I came across a post on Hacker News yesterday titled “How Google Authorship decreased our traffic by 90%” which was wrong on more levels than I knew how to count. I am not saying that it was intentionally wrong but the conclusions that the author came to (which you can probably guess from the title of the blog post) weren’t even close to correct.
Furthermore, the discussion in both the comments section of the site and again on Hacker News were kind of scary about again how misinformed a lot of people were and how very few people realised that the evidence didn’t stack up at all.
So as someone who has done enterprise level SEO for large multinational companies for a number of years I thought I might be able to clear up a few points to help educate people about what actually happened, but also what they can and should be doing if they find themselves in a similar situation in the future.
First just let me start off by breaking down some of the main points from the original article and from there we can see how the analysis starts to fall apart rather quickly.
1. Google Webmaster Tools Notification
This is where it all started for the guys over at Jitbit, the received a notification stating that search result clicks to a certain URL had dropped significantly. So from here he decides to rightly go and check his rankings, however this is where things start to go slightly wrong.
I don’t know if the screenshot is there just for demonstration purposes or what, but as a future tip this is not the way you want to check your rankings. I.e. while you are signed into Google and you have made zero attempt to remove any kind of personalisation. Personally, if you don’t already I’d recommend investing in a ranking tool such as AuthorityLabs or something similar (If your business relies heavily on SEO traffic this is a critical investment in my books).
At the very least sign out and append the pws=0 query string to the end of the URL to help remove the personalisation to try to determine your true ranking positions like so: https://www.google.com/search?q=macro+recorder&pws=0
However, once they saw they were number #1 still for this particular term they seemed to immediately jump to the conclusion that since traffic is down and so were the clicks according to GWT that the authorship markup must therefore be to blame. Instead here is what I’d have advised you to do instead:
- Check your rankings for a bunch of different keywords. The second you start to focus on just one like this you start to operate in a silo and you are absolutely going to start making decisions on bad / questionable data. But in real life that page ranks for a bunch of different keywords and just because you noticed that rankings for one of them didn’t change, doesn’t tell you anything what so ever.
- Check impression data in GWT for that URL. In the screenshot from the post the author is showing impression and CTR data for a bunch of different URLs. Some are going up and some are going down.Note: ALL of them have the exact same authorship markup on them which should be the first signal that Authorship wasn’t the problem here.
2. “Most people were clicking the number 2 result”
I have absolutely no idea how the author of the article was able to say that he somehow knew which results people were clicking but I’d love if he could share it with me, we can make a lot of money together with that information. However, the remainder of that article goes into a great depth of detail where the author is trying to fit bits and pieces of evidence to fit his initial theory in a great example of confirmation bias.
He then goes on to make a bunch of crazy and ridiculous claims based on all of this such as:
- Our rankings haven’t changed.
Actually, they had, massively but just not for that single keyword he mentioned.
- If you have authorship markup and you are the #1 result or the only result on the page with authorship people will skip you and click on something else
Almost no one at all in the SEO industry is going to agree with this statement, almost all of us have example after example after example of where the exact opposite has happened. I’m sorry but you’re wrong on this one.
- People will confuse you with paid ads if you have authorship markup on your pages.
Again, there is zero evidence at all to support this.
What actually happened
Immediately upon reading this I knew that something wasn’t quite right. Given that there have been some MAJOR algorithm updates in the past couple of weeks I knew that more likely than not, something else was to blame for this drop in traffic. This was later confirmed when Google’s Head of Webspam: Matt Cutts showed up to weigh in on the thread and said directly that the reason this site saw such a large drop in traffic was because of Penguin and the questionable link building tactics that they had been advised to undertake.
Personally I feel sorry for them in that they are just yet another data point of someone who was blatantly ripped off by a shady SEO provider and now have to suffer as a result. This is precisely the kind of reason as to why I am making an exit out of the industry myself as I can’t stand to be associated with this kind of stuff.
I don’t wish to go into detail and start publicly tearing apart their SEO strategy and link building because I don’t think they deserve it at all.
As a personal note to the guys at Jitbit, you have some work ahead of you here. If you would like a bit of a hand I am happy to make myself available to you for a couple of hours to help walk you through exactly what you need to do from here for free.
- Google Authorship DID NOT cause their drop in traffic.
- Don’t get caught up focusing on a single keyword and use it as the basis for major decisions.
- Authorship probably doesn’t really belong on product pages and while it’s not what happened in this instance, it could potentially not be amazing for CTR on some terms.
- When trying to figure out something like this at the future make sure to look at both the impressions and the CTR at a keyword level and at a URL level. It will quickly help you answer a lot of your questions.
- If you are hiring an external SEO agency, make sure they are providing you with complete transparency into everything they are doing.
Note: There is a discussion thread of this post including a response from Matt Cutts himself at: https://news.ycombinator.com/item?id=5801845