August 21, 2023 (Updated: January 24, 2024)
Google has always cared about who is publishing content and has used this info as a core of its algorithm. With the roll-out of Google Plus in 2011, Google started taking an interest in going beyond the concept of authority for a website through links, to actually connecting authors and writers as “entities” via their Google Plus profiles.
But the concept of Google understanding entities has continued since.
In the early days of search, this concept of authority was measured mostly through website links. But when Google rolled out Google+ in 2011, authority and authorship went beyond website links to establish authors and writers as the publishing “entities” through their Google+ profiles. Over time, this concept of authority grew beyond backlinks to encompass authored content.
So, what is authority? And how does this authority and authorship play a role in your SEO strategy?
In this latest SEO round table, CopyPress CEO Dave Snyder and Director of Content Analysis Jeremy Rivera discuss what this authority means with Jon Henshaw (Vimeo and publisher of Coywolf News) and Mark Traphagen (SEOClarity) and how Google looks at it in relation to the content you publish.
Jeremy Rivera: Hello and welcome I’m Jeremy Rivera with CopyPress, here with Dave Snyder CEO of Copypress, John Henshaw with Vimeo, and his own awesome newsletter, Coywolf News. We also have Mark Traphagan of SEOClarity and one of the earliest writers that I can remember — aside from the wonderful Bill Slowsky, who’s talked about authorship and that concept deeply.
So, Mark, let’s just spitball and throw out there… can you give a quick historical note about the concept of authorship and how we think maybe Google might have been using it? And then we can build through the years and see where we think it is now.
Mark Traphagen: I’d be glad to try to be as succinct as possible. I’m glad you mentioned Bill Slawski because you know we stand on his shoulders for a lot of things, but this is certainly one of them. For anyone who doesn’t know, you should go back and look up Bill’s work. He specialized in analyzing Google patents and he was kind of our informant in that world.
He had some experience and background in patent law, and he was able to look at these patents, understand them, and tell us what they might be saying about what Google’s up to in the future. He always emphasized might — he was careful to do that because just taking out a patent on something to protect intellectual property doesn’t mean you’re going to use it. And that was a controversial [aspect] that started this whole thing, which was something called Agent Rank patent.
In the early 2000s, they came out with the Agent Rank patent, which was basically a patent to say, like, this would be a method by which a machine can understand that this is an author, this author is a particular entity or a particular person out there, and that they are connected with these pieces of content.
It was pretty much that simple, it didn’t go into what they would use that for, but it spawned this whole way of thinking, could this be a ranking factor that Google would want to use? Would they want to be able to assess who wrote what, who are those people that wrote it? Should that be a ranking factor?
And there was rampant speculation in the SEO community that it was going to be involved with ranking somehow. That, you know, being involved in Google+ would affect certain SEO in some way, and that Google would connect it. But those pieces got put together when Google started around 2011 the rel=”author” authorship project, which was putting out a piece of structured data that they were going to use to connect anything you’ve written on the web to your Google profile.
The speculation since then on my part and others has been that Google said if they’re going to do anything with authors, it’s going to have to be at a machine-learning level. That’s the only thing that will scale. I think their attention has been turned, like, “We have this way, we don’t need you to code anything anymore.”
Google has ways of knowing who you are if you’re an identifiable author, and with reasonable certainty can connect you with your content. So, my attention has turned not so much to the [specific] method that they might be using. I think they’re always playing with that. It’s the question of whether they are using that in any way — is it a thing?
Jeremy: Google added an “experience,” to E-A-T, so if you don’t have a system that can connect or start to identify and connects the knowledge graph of this entity in this particular content, then, you know, is the expertise that you wrote a word in a way that says, “Oh I really know this,” or is there really a foundational connection of authorship? This idea of “this particular person wrote this, and this shows their expertise and their experience,” is either one of those E’s or authority.
All three of those imply that Google has reached some sort of capability that it’s confident in applying those concepts not necessarily as direct ranking factors — because that’s how they doublespeak — they know having an author there doesn’t necessarily change it one rank or two. But in general, we know that the brakes can get put on due to the Helpful Content Update and things being really low quality. It starts to connect those pictures, especially when you can see a knowledge panel result.
Mark: A few years ago, it was known that to become an entity — a human entity — a person entity in Google’s knowledge graph, there were certain databases you needed to be in. And I just never did the work to get into all of those, so I didn’t have a knowledge panel. Then all of a sudden this year I got one without doing any of that work.
