You must login or register!
Inbound.org uses Twitter to register and create accounts. Your Twitter handle will also be your username here on Inbound and registration/login will enable you to submit content, post comments and create/edit your Inbound profile. Use the button below to verify your Twitter account.
Login or Register
I wish there was more detail about the methodology and the charts. Very cool info, but hard to share/sell elsewhere without the background.
Have to agree with Rand about the data. I think Tim is onto something, but I hope he can share more insight into his methodology and the data behind those charts.
A cool test, but really not what I'm seeing in reality. Especially "bombing your site with anchor text doesn't seem to work anymore". Those aren't the results we're seeing, and in localized queries we still see a HUGE percentage (I have no fancy charts or data to support that claim, so take it with a grain of salt) of sites ranking on anchor text alone.
I am a little reluctant to share too much detail, as it refers to clients and the industries they work within. However, I am going to add an update to the post, just detailing the metrics used, some of the keywords and where I pulled the data from. I think if you look at any top 10 results on Google you're going to find exact match anchor text, however when you start breaking up these link profiles you begin to see that actually, these exact links only make up a small part of the overall profile. If you look at the UK results for 'car insurance', Tesco (a huge brand in the UK) are ranking second with little more than a few links from their other sites and a little link building work. Compare that to the competition in that set of SERP's and on paper they should not be ranking, especially not second for such a massive term. I am still looking into it, but in my mind, bombing anchor text is far too much of an easy tactic to rank, we all know this. So why wouldn't Google come along, devalue the exact match, and increase the weight of signals from keyword variations (as they correlate with search data) and general link quality. Food for thought. Rand/Cyrus, if I sent you guys over the methodology behind it, could you do a large scale version of this, my data was based on 20 keywords, I'm sure you guys could pull 1000 different SERP's together :)
Must agree with Rand. More over, I felt also that the amount of keyword tested was somehow too "little". Personally - and this is a suggestion for Rand as SEOmoz CEO - it would be interesting to bring the 12 hypothesis presented by Bill Slawski in his last post and test them.
The post is "ridiculous" according to: https://twitter.com/#!/neyne/status/176592019290984449
Yeah, really hard to apply a theory to "all SERPs," since all SERPs behave differently. Where some SERPs are driven primarily by brand signals and domain level authority (and exact match anchors don't appear to propel rankings), others are driven primarily by temporal factors while some are driven largely by document-level social signals and contextual links.
Wow, lot's of debate. For what it's worth, I liked Tim's article in the first place because it matched exactly what I was seeing online. (I suppose we all see just a slice of the pie, and that's why we need data and scientific method.) I still love the article and would recommend it again.
I love it, you try to suggest anchor text is anything but 'king' and people just can't handle it :) @Tad Chef, the tweets in mention are inaccurate, I never said anchor text wasn't a signal, or that it was a weak one, I said 'exact' anchor text seems to be a weak signal in that particular data set. There is a huge difference. In fact the data I collected showed that relevant variations of the main term seemed to correlate extremely well with higher rankings. Secondly, the data wasn't manipulated to match my opinion, I saw a trend across our clients, pulled the data and thought the fact exact anchor text percentages were really low was interesting. My opinion, before the data was gathered, is that exact anchor text is not as strong as it once was, may not be applicable to less competitive SERP's but social signals, brand signals and other factors may not be as strong in less competitive industries, hence exact anchor text will show as a strong signal. @Ken - take your point fully on board, as I stated in the article, this was across our clients who are in extremely competitive industry. This is not a dig at anyone, but negative comments that add very little to the discussion really bug me. Yes, I can't make a sweeping statement, and I didn't. I only have access to limited data, having said that I work in multiple industries across over 100 websites. When I am questioned by a client, I can't just say 'only Google knows', I have to work with the resources at my disposal. I have no problem with criticism, but people should add to it, what are they seeing, how do they back it up? If more people did this the industry would be more knowledgeable as a whole, and everyone would collectively benefit. As I say, not aimed at anyone.
Run a query for "SEO company" (Google US), grab the top three domains, take a look at the top exact match anchor text results for each in OSE. That's just one example where exact match appears to be a primary driver of rankings.
I still think its a factor, however those results aren't typical. Position 2 is an exact match, technically anchor text to this result is a brand anchor text (seocompany.com), the first result has 55% of the linking domains anchored with 'seo company' but hundreds of variations incorporating the term, and top SEO's in 3rd have only 7% of their linking domains anchored with SEO company, 125 linking domains, a lot less than the competitors below. It is a signal, but is it as strong as it was 2 - 3 years ago, I honestly don't think it is. You have to think, which SEO companies are doing good PR, good social/viral campaigns... as well as getting the SEO right?