In another “Grab Bag” video, Matt Cutts takes a question from England regarding image address. Paul Randall wanted to clarify whether the text of the URL is considered when generating search results for a relevant query. For example, can “” turn up if the keyword is “lolcatz”?

Matt points out that it is incredibly time-consuming and resource-intensive to conduct an OCR on all images in the Google Index. Therefore, they often rely on the available information surrounding the image to make inferences about its nature. These include the metadata, the nearby text, and yes, even the URL. The algorithm crunches all of these to ensure that the image will only be served in related searches.

Now some people might be tempted to use this information to game the system. For instance, webmasters could use layers of subfolders named after the target keyword like so: “”. Matt warns that he and the webspam team have already taken steps to combat this kind of brute-force tactic as well as others. Rest assured that as long as the URL does not indicate abusive behavior, it will be considered in the rankings.

Video Link:
Does Google consider the URL of an image? –