Anchor Text: You Must Kill Your Darlings
Three years ago, if you had told me to link build without a sole focus on competitive keywords as anchor text, I would have written you off as destined for the asylum. Keyword research takes too much time and effort to miss an opportunity for exact keyword links, right?
Enter Panda. Google’s Panda came down like a battle axe and smote low-quality content farms, leaving them scrambling to recover. Who knew such an adorable constituent of Animalia Chordata would become the mascot of an especially brutal SEM blitzkrieg?
In March, Google published a list of search quality highlights, one of which reads:
Tweaks to handling of anchor text. [launch codename "PC"] This month we turned off a classifier related to anchor text (the visible text appearing in links). Our experimental data suggested that other methods of anchor processing had greater success, so turning off this component made our scoring cleaner and more robust.
Several SEOs have waxed philosophical about the meaning of this cryptic paragraph, listed in Google’s March 2012 Changes. The change is hidden in a list of 49 other updates, some of which seem comically trivial (the tennis scores update – Codename “Double Fault” is going to change everything!), driving the search world mad with implications.
One thing is for sure – Google wants to get away from its dependence on anchor text. Whether Google is taking action now or later is negligible; it’s not as if Google will provide the SEO world with more descriptive warning, if any, than the paragraph above. With this in mind, it would behoove any search marketer to focus more on the quality of content in which your link floats, like a cute little buoy in relevant waters.
This will be a jagged little pill to swallow for companies in competitive marketplaces and passive income entrepreneurs. Campaign managers will cling to their too-precious anchor text without realizing how abusive their relationship with their links has become.
Y’all – it’s getting out of hand. And it’s only going to get uglier.
I’m a proponent for sustainable SEO – it makes me unpopular at conferences and in circles of SEOs who guffaw about their newest one-trick-ponies (exact match domains, anyone?) and getting the most out of them while it’s still possible. This sort of attitude is, quite simply, reckless.
A search engine specialist is nothing without his or her ability to adapt, and thus, preventative measures should be taken where consequence is foreseeable and inevitable.
I challenge you to take one week – just one week! – out of your link building endeavors to kill your darlings and take your anchor text blinders off: diversify your anchor text, improve the relevance and usefulness of the content in which your link resides, and submit fully to the flow of the content and external linking formats of link hosts. Pay more attention to directory linking, for the sake of a balanced diet.
Even if you don’t see any improved ranking or ROI, you will have done yourself a gigantic favor in link emergency preparedness for future algorithm updates.











I love the 1 week challenge. Making your anchor texts be more natural and diversified can really help you a lot. Many SEO even suggest to add at least click here or learn more anchor texts just to make your site be more natural.
Excellent post. Hard to swallow for some I am sure, but right on the money I believe. Another magic bullet bites the dust. Not it is just another bullet in the chamber…one of mnay in link building.
This is sure to be a tough pill to swallow for some seos. Nonetheless, so far it’s pretty clear that sites that contain more brand and “junk” anchors appear to out perform those with exact and even partial match anchors. I also think more attention should be paid to the context of text surrounding links.
@jaysen – Taking the challenge?
@Kevin Burke – “Another magic bullet bites the dust.” Well said.
@Gyi – You’re absolutely right about the context of surrounding links. If you’re sending a message and including external links, those links are contextually a part of that message; I think engines pay more attention to that than we might think.
Great Article!! Although you might want to update your thought about Directories… check out the latest Google Update here… http://www.seroundtable.com/google-directory-removal-15151.html
What I’d like to see a study on (or even hear opinions on) is how Panda handles exact anchor text which is also an exact match for the domain? In general branded keyword anchor text seems to actually give a lift, indicating that your brand is popular. What happens when your brand, your “best” keyword, and your domain are all the same keyword? Naturally people linking to your site are going to overwhelmingly also use the anchor text which is a very targeted keyword, in addition to being your brand and domain name. Does the brand/domain lift counter the fact that the majority of your anchor text is also a particular high traffic/conversion keyword? I find it hard to believe that the algo would be set such that my rankings could drop based on the fact that if my domain is examplekeyword.com and my brand is Example Keyword that say 50-60% of my anchors are also a version of “example keyword”. In the case of an EMD, that’s a natural link profile, where without an EMD that would scream overt link building. Thoughts?
I’m not sure I understand your question entirely, but if it’s what I gather: I’ve not seen any studies on this, but you pose an interesting quandary. In some verticals, I don’t see the point of avoiding keyword stuffing when there are only so many different synonyms for things, such as “carpet binding”, nor why any company would call itself “Carpet Binding” if not for the point of EMD/SEO benefit. The exception to this, I feel, would be in local search, where EMDs coupled with branded link building just makes sense. It’s more conceivable for companies to brand locally in some instances (i.e. Kernersville Dodge, Lancaster Archery) but I would be very interested in seeing case studies for generalized, international companies with EMDs.