Sometimes when SEOs run tests, we find hacks and workarounds for ways to get things done. I’ve recently been running some tests about rel=author, to try to get it to show for my site and others.
I now have conclusive proof that rel=author is not algorithmic. Rather, it is manual by site AND author. And I bet that the authority of the site has something to do with whether or not it shows as well. So we have a combination of who the author is as well as the specific site. If your photo shows for one site, it will not necessarily show for another.
Allow me to demonstrate.
Everyone in SEO knows who Rand Fishkin is. Once SEOmoz implemented the rel=author tag correctly, his photo popped up in the SERPs almost immediately. Here is his mug in the SERPs on his Social Media KPIs post:
Rand is also kind enough to trust me to test some things on his personal website, Randfishkin.com. I implemented the rel=author markup on this site on October 13th. I implemented it using the “?rel=author” parameter. I then validated it with the Rich Snippets tool and saw this:
And yet even today, even when it has been validated for almost 3 weeks, this is what we see on his VC funding post:
So either rel=author is manual, or Rand’s site is not old enough/authoritative enough (it has 265 LRDs though). I’m betting it’s the first one, though. Look at the next example.
Tom Critchlow is my boss here in New York. He implemented rel=author on his personal site as soon as it was available. He then pulled some strings and pretty quickly we saw this:
Tom even implemented this in the same way that I did on Rand’s site, with a link to his Google+ profile using anchor text with a “+” contained in it. The photo showed pretty quick. Then recently I was searching online and saw this for Tom’s posts on the Distilled site:
But when I went to validate this in the Rich Snippets testing tool, here is what I saw:
So Tom’s photo shows on BOTH his personal site (which has low authority and links, with a total of 19 linking root domains and is a Tumblr site) with valid markup and the Distilled site, which has 1279 LRDs, yet no valid markup.
Author authority. Tom’s on a whitelist that not even Rand is on yet. (Hellooo bragging rights).
Matt Cutts is the head of Google’s spam team. Obviously he is going to get special treatment, but come on Matt, what’s happening with you is ridiculous.
Here is Matt’s mug in the SERPs:
When I checked the Rich Snippets tool, it is telling me that his markup is correct:
But when we check his “author page”, which is the homepage of his site (mattcutts.com), we find no link with a +. In fact, what we find is a link to his old Buzz account using rel=me! Check out this code:
But guess what?
Matt’s Buzz link redirects to his Google+ profile!
Apparently Google is recognizing this as valid markup. Either that, or it is a case like Tom’s (which is my guess) where Matt is on a whitelist that allows his profile photo to show at any time.
I wrote about trying to get rel=author to show for my site about a month ago. Soon after that, I saw this in the search results:
But then guess what? It disappeared the next day and has not come back. The same thing happened to my friend Brian Whalley over at Hubspot. His image showed for that day and then disappeared and has not yet returned.
So what have we learned?
First, rel=author is a manual application Just because you have validated does not mean it will show (it is a pilot program, after all).
Second, you do not necessarily need to have the markup implemented for your photo to show. This leads me to believe that one of two things is happening. Either the authority of the site will be enough, or there are multiple levels of whitelist for photos to show.
Third, Google is not true to their word. Not a shocker to anyone, but they don’t play fair. Why are you going to reward some people who have not followed your guidelines, but not let those who have done it correctly, who are trying to play by the rules, have some satisfaction of doing right?
Like Ian Lurie said, I’m going blackhat. (Kidding).