for every $1 spent.
Both require that your conversion actions have meaningful values assigned to them. If your conversion value is set to a static $1 for all conversions, value-based bidding behaves identically to conversion-count bidding — there's no value differentiation for the AI to optimize toward.
Assigning Conversion Values: The Foundation
The accuracy of value-based bidding depends entirely on how well your conversion values reflect actual business outcomes.
For e-commerce: Use dynamic conversion values — the actual transaction revenue passed to Google Ads via the conversion tag. This is the most accurate form of value assignment. The tag reads the purchase total from your order confirmation page and sends it with each conversion.
For AI Max for Shopping (https://yositeup.com/blog/google-ai-max-shopping-replacing-performance-max-2026), dynamic revenue tracking is standard through Merchant Center's conversion tracking integration. Confirm this is set up: in Merchant Center > Conversion sources, you should see Google Ads conversion tracking linked and passing transaction values.
For lead generation: Assign estimated values based on your conversion funnel data:
- If 10% of form submissions become customers with an average deal of $5,000, the expected value of a form submission is $500
- If 30% of phone calls convert at $2,000 average, the expected value of a phone call conversion is $600
- Use these estimated values in your conversion action settings (Conversions > conversion action > Value)
For leads where value varies significantly by lead type: use conversion value rules (Tools > Conversions > Conversion value rules) to assign different values based on audience membership, device, or location signals.
Conversion Value Rules for AI Max
Conversion value rules let you adjust conversion values based on campaign context. Examples:
- Location-based: Users in high-value geographic markets have their conversion value multiplied by 1.5x (because conversions from those markets yield higher average deal sizes)
- Audience-based: Users from your Customer Match list of enterprise customers have their conversion value multiplied by 3x (enterprise leads are worth more than SMB leads)
- Device-based: Mobile conversions get a 0.8x multiplier (they're slightly less likely to progress to a deal than desktop conversions)
These rules apply in real-time during AI Max bidding — the AI doesn't just see "conversion value: $500," it sees the contextually adjusted value based on who the user is and where they are. This feeds directly into more sophisticated bid calculations.
Impact of the June 2026 Deletion on Value-Based Bidding Models
The June 2026 data deletion (https://yositeup.com/blog/google-ads-reporting-data-deleted-june-2026) removed historical conversion value data for AI Max campaigns active during the deletion period. The AI's tROAS model was built on historical conversion value distribution — after the deletion, it had less data to base its value predictions on.
Symptoms of this in tROAS campaigns post-deletion:
- Wider conversion value swings per day (less stable predictions)
- Either over- or under-spending relative to budget (the AI is less confident in which auctions will yield high-value conversions)
- ROAS reports showing higher variance week-over-week than before the deletion
Response:
- Temporarily widen your tROAS target by 20-30% to give the AI more auction access while it rebuilds its value model
- After 8-10 weeks, gradually tighten the tROAS target back toward your goal
- Avoid making large structural changes (new assets, major budget changes) during this rebuilding period — minimize the variables
Value-Based Bidding for DSA Migration to AI Max
When migrating DSA to AI Max for Search (https://yositeup.com/blog/google-ads-dsa-ai-max-migration-february-2027): DSA typically used CPA-based bidding or maximize conversions. If your DSA was optimizing for form submissions at a flat value, switching to value-based bidding in the new AI Max for Search campaign is an improvement opportunity.
Before the migration:
- Audit your conversion funnel to assign accurate estimated values to each conversion type
- Set up conversion value rules if your lead quality varies by audience or geography
- Launch AI Max for Search with Maximize Conversion Value (no tROAS) for the first 4-6 weeks to let the AI build its value model without a target constraint
- After 30+ conversions with values, introduce a tROAS target based on observed value/cost ratios
tROAS Setting: How Tight Is Too Tight?
A tROAS target that's too aggressive (set too high) prevents the AI from finding enough qualifying auctions and leads to underspending. Setting tROAS at 10x when historical ROAS is 4x means the AI can only enter auctions it's highly confident will yield 10x returns — which may be very few.
How to calibrate:
- Check your actual ROAS from the last 30-60 days in Google Ads reporting
- Set initial tROAS at 10-20% above your historical ROAS (slightly above current performance, not aspirational)
- After 4-6 weeks of stable performance, increase tROAS by 10-15% if you want to push efficiency
- Watch budget utilization — if spend drops below 80% of daily budget, the tROAS target is too high for current auction conditions
Under the July 2026 ToS (https://yositeup.com/blog/google-ads-tos-july-2026-ai-automation-what-changed), tROAS bidding and value-based optimization are explicitly supported AI Max behaviors. The ToS provisions about AI automation cover how the AI allocates bids toward value targets — no additional compliance considerations for using tROAS vs. tCPA.
Top comments (0)