Good Bye Anchor Text, Welcome Co-Citations: Revolution in Link Building for 2013 Revealed

SMS Text

goodbye anchor textThe Google Penguin update has changed the link building process forever. Gone are the days when it was easy to rank for desired keywords by building tons of links using exact keywords as the anchor text.

In the wake of Penguin, uncertain exactly where the thresholds and ratios are. Just how well distributed should our anchor text be in order to get the best results?

We have been busy with research on the subject and, not too long ago, discovered what Rand Fishkin at the SEOMoz blog had to say for their Whiteboard Friday series. Much of what we’ve observed has fallen in line with his predictions.

So why might Google think about diminishing Anchor text?

It’s too easily manipulated.

No matter how the link was obtained, you clearly have more control over the anchor text if have direct control over the link. Google isn’t, for the most part, interested in links that you have built yourself. There are exceptions. For example, if your reach is big enough to get links from prominent sites on the web, links that you directly build as a result are a genuine indication of your influence on the web.

But, in general, Google wants to see sites linking to you simply because they like the content, feel it’s relevant, and want to share it with their audience. A link with highly optimized anchor text simply doesn’t look natural. It means there’s a much higher chance you had direct control over the creation of that link. If most of your links have optimized anchor text, it means you probably built most of them yourself.

What are co-citations?

I found a page on that gives a nice definition of co citation.

“Bibliographic Co-Citation is a popular similarity measure used to establish a subject similarity between two items. If A and B are both cited by C, they may be said to be related to one another, even though they don’t directly reference each other. If A and B are both cited by many other items, they have a stronger relationship. The more items they are cited by, the stronger their relationship is.”

We believe that Google is placing less emphasis on anchor text (and perhaps links in general), and is leveraging it’s massive data set to look for statistical correlations that it can use to determine relevance and influence on the web. Links are a good proxy for this, but as the search engines get smarter they will get better at measuring “buzz” unrelated to links and especially anchor text.

How to earn co-citations?

  • Stay up to date on trends in your niche
  • Produce original content on the subject that hasn’t been covered extensively by others
  • If your content makes waves in the industry, it will influence the direction of posts made by others later on. Even if they don’t link to you directly, your impact on the industry will in some cases be measured by Google. The search engines will only get better at this in the years going forward.

Here’s an example. After the recent Google algorithm updates, “Post Panda/Penguin strategies” became a topic that people were looking for very frequently. As a result, that’s what we wrote about our earlier post. Here it is.

Afterward, we got a lot of natural citations from third party websites where people were talking about link building strategies that apply after the Google Panda and Penguin updates.

Here is an example from Quora:

Image1 (1)

Here is the citation we got:


As you can see, this link doesn’t use anchor text at all. However, the entire page is about post-penguin/panda strategies, and Google is increasingly using this kind of information to determine what the linked page is about.

Why would Google give more importance to co-citations than anchor text?

Simple. Logically, it’s harder to manipulate co-citations. You cannot have great co-citations unless you have actionable content which people love to read and share. Also, you cannot manipulate the industry influencers who make your co-citations possible. They only promote the content that they feel is best to share with their readers.

If you have an actionable post then you don’t need to manipulate things. You’ll get citations much like the one above.


It is too early to say for sure, but we believe Google is not only pushing away from anchor text, but toward co-citation as one of the strongest relevancy signals. Co-citation, if measured properly, gives the search engine much more data to work with and will allow it to take the temperature of the internet community at large. Anchor text, on the other hand, is a small html quirk that few people even know how to use properly.

So do you think Co-citations will be replaced by Anchor text in near future? Let’s talk about it in comments. I would love to hear everyone’s thoughts on this.

Image Credit: Shutterstock / woaiss

Pratik Dholakiya

Pratik Dholakiya

Co-Founder & VP of Marketing at E2M
Pratik Dholakiya is the Co-Founder and VP of Marketing of E2M, a digital marketing agency and, a mobile apps development company. Pratik has been featured on sites like Forbes, Moz, SEW, SEJ, KISSmetrics, Entrepreneur and FastCompany to name a few. Hit him up on Twitter @DholakiyaPratik for a quick chat.
Pratik Dholakiya
Get the latest news from Search Engine Journal!
We value your privacy! See our policy here.
  • Will

    It’s all well and good saying Google is doing something, but where’s the proof?

    Until Google can crack sentiment analysis, co-occurrence, which is what you refer to here, will only remain a weak signal.

    Why? Because Google’s spiders can’t tell if someone mentioning a site in relation to a certain keyword is good, or bad.

    If co-occurrence was a major raking factor, it would increase the likelihood of widely criticized sites ranking highly.

    Google will stick with links and anchor text until it can develop 100% reliable sentiment analysis which can be thought of as the do follow and no follow attributes of co-occurrence.

