Snapchat is using machine learning to introduce greater targeting to its ad stack
Snapchat will allow those advertisers buying ads from third parties to use machine learning to better determine which users are most likely to swipe on which type of ad.
Snapchat unveils update to its ad technology
Goal-based bidding allows advertisers who buy through the mobile messaging app’s fledgling application programming interface (API) to set goals on their campaigns beyond just impressions on ads.
While advertisers will continue to be charged on impressions, the update takes into account users who engage further with ads through ‘swipe ups’, which can redirect users to an app install page, webview, or long-form video content, and gives advertisers the option to assign value to trade on this on a CPM basis.
The aim is to allow advertisers to generate swipes more cost-effectively by reducing wasted impressions and increase the amount of time people spend on ads by serving them to users who are most likely to swipe on a certain type of ad. The feature optimizes bidding and delivery to those users automatically through its machine learning capabilities.
Goal-based bidding launched earlier this month and one-in-five (20%) of those advertisers using its API have used it. It’s a small update that comes in the wake of reports that advertisers are concerned by the average time their ads are being viewed by. A recent AdAge report claimed that the average viewing time for one unnamed advertiser was less than three seconds.
Snapchat has doubled down its adtech capabilities in the past year to give advertisers more control over campaigns in the app and better measurement as it attempts to compete with the might of Facebook and its sophisticated ad stack.