Link Building with Google's Search by Image Feature
Google yesterday announced that it was releasing a few new innovative features to its search platform, including the ability to carry out searches by images, search by voice, and one of the more interesting ones, Instant Pages (currently only available in Chrome’s development version).
We’ll save the write ups of the other two for another blog post, but I’ve just been experimenting with the search by images feature in Google Images, and noticed a couple of ways it could be used for link building. Hat tip to Clarke Duncan on Twitter who first inspired me to have a better look at the advantages this feature could have.
Claiming What’s Yours
There’s a handy link building tip that I came across some time ago on SEOmoz (can’t find the link at the moment), where you look for images that are hosted on your domain, but are being used on other people’s websites. This was good as it would allow you to contact them to ask them to credit you with a link, but unfortunately the Google Images query that let you do this no longer appears to work (please let me know if you have a working one).
Based on this principle is the link building technique of finding people who use your images and getting them to link to you, which is exactly what Google’s new search by image feature lets you do.
Firslty, if you head over to Google Images you should now see the ability to search by images, as shown below:
You can search by image by doing any of the following:
- Enter an image URL
- Upload an image
- Drag an image into the search box
Once you do so, you’ll be presented with search results based on Google’s interpretation of the image. For this example, I’ll use Kellogg’s:
As shown in the above example, there are two key areas of the search results we want to be focussing on: the Visually similar images, and the Pages that include matching images.
Visually Similar Images
Clicking the Visually Similar Images option within Google’s returned results allows you to view a set of images that Google believes are similar to the initial one you entered. Although there are likely going to be a number of images that simply aren’t relevant, there should be enough exact matches and variants for you to work from (providing the image you entered is well recognised/used).
Going through these results allows you to see which domain each of them is on, and to investigate the page a little further for link building opportunities. If a website has used your brand’s logo then there’s a strong chance they’re actually talking about you for whatever reason, and this could provide you with a nice target list of web pages to potentially get links from.
Pages That Include Matching Images
Moving on to the next option in the search results, the Pages That Include Matching Images feature provides a number of results where Google has found your image (or a close match) on – which again could provide a number of good link building opportunities. Using Kellogg’s example, one of the results on the first page is a case study on IGD.com, which talks positively about Kellogg’s but yet doesn’t link back to their website = link building target.
Expanding the Opportunities
The above example of link building using Google’s search by image feature is done so simply by using the brand/company logo, but this can obviously be expanded for a whole manner of other opportunities. If you create your own images and use them online, then finding instances of people using them allows you to gain links as credit for doing so. Similarly, if you have created any visual graphs or other representations of data, this too can be a good opportunity to gain credit for your work.
The above isn’t just useful for link building of course, you can get people to remove your images if they are breaching copyright as Clarke suggests, or if they are using your images inappropiately.
I should also note that the above process could probably be streamlined quite nicely by scraping all of the results for a set image, then checking whether they are linking to you automatically.