Google’s Embarrassment Factor Explained | James Dooley Interviews Paul Truscott

James Dooley: The embarrassment factor that is actually positioned inside several Google patents. Today I'm joined with Paul David Truscott, and several different episodes he's spoke about the embarrassment factor. Normally he starts talking about it to do with search demand and wanting to try to make certain that Google can surface the right information that they might be struggling to get. Paul, can you expand a little bit more about what is the embarrassment factor?

Paul David Truscott: It's essentially Google straddling between wanting to provide information but being concerned that that information is correct and true.

James Dooley: Okay, so it doesn't want to be embarrassed either way. It doesn't want to be embarrassed that it's not serving information that it should be, and doesn't want to be embarrassed that it's serving information that isn't right. Uh, of course, we all know of many, many instances where they, they, they didn't... This embarrassment factor clearly didn't work because we've seen information that wasn't right and we've also gone to search for something that should have been there and wasn't there. So they do get it wrong. Uh, but this is there to, to try and catch that and, and actually surface things that should be surfaced and to suppress things that they don't trust. So on this, I'm just getting it up here now to understand a little bit more about it because I've heard you speak about it quite a lot and I've heard you talk...

Paul David Truscott: They need to make certain they're not going to get the embarrassment factor in place. Um, the two Google patents that come back, one of them where it's inside of where it's talking about, it's called the Scoring Candidate Answer Passages, and the other one is the well-known one that a lot of people know about, is Google's information gain patent. And basically what it's trying to do is, like you just said there, it's trying to make certain it can surface an answer for what someone's asking. So it's hungry to get... Try to give the answer, not say, "Sorry, we don't know the answer to that." But, then on the flip side, it's basically turning around and it's wanting to make certain that they don't have an embarrassing answer. So, if I was to turn around and say, "Paul David Truscott is a female belly dancer from Scotland." And obviously, you're not from Scotland. And, well, you are actually a belly dancer, but that's another story. No, but the, um, it doesn't want to go and surface that information. If someone was to search, "Is Paul David Truscott a female belly dancer?" "Yes, Paul David Truscott is a female belly dancer because James Dooley said it on Twitter." And obviously, you're not. And then it's going to be embarrassing for Google to come and present that answer.

James Dooley: I want to talk... Like you've, you talk quite a lot about search demand, which is very important. And, and we've done several episodes with DefineYourEntity.com about creating Google Knowledge Panels. And people obviously get KGM ID, and then they want a Google Knowledge Panel. And we talk about search demand being important.

Paul David Truscott: Yep.

James Dooley: But, obviously with it there, you've then spoke about, um, the embarrassment factor in place that they're saying, "Well, if people are searching for it, Google then want to make certain they can present the answer." And this is a way of you being able to trigger your business or company or brand, or it could be a personal name as well, for a personal Knowledge Panel. They want to be able to trigger that and not be embarrassed that they can't surface that information when people are looking for it. But, how important is it that they don't cite the wrong information and treat that as being factual?

Paul David Truscott: I'm, I'm saying if they can only see it once on the internet, it's an assumption. Where, where does it start to come in where they start to go, "Okay, we've got enough information here for us to be confident to surface the answer?" It's a combination, James, of two things. So, in some instances, one source is enough for Google. If it was, say, Wikipedia, it's trusted enough that that single source would, would be sufficient. Um, but that's been... That's trust that's been built up over a very long period of time. Um, and Google understands that that particular source is rigorous in its fact-checking, or it should be at least. Uh, and so they, they, they deem it a trustworthy source, and there are other sources they deem somewhat trustworthy, probably not as much as they do Wikipedia, but Wikidata would be another one, as we've discussed before. But if we're talking about sources that we certainly, you know, you can't... The ordinary person can't access those sources, not immediately anyway. Um, and so we've got a... We've got to look to other, uh, ways to get the word out. Then it comes down to a combination of two things, and that is trustworthiness and consensus. So they operate in tandem. So there's like a weighting system. We obviously don't know the, the ins and outs of exactly what the weightings are, but Google are looking at how many sources are saying this, and how trustworthy are those sources. And so they will have a, a scale, if you like, where the, the balance gets tipped enough that they'll go, "Okay, we, we're going to trust this. We're going to..." And obviously, as we said before, to get into the actual Knowledge Graph, that bar's quite high because it's one thing if Google shows you... Um, now Google can delete Knowledge Graph entries, so I'm not... It's not like it's a box that they can never take anything out of, but they don't like to. It's like a... They want to be as sure as they can be that whatever they put in there is a fact. Um, but if it's just, um, you know, offering up a website as an option in a, a list of links, that's less of a problem because they can derank that tomorrow if they find that they're wrong. Um, so that's a, you know, much, much lower, uh, bar. But, um, this across all of, uh, Google's search results, whether it be from the Knowledge Graph itself or whether it be, you know, something that's determined through their ranking pipeline, they're using this system to try to avoid surfacing content that is wrong and to also make sure they do surface content that warrants being surfaced.