But that tells me that the knowledge panel was correctly connected to my content. That’s how they showed it. And it tells me it’s like they’ve gone beyond that now, they don’t need those databases anymore. They probably use them, but they don’t need them exclusively.
So I think the thing that’s interesting is they have the data, I don’t doubt that at all anymore. They have the data for anybody who’s a regular author on the Internet who actually exists. I think they have the data. The question is, what are they doing with it if anything?
Dave Snyder: So I’m going to be the dumb one on this Zoom — and I think it’s good.
The analogy I’ve always made is Google makes the best chocolate chip cookie, by a lot, right? In America, we think we have the option of Bing and whatnot, but you get outside of the U.S. and it’s Google for search. So if I make the best chocolate chip cookie and people are buying them, why the heck am I gonna change my recipe for the chocolate chip cookie, right? And with Google being a public company, every decision has to be backed by revenue.
So it’s like, what inherent monetary value is it to them to figure out the authorship? Better search? I mean that’s subjective. I think the thing that’s gonna change this now is AI [because there are obvious ways to detect AI]. But Google having results inundated with low-quality content…will have an effect on their bottom line now. I think they’re smart about not rolling out stuff that’s going to have an inherent effect on their business just because. Everybody was giving them so much heat for not rolling out the generative AI stuff as quickly as Bing.
I think it was the best thing ever because that was a total flash-in-the-pan gimmick and a low-quality feature overall. So my hypothesis right now, Mark, and what I’m interested in is schema. The fact that a huge portion of the web now has JSON-LD layers, most people have structured data, and there are pieces in that data where not only can you say who the author is right — which as a spammer I would destroy that all day if they were using just the author name — but there’s the author URL in that schema.
Jon Henshaw: The first thing I want to talk about is E-E-A-T, which — all I’m gonna say with that is that’s a human factor. That is something that has been set up that’s both for their quality raters, but it can also impact regular people and that can impact rankings.
The things that I think have the most impact? For one, I do think they use machine learning to understand a person’s writing style particularly. The more data they get from the actual human who’s written that stuff, they can be like, “Yeah I’m pretty certain this is the same person.” So, I think that’s a piece.
The other thing that was brought up, which to me is extremely important, is the idea around schema and entities. And to me, I wrap all that stuff up into the term disambiguation.
So, the general goal and way that you get anything to rank on Google is to disambiguate it enough so that Google is confident that — out of the billions of pages it could return for a particular query — it’s super confident that this is probably the most accurate and fits the intent, whatever it might be. And bringing that back to the topic of authorship, it comes down to consistent linking.
Jon: The other thing is schema and structured data. I will say I’m religious about how I do that and make sure that I also include a nice list of “same as” links associated with that author type. You know, there are certain things that I have done that if I’m good at interlinking back and forth, it creates this tight web of all these links to each other. And all of those ultimately link to that main author page, and that disambiguates for Google.
It tells them that this is all the same person and all the content where I see this JSON-LD, all these different things are related to this person. And then there’s more to it, but I think it’s the simplest version of it is consistently linking to one thing and then connecting all of those together and then writing.
Then the last thing I’ll say is about AI. To me, it’s about whether content creates engagement for the searcher and why. What I mean is, how long do they stay on that content before they bounce back if they bounce? That’s engagement.
If somebody searches for something, they click on the link from Google, and if they’re on that piece of content for a minute or more, what I have seen is that that content is likely to rank even without any backlinks to it. So take all the things I said on top of the fact that you have high engagement. That is a formula for [Google] understanding the person and for that content to perform really well.
Mark: E-E-A-T is an aspiration, it’s describing the way that [Google] wishes the web looked, basically. When I say it’s aspirational, I think the purpose of the quality raters is to ask, “How close is whatever is in the algorithms, how close is that coming to what real humans want to see and want to engage with in the ways that John described?
Dave: E-E-A-T has been the shiny thing for a long time, and when AI came around, even as much as I hated it, I was like alright at least somebody will talk about something new now. But I feel like AI’s been such a train wreck, a lot of the products have already been rolled around, and they’re so immature and terrible.
Everybody was ahead of themselves on it, and it just feels like it’s burned out fast. Right? I mean, I think we’ve seen it at CopyPress, we had a really slow first six months and we’re seeing people come around now, and I’m very sure that people were testing AI heavily to replace human writers. And everybody we’ve talked to has been like, “Yeah, that did not work.”
More from the author: