Anchor Text Test: Making The Most Of A Link Opportunity


Link building may not be rocket science, but it can be difficult. It’s not always easy to find “authority” sites in your niche that are willing to link to you – and when you find someone with a relevant site/blog who is willing, the link may not be on your terms. Likewise, while guest posting is a great way to build links, best practices change frequently and it can be tough to keep up.

Despite these inconsistencies, however, one factor remains that contributes greatly to any link building initiative and that factor is anchor text.

The importance of anchor text when link building

When link building, anchor text is key. Ideally, anchor text should take the form of keywords for which you want your site to rank. However, since other sites will invariably link to you, it can be difficult to control how these links are anchored.

Say, for example, you run an online snowboarding store… We’ll call it snowstuff.com. You want your store to rank well for relevant terms in your niche – terms like “snowboards” – and the way people link to you matters. Ideally, you’d have links from many “authority” sites in your niche and most of them would use the anchor text “snowboards.” Unfortunately, however, some people who link to you may use some other less-than-desirable anchor text, like “snowstuff.com,” or even “click here,” neither of which help your site rank for important keywords. You may try to convince them to change the anchor text, but this isn’t easy, and doesn’t always work. This may sound negative, but bear with me – I’m trying to set the stage.

How to successfully leverage anchor text

In the instance you are in complete control of the manner in which a webmaster links back to your site, you want to make the most of it by leveraging anchor text properly.

For example, let’s say you’re doing a guest post on a 3rd party blog to promote your online snowboarding store. Given the opportunity, an inexperienced SEO may feel inclined to include a plethora of backlinks to their site, using several different terms for which they want to rank, like “snowboards,” “used snowboards,” “cheap snowboards” and “snowboard gear,” all in one blog post, and all linking back to the same URL (snowstuff.com).

I have seen this strategy employed and was skeptical about its effectiveness. Now, a recent test done by Wordstream.com has confirmed my suspicions: multiple links to the same URL originating from one page is not effective. To be more specific, Google only considers the anchor text used in the first instance of a link to a URL. The anchor text used in subsequent links to the same URL is ignored.

What does this mean? Well, in the guest post scenario detailed above, only the “snowboards” link to snowstuff.com would be counted by Google. The other 3 would be ignored, and snowstuff.com would see no positive growth in the SERPs for the keywords “used snowboards,” “cheap snowboards,” or “snowboard gear”… At least not as a result of those 3 particular backlinks.

Does this mean you should only use one link to your domain in a blog post? Not necessarily.

Testing the effectiveness of multiple links within a post

By now, it’s understood that linking to the same URL more than once in a post does little to help a site’s ranking. However, many sites have more than one page they’d like to rank.

For example, in addition to its homepage, snowstuff.com might have a landing page for snowboard bindings they’d like to rank as well. Would it be effective to link to snowstuff.com/snowboard-bindings (using the anchor text “snowboard bindings”) in addition to snowstuff.com (using the anchor text “snowboards”) in the same blog post?

I decided to throw together a little test to find out.

To start, I created a little div at the bottom of Wpromote’s “About the Team” page. This div contained a blurb with 3 links going to johnvantine.com, a domain of mine that has more or less been gathering dust for the past few years. I chose 3 unique anchor texts that, as far as I could tell, didn’t occur anywhere else on the web.

Initial anchor text test

A screenshot of our anchor text test, which temporarily resided on the Wpromote About Us page

As you can see in the ridiculously good looking diagram above, the first link (“j0hnv4nt1n3.com”) pointed to johnvantine.com. The second link (“wpr0m0t3″) also pointed to johnvantine.com. So, based on what we learned from Wordstream’s test, the second link should be ignored by Google, because it points to the same URL as the first. The third link (“mamachajj”) pointed to johnvantine.com/anchor-text-test-landing-page/; this link was the test variable.

So what happened?

I setup Google alerts for all 3 keywords/instances of anchor text used in the test blurb. It took Google about a week to detect the text on Wpromote’s “About the Team” page, and another two weeks or so for the links to be followed and indexed. When they were finally indexed, a search for the first instance of anchor text revealed the following result:

What happens when we search for our first keyword, "johnv4nt1n3.com"?

