Logo

The Data Daily

Facebook Brings Machine Learning Into the Dynamic Ads Creation Process

Facebook Brings Machine Learning Into the Dynamic Ads Creation Process

Facebook is adding a touch of machine learning to help brands more accurately personalize dynamic ads.

The social network’s new dynamic formats and ad creative solution delivers a personalized version of an ad to everyone who sees it, based on formats they have proven more likely to respond to, such as collection and carousel.

Facebook said in a blog post that this solution can be leveraged within Facebook dynamic ads when they are created for the catalog sales, traffic and conversions objectives in Ads Manager or via the Marketing API (application-programming interface).

The social network added that it recently conducted a test with 12 ecommerce and retail advertisers and found that dynamic formats and ad creative outperformed carousel-only ads in driving lift across content views, add-to-cart, purchases and sales, with a 34% improvement in incremental return on ad spend, a 10% boost in life and a 6% drop in cost per incremental purchase.

When brands are creating single-media ads for traffic, application installs and conversions objectives, they can add multiple text optimization, enabling them to input multiple options for the primary text.

Facebook said in its blog post, “This allows our ad system to optimize for delivery and performance using variations of the text options provided, based on individual preferences identified by our machine learning models.”

Advertisers can also use auto-translation for single-media ads, enabling them to reach international customers in their local languages and save on the resources they would have needed to do the translations themselves.

Brands opting for this feature have controls that allow them to review the translations it produces and, should they wish, to provide their own.

The social network wrote, “By leveraging machine learning to dynamically select ad formats, creative and text at the ad impression-level based on what might work best for the individual in question, we can improve advertiser performance and the consumer experience.”

Images Powered by Shutterstock