The Word Bay

Google EMD update – not just about EMDs

October 18th, 2012

So, Google is really on the warpath, there can be no doubt now. The latest update, on September 28th, has seen them apparently targetting EMDs (exact-match domains). If you are not familiar with the term, it has been something of a buzzword in SEO circles for the last few years – it refers to domain names containing the exact keywords that you are trying to rank for. E.g. you are trying to rank in Google for the term dog training tips, so you register the domain dogtrainingtips.org or whatever (fat chance of THAT being free right now, but it’s just an example!), and Google always tended to rank your domain artificially highly for those terms because of it. This makes sense – the Google algorithm had difficulty telling the difference between when someone had created a genuine site about dog training tips and when they were just gaming the search engine results – and if a site is called dogtrainingtips.org then it had to take seriously the possibility that this is exactly what the site was about.

Well, it seems Google have finally gone a long way towards cracking this “problem” (for them, exploit for us) because the latest update has hit EMDs HARD. Probably the best evidence of this is in reports from people who were doing LARGE-scale EMD site-building, like the Adsense Flipper guys, who report a 69% drop in traffic across their hundreds of niche sites.

Personally, I took a hit on one site which wasn’t even exactly an EMD, but had the keywords in a different order in the domain name. The content is pretty high quality, like, actually well-written and informative, but I don’t think EMDs and quality of content tells the whole story about this update.

What is this update all about?

Well, I can say that I have another EMD (I had got away from those lately, but I had these from before) which is still ranking no. 1, no problems, though it took a hit on its inner pages, almost certainly because they are a bit thin and with too many affiliate links. The difference between site A (the hit site) and site B (the unaffected site) is as follows:

Site A:

  • classic “microniche” homepage, with a long article that was, though not “keyword-stuffed”, obviously optimised around the main “EMD” keyword and variants of it.
  • a number of key inner pages optimised around “sub-keywords”, e.g. dog training tips for students, alsatian dog training tips etc.
Site B:
  • a homepage that is more like a portal, with lots of excerpts from different articles on the site, recent posts, etc.
  • inner pages that are much more loosely related to the main keyword, or do not directly contain the main keywords at all
I am almost certain, not just based on my experience (two sites is hardly a sample), but on the trend that has been happening in the last year or so, that Google is basically still targetting keyword over-optimisation here. A site that has the same keywords in the domain name, in the homepage article, title, subheadings etc. and in the main subpages is fairly obviously out to game the rankings – possibly Google doesn’t even care about the content in that case, or perhaps you have to have some VERY good authority backlinks to get away with it.
Here is something I wrote a few months ago, before all this happened, in an ebook I am trying to write (if Google will stop changing things the whole time!) – I still absolutely agree with what I wrote, and I think it is being proven true:

I’ve talked a lot about avoiding some of the excessive on-page optimisation that goes on, and how I firmly believe that Google will be, and probably is already, looking at typical micro-niche footprints, especially obvious over-optimisation for particular keywords. These are some of the things you get taught to do in SEO courses, which I strongly recommend you do not go crazy with – people may be getting sites slapped for these very reasons, and still be completely in the dark as to why it happened:

  • your “main” keyword twice in every title tag
  • your main keyword in every heading and sub-heading (h1, h2, h3 etc.)
  • your keyword bolded or italicised in the first and last paragraphs of your article (c’mon! That’s just silly, it’s like, “Look at me, I’m a microniche site”!)
  • constantly linking back to your homepage with your main keyword as anchor text
  • your keyword in every alt attribute of every image
  • etc. etc. etc.!

Look, all those things make some sense, but how long is it before Google starts to get really smart about working out that sites are unnaturally optimised for certain keywords? In fact I am pretty sure they already are looking at this stuff. I have not seen any long-term detriment to my rankings in NOT going overboard with this kind of on-site SEO. Authority counts for much more, and that’s what we are aiming at with this method [the topic of my ebook].

 

Remember, I wrote this several months ago, in fact even before the first Penguin update, and I think I have been proved totally right. Actually, I had stopped doing most of these things, but I think I still did not do enough on some of those old sites to correct the problems. I will be adding EMDs to this list, for sure – I had my doubts about them before, but now I would definitely avoid them. I will also be thinking even harder about how to break up and “deoptimise” some of my sites, especially this affected one, and will report back with the results.

What has your experience been? Have you been slapped too? Does your site fit in with some of these “footprints” I have described.