receiving, by a computing device used by and local to a user, first media data having first characteristics of contextual characteristics inclusive of a type of media, and a resolution and frame rate of the type of media;

determining that the user chooses to view the first media data exclusively on the computing device, instead of one or more additional computing devices locally available to the user, due to the first media data containing the first characteristics preferentially viewed by the user on the computing device;

receiving, by the computing device, second media data having second characteristics of the contextual characteristics different than the first characteristics;

determining that the user transmits the second media data to a secondary computing device of the one or more additional computing devices;

determining that the user chooses to view the second media data exclusively on the secondary computing device, instead of the computing device, due to the second media data containing the second characteristics preferentially viewed by the user on the secondary computing device;

correlating that the user views the first media data on the computing device because the first media data includes the first characteristics, and views the second media data on the secondary computing device because the second media data includes the second characteristics, wherein the correlation is stored on at least one data source;

commensurate with performing the correlation, classifying content portions in each of the first media data and the second media data based on the contextual characteristics identified within the content portions, wherein the contextual characteristics further include identified persons, places, objects, sounds, and utterances within the content portions, modifications made to the media data and an application rendering the media data by the user while consuming the content portions, a color composition, shape, and texture characteristics of the content portions, audio signal properties of the content portions, and a type of the various forms of media data making up the content portions;

detecting third media data having a new content portion is received by the computing device;

identifying the contextual characteristics within the new content portion;

selecting a storage scheme for the new content portion according to a predicted utilization of the new content by the user based on the correlation in the at least one data source associated with the consumption of the content portions by the user, wherein the correlation has a confidence level that exceeds a predetermined threshold, wherein at least a first storage scheme comprises storing, on the computing device, a reduced-size version of the new content portion according to content optimization settings of the computing device, and wherein the storage scheme is further selected according to technical specifications of the secondary computing device utilized to render the content portions such that the new content portion is stored on the computing device irrespective of whether technical characteristics of the new content portion, as stored based on the storage scheme, exceed technical specifications of the computing device yet align with the technical specifications of the secondary computing device;

causing the new content portion to be stored on the computing device according to the selected storage scheme; and

responsive to detecting that actual utilization of the new content differs from the predicted utilization, restoring a full size version of the new content on the computing device.