Performance-based content delivery

A system comprising:

• a data store configured to store computer-executable instructions;
• and
• a hardware processor in communication with the data store, the hardware processor configured to execute the computer-executable instructions to at least:
• define a first performance category based, at least in part, on an analysis of performance data associated with requests for network resources received by a content provider to form a first performance category definition;
• determine a performance characteristic corresponding to the first performance category, wherein the performance characteristic identifies a characteristic of a subset of client computing devices in a plurality of client computing devices that belong to the first performance category;
• generate a prediction that a first client computing device in the plurality of client computing devices that transmitted a request for a target network resource belongs to the first performance category;
• and
• cause transmission of at least one version of the target network resource to the first client computing device based, at least in part, on the prediction.

A non-transitory computer readable storage medium storing computer executable instructions that when executed by a processor perform operations comprising:

• defining a first performance category based, at least in part, on an analysis of performance data associated with requests for network resources received by a content provider to form a first performance category definition;
• determining a performance characteristic corresponding to the target performance category, wherein the performance characteristic identifies a characteristic of a subset of client computing devices in a plurality of client computing devices that belong to the first performance category;
• generating a prediction that a first client computing device in the plurality of client computing devices belongs to the first performance category;
• and
• causing transmission of at least one version of a target network resource to the first client computing device based, at least in part, on the prediction.