Google “Farmer” Update – Before and After
Farmer still hasn’t hit the UK yet, but if you’re interested to see it in action you can actually do a live before ‘n’ after comparison. Go to google.com from the UK, do a search and note what you see. Then add &gl=us to the end of the URL and you will see the updated SERPs. I’ve dug out a couple of entirely random examples below that hopefully illustrate what you might find. I’ve also been completely dumb and not done a like-for-like comparison at all, as some commentators have pointed out. I’ve never got a handle on Thursdays.
The first thing to note is that Farmer has been billed as a conscious attempt to wheedle out what Google considers to be low quality content farm sites. This is the area where you find affiliates and site-owners chasing the long tail with very specific articles or pages – either directly on their own site, or on a third party site like Squidoo, ezinearticles and so on.
A typical search term for this kind of thing might be to target a phrase like “where can I buy new jeans?” I’ve highlighted the results that I consider to be ‘articles for SEO’ in yellow.
“Where can I buy new jeans” – pre-Farmer
“Where can I buy new jeans” – post-Farmer
The post-Farmer result reflects something that Google has been following for some time: the elevation of brands and Google’s own properties – in this case Google Shopping.
The big oddity is at the #1 slot – which goes to a record by some band or other sold on Amazon, which is clearly not relevant to the intent of the search. The remainder of the results see almost all of the ‘article’ type content expunged from the rankings in favour of shopping results – say goodbye to 43things.com, Esquire, About.com and so on. The notable exceptions are Yahoo! Answers (promoted) and Ehow.com (down a few places but still there), which sail through relatively unharmed.
What is interesting is that none of the sites in the new SERPs even have the word ‘buy’ in their title. Google seems to have decided on the intent of the search, rather than just looking at on-page SEO. “This person is looking to buy new jeans: ergo let’s not serve them some Noddy’s-Guide-To-Finding-A-Shop”. If that trend is seen across the board, it has quite a few ramifications for trad long-tail SEO.
“Is he cheating on me?” – pre-Farmer
“Is he cheating on me” – post-Farmer
The big losers here are clearly two very specific sites – ezinearticles and Hubpages.com, which suggests that there’s something in early reports that some sites are being specifically hit very hard, either by accident or design. By contrast, 2knowmyself.com – which appears to be stacked to the gills with long tail articles – isn’t touched at all, and neither is iVillage.
Perhaps interestingly, these sites are closed content sites: you can’t ‘submit’ your ‘articles’ to them. It’s all coming from within.
Speculative conclusions from a very narrow sample base
Farmer has two components: firstly some apparently site-specific components which have worked to dampen the power of free-for-all content farms. It might seem harsh lumping in Ezine and Hubpages into that bracket, but they’ve always been go-to destinations for anyone with 400 words and a link quotient to reach. For every individual diligently working on their original content and pushing out through these channels, there’s been 6 guys with bits of spun content and it could be that Google has just decided the signal-to-noise ratio has just gotten too poor.
Secondly, Farmer is about (repeat after me) “improving quality”. Google always say this, and to be fair, they’re generally right but here it’s less clearcut as to whether the changes actually improve quality. There are great Hubpages, after all, and if you don’t know where to buy jeans or whether your boyfriend’s cheating on you, there’s no reason why an article site can’t give you just as accurate information as the Wall Street Journal.
But, on average, Google have decided to go with their ‘brand’ thing. iVillage is a brand so regardless whether their thin content is worse than the thin content on articles-for-my-personal-business.info is neither here nor there.
What difference does it make?
Firstly, if you’ve been using article sites to pad out your link profile, it’s pretty likely that any residual power from those articles will be lost. Building links out from article sites has always been something of a wash from an SEO perspective: you get the link fairly easily, and you can gain some long-tail relevancy from it, but you wouldn’t expect to rank a site for a head keyword on the back of it.
But if you’ve been hiring Indians to churn them out by the score to support your rankings, that’s bad news (especially if you’ve already been burnt by the Feb 9th update, which took out swathes of scrapers and spun-content hosts).
If you’re an affiliate or supporting yourself through AdSense… well. Having 6000 hubpages might have been a living a fortnight ago, but now you’re going to find other sites eating your lunch. As ever, the game will just move on: if Ezine’s burnt, there’s alternatives out there, and ways to monetise them. The smart will survive.
But poor Joe Schmo just found that his “Make a million from Squidoo!” ebook isn’t worth the pixels it’s written in.
Concentration of power is a recurring theme of recent Google updates: brands and big domains benefit every time. So, in a way, this changes nothing for big-term SEO. It’s still about building a presence. What’s less clear this time around is how to approach the long tail.
Building up a big content-heavy resource of FAQs, technical support and insider knowledge has always been a winning strategy to pull in that sweet long-tail stuff, but as in the jeans example above, you can see how even a site like Esquire can see their shit get rocked. Suddenly, those 400 variants on ‘winning bingo strategies for dogs’ you’ve got coming out of your blog just got a little less valuable.
It’s an extension of the phenomena we’ve been seeing sporadically for a good couple of years now where you’d see BMW ranking for a part that they don’t even sell: the trust of the BMW brand in Google’s eyes means that it is most likely to be able to give you the relevant answer, even if there is nothing specific that it can find that matches the query.
Taking all these factors into account, the answer is wearyingly familiar. Build trust in your site. Engage with your customers directly – either on your site through good content… by going out onto the web and finding the complaints and questions about your products and answering them… by cutting deals with partners and sponsors where a little Google-friendly exposure wouldn’t go amiss… by developing a brand.