Methods and systems for prioritizing content are described. A request for a first content object is received at a caching proxy. A search of an access log of the proxy cache for a portion of the first content object is then performed. The size of the portion of the first content object on the proxy cache is calculated and the normally utilized size of the segment of the first content object is derived. A comparison is then made between the size of the portion of the first content object on the proxy cache and the normally utilized size of the first content object requested.