What happens when we search for our first keyword, johnv4nt1n3.com?

As you can see above, a search for our first keyword, “j0hnv4nt1n3.com,” yielded 2 results: a link to it’s non-leetspeak counterpart, johnvantine.com, and of course the wpromote.com page where the anchor text originally appeared.

What about our second keyword, wpr0m0t3?

What about our second keyword, wpr0m0t3?

A search for our second keyword, “wpr0m0t3,” only returned the wpromote.com page where the keyword appeared. So this confirms Wordstream’s findings: only the anchor text from the first instance of a link to a URL is counted. Am I starting to sound like a broken record here?

...And the third keyword, mamachajj?

...And the third keyword, mamachajj?

Our third keyword, “mamachajj,” yielded 2 results: the page on johnvantine.com I created for it, as well as the wpromote.com page where it originally appeared.

SIDE NOTE: In viewing the search results, you’ll notice Google replaced the title tags for johnvantine.com with the keyword being searched in each instance. Had I provided the keyword in the copy on the site (or in the title tags perhaps), I don’t think that this would have happened.

The final word on multiple links = good news

The results of my test indicate that using multiple links within a post works. While linking to the same domain multiple times using different anchor text within a post does not contribute to ranking beyond the first instance of the link, linking to different pages on the same domain using different anchor text does help ranking, and can be leveraged accordingly!

So the next time you’re building links for your own site (or for a client), keep in mind that more than one link can be effective. Personally, I would use no more than 3 links to the same domain on one page, as the value of links diminish as more of them are present. Of course, if you decide to include multiple links to the same URL on a page, make sure that the first link contains your desired anchor text, as subsequent links will be ignored.

I hope this experiment deepens your understanding of anchor text, and I look forward to sharing the results of future tests. Now get out there and start building links!

  • christian

    Great post John!

  • Jesse

    Wow, I knew anchor texts were important, but that’s a lot of great info. Thanks.

  • http://www.wordstream.com/blog Ken Lyons

    Hey, John.

    Cool anchor text experiment and compelling results. Also, glad you were able to repeat our test results from previous testing. Would’ve been a little embarrassing if you were unable to ;)

    Cheers!
    Ken

  • http://www.wpromote.com/quicklist Chris Laub

    Very very good article. Do you work for google ;) ?

  • http://modelurl.com Ny

    John, great post! The diagram looks awesome. This is definitely important for SEOs/Linkbuilders to know.

  • Pingback: Oatmeal Stout & Anchor Porter | Beer Festen

  • Andrew

    Great post John! This is great information to know – get this guy a raise!

    So now what do you think about some follow up tests?

    1. Are the results the same when you’re talking about internal links? Say snowboardingstuff.com had bindings AND boards on the same page and on the home page there was one link that pointed to that page with the anchor “snowboarding bindings” and one with the anchor of just “snowboards”, would you expect the same outcome, only the first would work?

    2. How about when an image comes into play. Say the first link to snowboardingstuff.com on a third party site is through and image w/ a specific alt tag and later in the page there is a text link to the same page with specific anchor text. Would only the image anchor carry weight and not the text link? What if their positions were reversed?

    Great stuff!

  • Mike

    Terrific post on Link-Building! This will be very helpful for explaining link-building to clients.

  • http://johnvantine.com/ John

    Andrew, those are both great ideas. Marios over at AllThingsSEM did a test similar to the one you mentioned with images (here), but I may have to give the anchored links test a shot. Thanks for your feedback and ideas!

  • Pingback: Pages Linked From Your Syndicated Articles | How To Make Big Money

  • http://www.keralatourismpackage.com Kerala Tour Packages

    hi i was about to link same url with different anchore test from one page..i searched in the web and found your article..very useful for me..thanks

  • http://www.whistler-outdoors.com/ Joe MmacMillan

    Thanks for taking the time to explain this. You are a great help to those of us who may not be up to date on the tech stuff.

  • http://www.sanantoniobarsandclubs.com San Antonio Clubs

    Exactly what I was looking for, I guess it’s gonna take more time to rank my keyword/s but at least I’m using my time more wisely bc of your proven testing.