When there is a mix of asset types, like image and video, in ads in Demand Gen the system utilizes performance data and tests different combinations of headlines, descriptions, images, and videos to optimize for the best campaign results. As a result, more budget spend is allocated towards combinations that are predicted to perform better and may lead to uneven delivery between formats. If one asset isn’t serving but another is, it’s likely that the asset generating more clicks is being prioritized in serving. To check, you can pull the report for the metric in question to verify the bidding strategy and see which ad is the most productive.
If you want to deliver the assets equally, structure the campaign with dedicated components by creating separate Demand Gen line items or distinct ad groups specifically for each type of asset, country, or language. This separation allows for greater control over budget spend, bidding strategies, and performance tracking.
Demand Gen Bidding strategies and limitations
- There should be a minimum of 50 conversions in the last 30 days at the line item level.
- Fluctuations may happen during the algorithm’s 30 day learning period.
- Don’t make changes to the budget since this disrupts the learning period.
- Note: Using the Automatic budget allocation option switched on in the insertion order level changes the budget of line items daily, and may impact the learning period of the algorithm. It’s not recommended to use it, especially at the beginning of the flight.
- If there aren’t any conversions, the data needed to provide the stable spend isn’t available and spend will be impacted since bid prediction is low.
- If the above conditions are met and line items aren’t spending, the issue may be somewhere else.
- Demand Gen campaigns should start serving right away.