Abstract: One or more embodiments of the present disclosure involve training and utilizing a recall machine learning model to predict recall lift on a per-user basis with respect to digital content items. For example, systems described herein train a recall machine learning model based on poll responses from exposed users and non-exposed users with regard to sample digital content. In particular, the systems described herein train the recall machine learning model to output recall lift scores for a target user based on an assumption that the target user has been exposed to digital content and an assumption that the target user has not been exposed to the digital content. The systems described herein further involve delivering digital content in accordance with the recall lift score.

Abstract: An online system generates a custom audience based on user engagement with distributed content items. The online system monitors and stores user engagement with content items. A content provider submits to the online system a request to generate a new custom audience and selects audience parameters and user engagement types, such as users who watched a specified amount or percentage of a video, users who interacted with an online system page associated with the content provider, or users who clicked through online system content to a page associated with the content provider. Responsive to the request, the online system retrieves the corresponding user engagement data and applies the audience parameters to generate the custom audience and present it to the content provider.