Optimizing value of content items delivered for a content provider

A non-transitory computer readable storage medium storing computer program instructions executable by a processor to perform steps comprising:

• receiving, by an online system, a plurality of content items from a content provider, the plurality of content items for delivery to users of the online system, and each content item associated with one of a product or service;
• receiving, from the content provider, a target return on spend value for the plurality of content items;
• delivering the plurality of content items to a plurality of users;
• determining, for each of the plurality of content items, the cost to the content provider of displaying each content item, the determination based on the delivery of the plurality of content items;
• monitoring a rate of user interactions associated with the plurality of content items for training a machine learning model to predict a likelihood of a user taking a desired action regarding the content item;
• providing, as input to the trained machine learning model, factors associated with each of the plurality of content items, the factors including the monitored rate at which users interact with the content item;
• applying the trained machine learning model to the provided input for determining, for each of the plurality of content items, a rate at which the content item is subsequently delivered to users, wherein the rate of delivery of each content item is determined such that a sum of the weighted rates of delivery of content items being delivered for the content provider is equal to the target return on spend for the content provider, wherein a weight of a rate of delivery of a content item is indicative of a value of an item associated with the content item;
• and
• delivering the content items to users at the determined rates.

A computer system comprising:

• one or more computer processors for executing computer program instructions;
• and
• a non-transitory computer-readable storage medium storing instructions executable by the one or more computer processors to perform steps comprising:
• receiving, by an online system, a plurality of content items from a content provider, the plurality of content items for delivery to users of the online system, and each content item associated with one of a product or service;
• receiving, from the content provider, a target return on spend value for the plurality of content items;
• delivering the plurality of content items to a plurality of users;
• determining, for each of the plurality of content items, the cost to the content provider of displaying each content item, the determination based on the delivery of the plurality of content items;
• monitoring a rate of user interactions associated with the plurality of content items for training a machine learning model to predict a likelihood of a user taking a desired action regarding the content item;
• providing, as input to the trained machine learning model, factors associated with each of the plurality of content items, the factors including the monitored rate at which users interact with the content item;
• applying the trained machine learning model to the provided input for determining, for each of the plurality of content items, a rate at which the content item is subsequently delivered to users, wherein the rate of delivery of each content item is determined such that a sum of the weighted rates of delivery of content items being delivered for the content provider is equal to the target return on spend for the content provider, wherein a weight of a rate of delivery of a content item is indicative of a value of an item associated with the content item;
• and
• delivering the content items to users at the determined rates.