Value-based bidding (VBB) means bidding more for conversions that are worth more — not treating every conversion as equal. In AI Max campaigns, VBB is not just a tactic; it's the mechanism through which the AI optimizes for revenue rather than conversion volume.
The Fundamental Shift: From Volume to Value
Traditional Smart Bidding with Target CPA tells the AI: "get as many conversions as possible at this cost." Target ROAS tells the AI: "maximize conversion value relative to cost." When you set a ROAS target, you're already doing value-based bidding in its basic form.
But if every conversion is reported as having the same value (or no value at all), Target ROAS degenerates into Target CPA behavior. The AI can't distinguish between a $20 sale and a $2,000 sale if both are reported as "1 conversion."
VBB works when conversion value is reported accurately and varies across conversions. The wider the value spread in your data, the more value VBB adds.
Dynamic Conversion Values
The foundation of VBB: passing dynamic conversion values from your e-commerce platform or CRM.
For e-commerce (AI Max for Shopping: https://yositeup.com/blog/google-ai-max-shopping-replacing-performance-max-2026): the purchase conversion event should pass the actual order total as conversion value. GTM variable: {{dlv - ecommerce.purchase.revenue}}. This is dynamic per transaction — each order reports its actual value.
For lead generation: assign a proxy value based on expected close rate and average deal size. If your close rate is 10% and average deal is $5,000, a lead is worth approximately $500. Use this as a fixed conversion value initially, then refine with conversion value rules as you gather data.
Conversion Value Rules
Conversion value rules (CVRs) let you apply multipliers to reported conversion values based on audience, device, or geographic signals. Under the July 2026 ToS (https://yositeup.com/blog/google-ads-tos-july-2026-ai-automation-what-changed), the AI uses CVRs as inputs to its bidding model — they are directives that adjust what the AI treats as the "real" conversion value.
Audience-based CVR: if Customer Match list members close at 2x the average order value, set a CVR of 2.0 for that audience. The AI will bid as if those conversions are worth twice as much.
Geographic CVR: if conversions from specific states (for US-targeted campaigns) have significantly higher average order values, set geographic CVRs to reflect this.
Device CVR: historically, mobile had lower conversion rates. In 2026, mobile conversion patterns differ by business type. If your data shows desktop converters have 40% higher order values, a device-based CVR of 1.4 for desktop adjusts the AI's bidding.
Path: Google Ads > Tools > Conversions > Conversion Value Rules > New Rule.
The Data Deletion Impact on VBB
The June 2026 data deletion (https://yositeup.com/blog/google-ads-reporting-data-deleted-june-2026) removed the historical conversion value data the AI had used to calibrate its understanding of value signals. Campaigns that were optimizing effectively on VBB may show degraded performance post-deletion as the AI rebuilds its understanding of value patterns.
For VBB campaigns recovering from the deletion period:
- Check if ROAS targets are still calibrated to current reality (not pre-deletion reality)
- Review CVR multipliers against current data (is the 2x Customer Match multiplier still accurate?)
- Consider pulling GA4 data to recalculate average order values by segment and update CVR multipliers accordingly
DSA Migration and VBB
DSA campaigns (retiring February 2027: https://yositeup.com/blog/google-ads-dsa-ai-max-migration-february-2027) often used fixed conversion values or simple ROAS targets. When migrating to AI Max for Search, implement dynamic conversion values and CVRs at the same time. Migrating to AI Max without VBB misses a significant optimization opportunity.
Migration VBB setup checklist:
- Confirm purchase/lead conversion actions pass dynamic values
- Create conversion value rules for your highest-value audience segments
- Set Target ROAS (not Target CPA) as the bid strategy for AI Max for Search if value varies across conversions
- If value is uniform, Target CPA is fine — VBB adds marginal value when all conversions have the same value
Profit Margin Bidding
For advertisers with significant variation in product margins, the most advanced VBB approach is profit margin bidding: reporting conversion values as profit (not revenue).
If you sell Product A with 60% margin and Product B with 15% margin, and you target ROAS at the revenue level:
- Product A ROAS target implicitly includes high margin — acceptable ROI
- Product B ROAS target at the same rate may be unprofitable after accounting for margin
By passing gross profit as conversion value instead of revenue, the AI optimizes for profit margin rather than revenue. The AI Max bidding model then naturally allocates more budget to the high-margin product queries.
This requires a Merchant Center or data layer adjustment to pass margin-adjusted values, rather than order totals, as conversion values. For AI Max for Shopping, this can be done through Merchant Center feed customization (custom labels for margin tiers) combined with CVRs.
Top comments (0)