How does that work? Well, you might have noticed that Google Images doesn’t just return pages from where a picture originates; for instance, querying Google Images for site:google.com, Google among other results shows a page from bbs.mydriver.com... which simply includes a JPG from somewhere at groups-beta.google.com. When you click-through to that page, it will then not be the origin site’s ads that will be displayed, of course, but the ads of MyDrivers.com, even though they likely don’t own the pic.
Michel, who tried this out, says “I do one of this hotlinking with one flickr image in a new domain (3 weeks, no backlinks) and in one week I was ranking #3 for that keyword in [Google Images].”
Case in point – do an image search for paris in Spanish Google, and of the top 5 results, all but 1 are hotlinking the image. Even when you do a Google web search, thanks to the direct image results box on top of the web results, a hot linker ranks top for paris. Michel says, “I think there are zillions of examples...”
Now, these aren’t all highly optimized spam sites, but partly just your average innocent blogs (and some may even hotlink because they pay the origin domain for hosting, though that’s certainly not true for all hotlinkers). Indeed, hotlinking an image without asking exposes your site to the risk of defacement; the image owner just needs to replace the picture with something shocking, and it will then display on your site.
Still, it seems a little wrong for Google to return these context pages before the origin HTML page, in those cases where there exists one. Of course, Google may not know the origin page at all times (I cross-checked this one one of my domains, where images were indexed due to hotlinking from other domains, while my context HTML page wasn’t yet in the Google index). Even then, Google could down-rank those results which just hotlink something. Then again, even if you copy an image to your own server, it doesn’t mean you’re always the photo owner...
>> More posts