Attribution in Google Ads has been evolving for years. In 2026, data-driven attribution (DDA) is the default for new campaigns, last-click is deprecated for many campaign types, and AI Max campaigns use attribution models in ways that differ from traditional campaign types. Here is what you need to know.
The Shift to Data-Driven Attribution
Data-driven attribution assigns fractional credit to multiple touchpoints in a conversion path. The model is trained on your account's conversion data to determine which touchpoints (impressions, clicks) correlate with conversions in your specific account.
For AI Max campaigns, DDA is the default and the recommended model. The AI makes bidding decisions that incorporate multi-touch value signals, not just last-click. This means AI Max bids differently than traditional campaigns on the same keywords — it values impressions and earlier-funnel clicks more than last-click models would.
The Data Deletion Effect on DDA Models
DDA models are trained on your account's historical conversion data. The June 2026 data deletion (https://yositeup.com/blog/google-ads-reporting-data-deleted-june-2026) removed approximately 14 months of conversion path data from the training set.
This matters because DDA models need conversion path data — sequences of touchpoints that led to conversions — to identify which touchpoints correlate with outcomes. With Q4 2025 data removed:
- DDA models have less data on Q4 seasonal touchpoint patterns
- The model's understanding of which upper-funnel interactions correlate with Q4 conversions is reduced
- Attribution weights for Q4 2026 will be calibrated from earlier data (pre-April 2025 and current 2026 data)
For accounts with high conversion volumes (1000+ monthly conversions), this impact is likely small — the DDA model has enough remaining data. For accounts with lower conversion volumes, the impact may be more significant.
Attribution and the AI Max Migration
DSA campaigns being retired in February 2027 (https://yositeup.com/blog/google-ads-dsa-ai-max-migration-february-2027) commonly used last-click or linear attribution. When you migrate to AI Max for Search, the campaign defaults to data-driven attribution.
This creates a reporting discontinuity. Before migration (DSA + last-click): 100% of conversion credit went to the final click. After migration (AI Max + DDA): conversion credit is distributed across the conversion path. Conversion counts may appear to change not because performance changed but because attribution changed.
To manage this comparison: during the migration period, run a parallel view in Google Analytics 4 (which has its own attribution settings independent of Google Ads). GA4 shows consistent attribution across the migration event.
The ToS Connection
The July 2026 ToS update (https://yositeup.com/blog/google-ads-tos-july-2026-ai-automation-what-changed) includes provisions about AI optimization authority. Attribution model selection affects how the AI values different bidding opportunities. Under the new ToS, the AI's interpretation of DDA signals (multi-touch value) is an approved form of optimization authority.
Practical implication: don't switch from DDA to last-click to "see clearer data" — this reduces the AI's ability to optimize AI Max campaigns effectively. The AI Max bidding model is calibrated for DDA. Switching attribution models mid-campaign disrupts the AI's reference frame.
AI Max for Shopping and Attribution
For AI Max for Shopping (https://yositeup.com/blog/google-ai-max-shopping-replacing-performance-max-2026), attribution becomes more complex because shopping campaigns can appear across multiple surfaces — search, display, YouTube, Maps. Each surface creates different touchpoints.
DDA in AI Max for Shopping weighs these different surface touchpoints based on their conversion correlation. The model learns which surfaces drive outcome-correlated impressions in your account specifically.
This means two accounts with identical AI Max for Shopping campaigns may receive different attribution weight distributions — because their customers' conversion paths differ.
For Q4 2026 planning: review your current DDA attribution report (Tools > Attribution > Paths) to understand which surfaces and touchpoints your DDA model is currently weighting most heavily. This provides a qualitative understanding of how the AI is interpreting your conversion signals.
Practical Attribution Actions
Don't change attribution models during Q4. Model changes during high-volume periods create measurement discontinuity that makes performance evaluation harder, not easier.
Use GA4 as your attribution-stable reference. GA4 attribution settings are separate from Google Ads. If you're comparing Q4 2026 to Q4 2024 (the surviving baseline year), GA4 provides consistent measurement across both periods.
Review the Model Comparison tool. Tools > Attribution > Model Comparison lets you see how conversion counts would differ under different attribution models. This is useful for understanding the gap between what your current model reports and what last-click would report — but it doesn't change how Smart Bidding is optimizing.
For DSA migrants specifically: document your DSA campaign's attribution model and conversion volumes before migration. After migrating to AI Max, give the new DDA model 30 days to calibrate before comparing performance to the pre-migration DSA baseline.
Top comments (0)