Your customer is considering their reports and notices that Search Ads 360 is using a previously generated basic linear attribution model to track conversions instead of the data-driven attribution model they set up last week. Why might Search Ads 360 be defaulting to the previous model?

The model was created during the bid strategies learning period.

There are too many keywords included in the bid strategy.

Historical data is transferring to the new model created.

The number of clicks and conversions falls below the data requirements.

Explanation

Analysis of Correct Answer(s)

Search Ads 360’s Data-Driven Attribution (DDA) models require a sufficient amount of historical data to function accurately. These models use machine learning to analyze conversion paths, and to do so reliably, they must meet specific minimum thresholds for clicks and conversions over a set period (typically 30 days).

  • If an advertiser's account volume drops below these minimum data requirements, the DDA model can no longer be considered statistically significant.
  • In this scenario, Search Ads 360 will automatically revert to a default or previously selected basic model, such as a linear or last-click model, to ensure that conversion tracking continues without interruption. This is a safeguard to maintain reporting integrity when the data is too sparse for the advanced model.

Analysis of Incorrect Options

  • Too many keywords included in the bid strategy: The number of keywords does not directly impact the eligibility of an attribution model. The key factor is the total volume of clicks and conversions across the advertiser account, not the keyword count.
  • Historical data is transferring to the new model created: While a new model uses historical data to build itself, this process is part of its creation. A reversion back to an old model indicates a failure to maintain the new model due to insufficient ongoing data, not a data transfer issue.
  • The model was created during the bid strategies learning period: The bid strategy learning period and the attribution model's data requirements are separate concepts. A bid strategy can be in a learning phase while a valid DDA model is active, as long as the data thresholds for the DDA model are met.