Utilizing multi-point optimization to improve digital content distribution

A system comprising:

• at least one processor;
• and
• a non-transitory computer readable storage medium comprising instructions that, when executed by the at least one processor, causes the system to:
• receive a content item from a content provider for delivery to users by way of a digital content distribution platform;
• determine expected delivery metrics of the content item by way of the digital content distribution platform;
• utilize a first machine learning model trained to optimize content delivery for a first target event to determine probabilities that the first target event will result from delivery of the content item to corresponding users;
• utilize a content selection process to determine whether to deliver the content item based on the determined probabilities that the first target event will result from delivery of the content item;
• determine that actual delivery metrics of the content item fail to satisfy the expected delivery metrics of the content item;
• based on the determination that the actual delivery metrics of the content item fail to satisfy the expected delivery metrics of the content item, utilize the first machine learning model and a second machine learning model trained to optimize content delivery for a second target event to determine weighted sum probabilities that the first target event or the second target event will result from delivery of the content item;
• and
• utilize the content selection process to determine whether to deliver the content item based on the determined weighted sum probabilities that the first target event or the second target event will result from delivery of the content item.

A non-transitory computer readable storage medium comprising instructions that, when executed by at least one processor, cause a computer device to:

• receive a content item from a content provider for delivery to users by way of a digital content distribution platform;
• determine expected delivery metrics of the content item by way of the digital content distribution platform;
• utilize a first machine learning model trained to optimize content delivery for a first target event to determine probabilities that the first target event will result from delivery of the content item to corresponding users;
• utilize a content selection process to determine whether to deliver the content item based on the determined probabilities that the first target event will result from delivery of the content item;
• determine that actual delivery metrics of the content item fail to satisfy the expected delivery metrics of the content item;
• based on the determination that the actual delivery metrics of the content item fail to satisfy the expected delivery metrics of the content item, utilize the first machine learning model and a second machine learning model trained to optimize content delivery for a second target event to determine weighted sum probabilities that the first target event or the second target event will result from delivery of the content item;
• and
• utilize the content selection process to determine whether to deliver the content item based on the determined weighted sum probabilities that the first target event or the second target event will result from delivery of the content item.