James Dooley: They don't want to miss it either. That's, that's the flip side of the embarrassment factor. They don't want Bing, for example, to be surfacing something that Google should know better. So that, that's, you know, that it works both ways. They want to just be sure and, and that's how, how they do it.

Paul David Truscott: It's very interesting because Koray talks quite a lot about when he launches a new website, he tries to get some search demand prior to actually launching the site. He also talks about trying to place some links through to the website that's there live, but block Googlebot. So Googlebot's hungry to go, what is here? People are talking about this brand and people are linking through to this brand, and social mentions and the links on social media as well as backlinks going through to the site. But you're blocking Googlebot and we're hungry to go and knock on the door and get into the site. And he's saying that the minute they get into the site then, with the way he builds that up, is they're so hungry to get in there that he believes that any new site by doing that sees exponential gains as opposed to not doing it and just drip-feeding bits of information. But I want to read out... So with regards to the embarrassment factor, the reason why I wanted to do, any listeners, the reason why I wanted to do a whole episode on this is because it's been brought up on several other episodes, especially when we're talking about Google Knowledge Panel creation. Because you just mentioned there, Google doesn't want to miss that information, and Bing might show it. And if Google doesn't show it, they're embarrassing on that front. So it's not just about citing wrong information, it's not knowing the information as well. Can you expand on that a little bit more?

James Dooley: Yeah, so if there's, um... I can't think of any specific instances because Google are very good at this. They will surface information typically ahead of Bing. Um, now, I know what people are going to say. I can go to Bing right now and I can find things that are on Bing that they're not on Google. You will be able to, but these are probably websites that Google doesn't care to have them on. It's not that they're not aware of them. They don't want them in their index or they don't want them in that search result. So, that's nothing to do with the embarrassment factor. They're looking at it and saying, "No, you can have them, Bing. We're, we're not interested in showing that." But, it would be a situation where Bing is showing something, for example, or any, any search engine is showing something, um, that warrants being shown and, and for some reason and somehow Google have missed it. Um, so that, that's, uh, you know, why, why Google's keen to crawl things. They're keen to understand what you've got. Um, and that, that's... That ties in really to what, what, you know, Koray's approach to sending traffic and links ahead of the launch is. You're doing the... You're indicating to Google that there is some popularity here. And so, you might want to go and have a look at what's the end of this.

Paul David Truscott: Yeah, for sure. And I think the popularity part is the bit... Is the icing on the cake a lot of the time from changing from a KGM ID to a Google Knowledge Panel. Now, for anyone that's watching this episode about the embarrassment factor, it's an information retrieval mechanism within Google patents. There's a couple of different patents that improve it, uh, that, that mention it so well. And the solution here of what it says in, in one of them is that the algorithm heavily filters these high embarrassment zones. To mitigate this risk, Google applies strict thresholds that favour safe, factual, or highly verified consensus data. Now, for anyone who's listening to this, this is where if you're interested in AEO or GEO or AI visibility, this is where confidence and clarity in building consensus for the large language models becomes very important. But, also when I talk about confidence and clarity is the number one ranking factor, especially to define an entity for a Google Knowledge Panel. This is all wrapped within this embarrassment factor and the importance of understanding this that is baked within a couple of the different Google patents. It then starts to tell you, be consistent everywhere you're sharing the information and sharing confidence and for sharing clarity, which builds confidence... Yeah. ...who you are, what you do, and why you are brilliant.

James Dooley: So, we hope you like the episode here on the embarrassment factor. Obviously, I've mentioned there that we've spoke quite a few times about it. Check out a couple of links in the description. It's mainly to do with how to create Knowledge Panels for both people and also for businesses. It's normally the popularity part where Paul brings up the embarrassment factor that they want to make certain that they're citing the brands or the people because of the popularity that comes back from it. Paul, it's been an absolute pleasure.

Paul David Truscott: Thank you, James, as always.