    • Tommy Redmond

      In asking “where is the proof, ” you bring up some valid concerns. However, you can’t say for sure whether or not Google has ways of factoring in sentiment. In fact, Google has been awarded at least 1 patent for sentiment analysis (

      I think the more interesting and pertinent issue here is that there seems to be a lot of general confusion in understanding the difference between co-citation (which the author correctly defines) and co-occurrence (which is what the author is using in most of the examples in this post).

    • http://http// Gary

      Swill, I have to disagree with your last point, “Google will stick with links and anchor text until it can develop 100% reliable sentiment analysis” A few years ago this may have been true, however in the last 2 years or so Google has rolled out changes to SERPS such as Google Local which was a poor product and I still think is too this day, it had little data and lots of fake reviews and Google still saw fit to place it right up there with everything else, they knew it was a poor product a knew that if they showed it in a high position in SERPS then people would come and fill their own profiles in and make the service a success. This goes against everything Google allows US to do. They say if its not good enough for the public then why should we rank it.

      The point I am making is Google will happily roll out a bad algo these days, they did it with Local, they did it with Penguin and Panda. So no reason they would not do this as well. not to mention ever time a good site gets hit they are forced to use the adwords cashcow to supplement temporary mistakes that can take many months to fix.

      There are so many ways to manipulate this as well.

      Here is what I think should have been done a long long time ago.
      Tell everyone how to rank, its always been the spammers who learn how to manipulate it first, it will always be like this. That is the nature of the internet.
      If the ranking factors are genuine then it would be much harder to rank a spammy site especially if genuine sites are also in the loop on how to rank well, they would have competition immediately and with a much stronger site. Simple logic.

      It seems to me there will be lots of fake citations coming from sites which are run by spammy authors that create tons of content because they have cheap access to it. They create the appearance of a strong author/brand that really is just a large network of sites created in some far away offshore country and they will just pump citations out for high prices.

      Google’s customer service is so poor, if it took the time to create a paid for system that would verify certain types of business as genuine, I know many would pay. This would go towards the cost of their employees to determine if the business is registered really has a fixed address etc.. This would entitle those users to genuine help when SERPS positions fall. As it would be in Google’s best interest to serve traffic to businesses that they know are indeed genuine and additionally have good reviews, social etc.. This does not work for all but that is Google’s problem, they have a one size fits all solution that simply does not work.

      Right now one of my competitors has 38 Google Review, the first 32 are fake and the remaining are complaints that the company s terrible all asking if the reviews below could be legit., The whole industry knows the company to be a scam they manipulated SERPS and remain top for major keywords. There is no way to report bad natural SERPS results so you are unable to inform Google about the bad job they are doing.

      95% of our customers are using Google so it is not as if I can even go elsewhere, so I am stuck with a search engine rife with spammers and scammers and waiting for the next ridiculous update that all the SEO scammers will learn before any of us, rank a site for a few months knocking all the genuines sites off top again damaging the industry until they go and we all get 1 month back at the top before the next update and the next scam site.

      Google is a total joke. they are the reason for almost ALL SPAM ONLINE because they are unable to think clearly about people’s motives to game the system.

    • Ross Dunn

      I understand you said it couldn’t be a “major” ranking factor but your last paragraph was more absolute towards co-occurence not being used at all. I would just counter that Google is likely using variations of co-occurence weight in their algorithms to build reliable results using a better blend of links, anchor text and co-occurrence. So to say this ranking factor isn’t being used is not likely considering Google’s continual testing and improvement of their algorithms.

      Of course all of this discussion is still based on conjecture – albeit conjecture amongst enthusiasts and experts in the field. Personally I believe co-occurrence has been in use at least as long as citations played a role in local results – at least it seems logical.

    • Naveen kumar m

      Hi Will

      Pratik was given such a good information on anchor text but how we can stop the work on exact anchor text and continue with co-citations. Yeah exactly we didn’t have any proof to consider the co-citations.

  • Kenneth A. Steele

    I tend to agree with Will a bit here simply because I believe relevancy between anchor text, link url, page content and keyword density will ALWAYS have a direct effect and indicate to Google spiders how to index you properly. Co-ciatations (for lack of better terms) are something the blackhatters and seo junkies have been doing for years (esp. in the adult industry) by buying up hundreds, even thousands of domains and publishing amongst multiple platforms and servers on a global scale to create their own web portals and hubs, all interlinking using controlled backlinks. This convo is moot.


    • Kenneth A. Steele

      Forgot to mention one more important factor; that being, most of (if not all) the backlinks you acquire using your method use the “nofollow” parameter anyway… so, calc the time, effort and money spent to get these links and measure against building out = rethinking the future of your web dev plan.

