Received an email today via a private discussion group about some hotlinking and Google Image spamming for the term “hawaii pictures.” If you search for the term on Google, you’ll see that Google is serving three image results above its normal web results.
A mouseover on the images results in some fishy target URL’s (truncated for publishing purposes) :
http://images.google.com/imgres?imgurl=http://www.people.fas.harvard.edu/~junliu/Photos/hawaii.jpg
&imgrefurl=http://www.hawaiispecialoffers.com/&h=676&w=576
After clicking on the images, you’ll see that the user is taken to parked domains where the images are not even present; but plenty of Google AdSense ads are. One of the images is being pulled in or ‘hot linked’ from a Harvard.edu website somehow to a parked domain.
So, how are these spammers gaming Google? My buddy Everett stepped in to explain :
The folks over at http://www.hawaiispecialoffers.com/ are hot linking to an image from people.fas.harvard.edu so Google is showing you the hawaiispecialoffers.com domain in the iframe, but if you click to see the actual picture on it’s URL you go to the Harvard subdomain. Normal, but I hate bandwidth thieves who are to damn lazy to just save the file and upload to their own server.
But then another group member replied:
Yes… But… The image is nowhere to be found on HawaiiSpecialOffers.com
They have hot-linked an image from Harvard.edu to get a parked domain in the top position in Google… Without even displaying the image on the website.
I don’t see the image on the page, or even a link to it in the source code. But… There it is, driving traffic to a parked domain for a competitive keyword.
Two out of three of the images displayed in the results page use this technique. (Whatever it is)
Harvard.edu also has much more Google Juice than some parked domain, so by piggybacking the Harvard.edu domain, the images are appearing as top images in Google search. This is a popular keyterm and the image search spam is driving users to Google AdSense arbitrage/spam sites, where users are probably clicking on these ads. A classic Google AdSense scenario.
Does anyone know how this is being done and hey Matt Cutts, can’t Google do something to stop this trickery?
Wow – That sucks! I’m sure the people who click on the images are disappointed as well.
It’s a neat trick though… Any ideas on how it’s done?
Those lazy bastards are hotlinking to all kinds of images from different domains:
http://www.eastes.net/content/files/adventure/hawaii%20beach%204.JPG
http://www.travelteam.org/images/Maui_bch.jpg
…etcetera.
The rainbow pic:
http://www.people.fas.harvard.edu/~junliu/Photos/hawaii.jpg
Was probably an image they were previously linking to, or perhaps even using the good ‘ol 302 hijack that was supposed to have been fixed up five years ago.
Hey Hawaii Special Offers – You have a server to put files on – USE IT you thieving jerks!
Dave,
Try putting an image from one of those harvard pages on your site and instead of linking directly to it use a 302.
See what happens.
I don’t think I’ll be doing anything like that… I was just wondering how it’s done. (It’s like watching a magic trick)
In this case… It looks like a dirty trick and I don’t want to get myself in trouble.
Everyone always poo-poohs this idea…but I’m telling ya: Many of these spam problems could be solved quickly and easily with a well-applied machine gun! ;-)
Being some that has parked domain myself, it is against the TOS of domain parking companies to drive traffic to parked domains.
Damn! Thats a neat black hat trick for a highly competitive term. Too bad their adsense account will be shut down… lol All that money is going to be GONE! lol
We hould start doing that and get important traffic from Google in that way. http://www.paroles32.com/paroles/tamia/index.php
Or no?
nice!
Look at the source. Kinda lazy coder. Cut and paste from Dynamic Drive. Cut and paste from bravenet. Cut and paste Adsense blocks, analytics, etc…
Figures that he’d/she’d hot link images.
But where did they originally find the images?
Google image search, perhaps?
ROTFLMAO!!!
I asked the image team about this, and it looks like the images were there when we crawled, so 1-2 of these images may just be slightly older/stale data.
If someone is thinking that more nefarious stuff is going on, I’d restate the problem in simpler words for my puny brain. :) Something like, “This image exists on domain A, but domain B is trying to get credit for the image and show up instead.”
Thanks Matt!
I thought it might have been some sort of cloaking but that’s a much simpler explanation.
(It was cached on Jun 30, 2007. )
Hi Matt,
About image hot-link spam…
So… why not display the page where the image is originating from versus the page that just has an image hot-linked.
It stands to reason that the Harvard.edu page should be more relevant and more deserving of being listed in the results and not the parked domain.
Or… Why is the parked domain that was only hot-linking to this image more relevant than the original page from Harvard? How does sending people to the parked domain versus the Harvard website create a better user experience?
Just an update… The page was recently cached on 8 Jul 2007. The image is still there.
I’m just trying to determine how long this phenomenon takes to correct it’s self.
Cashed again on July 11th. The image is still there.
Now that the page has been visited twice by Google, after the image was removed, we can likely assume that there is some sort of disconnect between when a page has stopped using an image and when Google refreshes the Images database to reflect the change.