Latest / Semantic SEO Podcast / The Topical Authority Formula Explained: Coverage, Historical Data and Cost of Information Retrieval
Transcript
- James Dooley: Topical authority equals topical coverage multiplied by historical data divided by cost of information retrieval. Today I'm joined with Pavel Klimakov, and we're going to be going through this formula of what topical authority is. So, Pavel, looking at the formula over there, my first question to you is what is meant by topical coverage? Pavel Klimakov: Topical coverage is a kind of, let's say, measurement or a metric for how many topics the website touches on or ideally covers in detail. In another, maybe more technical way, you can think about it as the amount of pages, separate pages, that the website has that are touching on a kind of topic or a subtopic, in an ideal scenario where all of these pages have some kind of connection or relation to themselves. So it's not just a kind of everything, like every single possible topic randomly disconnected from each other. It's a tightly connected circle of, uh, let's say, topics that are very strongly related. James Dooley: Yeah. For anyone who's watching this, I strongly recommend checking out a link in the description about the site radius score. I go into detail about how wide you can go on a website. Pavel has got an amazing way and a concept about playing a game and how connected they are, because the amount of people that seem to go too wide and the topics aren't connected, and then you're leaving orphan pages, and it actually can have a detrimental effect sometimes. So many people, Pavel, obviously you know this, we've spoke about it off-air, and so many people with topical coverage just think the more topics, the better. And they talk about everything and think, "Okay, I'm going to keep doing it," and they go, they expand, and the topic dilution of it can actually harm the website. But the next part on that, so you've got topical coverage and it's all interconnected all nicely together. What is meant by multiplied by historical data? Pavel Klimakov: Historical data, in a simple way, is just the, let's say, kind of historical log that Google stores in terms of how people behave on this website. How often do they visit? How many pages they visit per session? Do they return back to the site? Do they store it in their bookmarks, favourites? Basically, how do they access the site? And what's this overall, let's say, retention of the people inside the website? It can include additional things like copying certain elements from the website. Uh, checking the images or hovering over different buttons, elements and things like that. In a simple way, Google is just paying attention. What's the human response to everything that this website is presenting on there? James Dooley: With the historical data, is this part of Navboost or is Navboost something completely separate? It is. And then, does traffic, does CTR come into play, or even, like, social shares or backlinks? Does that form part of historical data as well? Pavel Klimakov: What I believe, it's mostly connected to the human signal in itself. Um, however, there is a kind of, let's say, grey area or a bit what I see, a kind of, um, not so clear components. Because, for example, impressions, uh, can be signalling, uh, historical data or a user behaviour. But at the same time, um, impressions can be triggered to, let's say, slightly different, um, behaviours. Even it can be imitated through things like CTR and other things. And the point here is that technically, even though somebody searched for a topic, your website showed up somewhere, it was not an available option for them to click on, you still triggered an impression. That starts to accumulate a kind of, uh, like a, a small signal for the accumulation of the historical data. James Dooley: Yeah. And then with regards to the initial formula, the first time I heard the formula was from Koray, right? So initially I heard it as being topical authority equals topical coverage multiplied by historical data. And then the first time I heard the divided by cost of information retrieval was actually by Lewis. Um, but from there then, what is meant by the divide? Now, the whole concept, the Koray concept... I'm going to speak nice of him. Koray is a legend. I love him to pieces. He's like a brother. But I don't normally say that to his face because he'll get too big-headed. But anyway, Koray's methodology is all about the cost of information retrieval. How can you make it cheaper for Google to be able to rank your topics? So they divided by cost of information retrieval. Can you explain that in more detail? Was it Lewis that said to add that on, or was it Koray, or where did that come from, the divided by? Because that wasn't initially in the Koray course. Pavel Klimakov: I do not remember precisely where it came from, but I can definitely say I heard Koray mention this as well. Um, the whole concept of cost of retrieval, the way I see it is, to be honest, this is the whole point of SEO. For Google to prefer your website, there needs to be a strong reason why. And the strongest reason in today's world is the economic benefit. If it's cheaper to work with you, we will work with you. If you are sourcing materials from somewhere and they offer you the same or better quality for a cheaper price, it's a no-brainer that you'll choose that partner, right? This is the exact same case. If you're able to provide better quality of information at the lower cost, it's a no-brainer to prefer this website over the other. James Dooley: Can you give one or two examples of where people make it more expensive for their website to actually be ranking, like with regards to cost of information retrieval? Pavel Klimakov: Sure. So, for example, the simplest one is having some kind of cheap or broken hosting that takes a ton of time for Google to access. They have the crawlers have to be stuck for a few seconds and wait until everything is loading. So, as these seconds pass by, the crawlers could be doing something else, but they're wasting their time over here. So, that's an expense. Another point, let's say we have a, um... Let's say an article specifically dedicated to, um, should I install a kind of, like, snow or ice-blocking elements on the roof? Or, like, what's the benefit or what's the reasons to do that? And imagine that the article starts to talk about, like, "Hey, I am this roofer and I've been doing this for many years and I live there and work there." That was not the main promise of the article. We were promised to discuss this mechanism and whether we need it or not. So please give us the actual, let's say, reasons, the pros and cons, the costs. Give us proof, like, how much snow it blocks in, let's say, in a specific area in the UK, during which months, where is the highest risk of, you know, the snow could be melting and falling down. It can cause some injuries. These other things. Maybe in other specific places or countries there's a regulation or law that you're obligated to have them installed. So things like that. And, uh, since you mentioned Lewis, um, let's say, um, we can, uh, think of something like including just simple schema elements that takes your components of the information you provided and, let's say, zips them or compresses into a very, very easily structured type of format, which is extremely easy for bots and crawlers to process. It's like a near-instantaneous understanding of everything that is presented without having to do any kind of extra computation to try to interpret what has been said on that specific page. James Dooley: Sounds great. So anyone who's watching this with regards to the topical authority framework, topical authority equals topical coverage multiplied by historical data divided by cost of information retrieval. Let me know whether you've studied the topical authority framework in the comment section. Pavel, it's been an absolute pleasure. Pavel Klimakov: Likewise, James. Thank you.