  • Stu Morris

    Have to agree with Will and I quote PD “It is too early to say for sure, but we believe Google is not only pushing away from anchor text,”. Co-Citation has been around for years and actually an SEO Strategy discussed for years also. I believe Co-Citation is a factor of many, of which can be manipulated as easy as anchors, links, and social. Do I have an answer, NO, I can only say do the best you can with all the tools in your box. You can’t build a house with just a hammer. The more quality work “nails, lumber, paint” the longer it out last it’s neighbors house.

  • Aaron

    It would be foolish to think Google is not well under way with its implementation of a more advanced mechanism for measuring a pages authority on the web. Co-citation sounds to me like a very possible avenue and one I am sure they are working on right now.

    It will be some time before anchor text is no longer a strong indicator to search engines, however we must not forget of all the other key ingredients that go into any well thought out and structured SEO strategy, on-page cannot be overlooked and of course the all important social signals.. which in my opinion holds far greater weight than many people think.

  • Stacey Mayer

    I think it’s a good move, as those of us who are producing original, useful content, will still be reaping the most rewards, while those just adding links will be found out.

  • Dave Davies

    Interesting idea – so one might argue that if would benefit a newer site with little co-relation to the top dogs to build links to the top rankings sites on the same pages that they’re getting links on for themselves.

  • Mary Grace Viado

    This article really made sense. Close-minded and black hat practitioners may not embrace the idea but this is a reality check for every internet marketers, especially the SEOs. Although, there’s no direct effect seen, as of yet, Co-citation could really be a major ranking factor in the essence of being authoritative on the field. We all know that people mention powerful personalities because they’ve established trust and excellence in a certain field. These people (Powerful Authors), are change agents who never stops looking for ways to keep up with the now smarter and stronger search algorithms. Co-citation is also a fail-proof to having quality and evergreen content; which information seekers are most interested in.

  • Subhash Jain

    SEs crawl websites through hyperlinks. The anchor text in the link is the deciding factor about the relevancy of the landing page. There is definitely a change in the Link Building strategy. “Content Marketing” is the new “Mantra” of SEO. This my perception of Google thrust on 2013 Link Building strategy.

  • Mike C

    Speaking as someone who occasionally dabbles in Blackhat (Not professionally, just on my own unmonitized sites so I know how it works and can effectively analyse competitors doing it. No point going all PI on me, Mr Google Human Reviewer) this matches up with the strategies that seems to work currently.

    Where this falls down as an anti-spam technique, of course, is that there are plenty of blackhat techniques where the spammer has control of the entire pages content. Spun article syndication, for example. Or automated forum profiles. I don’t think Google can class content on a sitewide basis – many legitimate sites are ‘about’ more than one thing. What is the BBC ‘About’ for example? And page-by-page stuff, like I said, can be manipulated.

    Cat, meet mouse. Mouse, Cat.

  • Tomer Harel

    Hello Pratik, thanks for a great article.

    Personally I don’t think Google will be ditching anchor text, but there is no doubt in my mind they will get smarter at knowing when to make it count and when to ignore it.

    For example, getting a link with exact match anchor text from a high authority site should still count because there are much lower chances for manipulation.

    On the other hand, co-citations can come from regular people, not just known social influencers and that opens a door for manipulation, similar in a way to the old forum spamming techniques.

    Bottom line, this is far from being a black & white world and the grey zone is probably larger than both ends combined ūüôā


  • John OK

    The points raised in this post are definitely worth considering . I can’t see Google being overly concerned with anchor text. Using optimized anchor text will do no harm, But by generally producing quality content co-citation’s should follow.

  • Noman Ali

    Well this need huge analytic skills, and right now people around 80% didn’t think about this, Yes if Google replace this thing all over than it will be huge task for all Marketers. And i think Google will not change this huge thing at once.

  • Fabio Buda (netdesign)

    Co-citations are important for search engines not only to discover the relation between resources but also to better understand the topic of the pages. It is one of the most important steps in NLP.
    By the way, great article, thank you for sharing your SEO know-how with us!
    Have a nice day!

  • Scott McKirahan

    Well, I’m not sure if your SEJ example is any more an example of co-citation than it is of an article written about a subject ranking well for that subject when relevant websites link to it – regardless of what the linking website was writing about (as long as it is somewhat related). I do agree that anchor text is pretty meaningless and that the relevance of the referring page/website combo carries much more weight.

  • Asif Anwar

    Anchor Text is actually dead, in my sense. It’s not sure if it is punishable (i.e. you get negative points for abusing it). But, the debate is more on whether:
    1. It does not work any more (you don’t get any points)
    2. It does not work as good as before (you don’t get much points as you did in earlier stage of SEO)

    So, the later 2 needs to be tested before we can say that Anchor Texts punishes you for abusing it.

  • Foxcrawl

    I do not see anything suspicious. In the end co-citations represent nothing else than the backlinks to an article/post. Isn’t that what Google mainly used so far for ranking? well, maybe now the algorithm pays less attention to the keywords because of the stuffing.

  • Jared

    Anchor text isn’t dead and I don’t expect it to die anytime soon. I do agree though that co-citation could be the next big thing.

  • John

    Co-citations may be the next best thing…but we have to wait and watch for Google with their next update.

  • kshitija

    I agree with Jared. I dont think anchor text can be dead completely. But we should try to move out of the exact match keyword anchor text. Instead, we should dilute it by incorporating long tail and branding elements.

  • Krishna

    Hi Pratick, Thanks for the post, I am almost spell bound by the new tactic for ranking which may be implemented by google. I think it is one of the best ideas I have heard over along time, Really man this tactic will make the world hell for Black hat technique users.

    I am looking forward for Google to implement this technique just to see how it will effect the rankings of sites continuously using Anchor text or Keywords to get links and ranking……


  • The Viral Marketer

    Co-citation is a nice idea and you’re right in saying that it looks far more natural than anchor text simply because people do share links in this manner naturally through social networks for example, however, for now I think anchor text is still a very big thing.

  • Paul

    I am looking for real proof here. Where is it? Your post is good and it conjures up lots of question in my brain. Actionable content…it wouldn’t be long before spammers find ways to mass produce “Actionable Content”. How long before co-citation gets over abused and Google decides to no longer use it in their algorithm? What will replace it? Will we we have to start all over again? There will always be an ongoing battle between all search engines and the spammers.

  • Szilard

    Placing your keywords in the anchor-text of the inbound links is an allowed, natural, normal SEO technique. Absolutely nothing wrong with it. Why is everyone talking about it as if it was a devilish no-no?!
    The co-citation sounds good and even more natural. However, some questions pop into my mind: 1, All the no-follows will be in vain from now on? 2, Does co-citation work in the comments section of a website? If it does, is Youtube becoming the nr.1 link-building platform? 3, If I create a buzz in a niche forum using a keyword and a website’s brand name and responding from several accounts, does that equal to the “future of link-building” as people mention co-citation nowadays? This could be manipulated and abused even more easily than anchor-text spamming.
    Informative article, appreciate it!

  • Bryan

    First off, great article! I enjoyed reading it. I think anchor text will always play a role in SEO, but I do think that it’s importance is fading when trying to acquire good rankings for specific keywords. As you pointed out, anchor text is too easy to manipulate, and while it has always been one of the most important ranking factors in Google’s algorithm, I think Google is too smart will indeed find un-manipulative ways (co-citation) to rank sites. We’ll just have to wait and see…

  • Don Sturgill

    I don’t know that anything is ever the “next big thing,” Pratik. There are many, many factors involved. Surely co-citations are a part of the puzzle, though … and a growing part, for sure.

    The thing that amazes me is how often I get pulled in by headlines declaring huge break-throughs and must-know information. Goes to show how important it is to use powerful words (Revolution) to get Attention.

    My guess is the Revolution has been Revealed already … it maybe even be beginning to die down a little. The immensity of information generated via the Internet is forcing us to either freak out …. or focus.

    And Google is listening.

  • 9Dotstrategies

    Actually, we must use anchor text in some small percentage of our link building.

    Anchor text variations are really helpful since Google have detected those linking using the same exact match anchor text over and over…

    Specially using naked url as your anchor text variations…

    Read more

  • Petric Massa

    Hi Pratik,

    As per your article if google dosn’t like anchor text link building so why are you doing same thing in your profile for “Advance link building” keywords.

    Can you explain to me?

  • faisal

    Basically one should get natural links without any anchor text, this way you can build a better brand too.

  • Mann

    Thanks to Pratik for wonderful share of thoughts on co-citations. Also, I hold on hands with concerns expressed by other professionals for anchor text being the liable source for ranking a page on SERP in present. However their is no full fledged evidence for the workings of co-citations in actual s.

    As per my understanding, co-citations follows the flow & link juice like SMO works. The more the number of people sharing or commenting onto your post, the better is your site or content visibility on SERP. The author, could urge more & more profiles & can have better presence on web for their post.

    Please correct me if I am wrong?

    Lets wait for coming furture how Google could help us in exploring this more effectively in future times.

  • Tom Dupuis

    Nice article Pratik. Quick question that I didn’t see here:

    What % of links do you recommend having an exact anchor text match of your primary keyword?

  • Yogita Aggarwal

    Hey Pratik, Well informed post. I just want to ask one thing that use of supporting keywords as prefix or suffix with the main keyword would be harmful or not.

    For Example:

    technology news – main keyword
    sensational technology news – modified keyword

  • Bruce Jones

    Thanks for giving us new things to think about, I think the ratio of anchor text compared to brand links is what to watch for.

  • Richard Takemura

    Thank you for the info Patrick. It actually has me a little more confused than I was before. Just going to have test I guess.