Thumbnail: Big Tech Algorithms are Children

Like any business, Google, Bing, Facebook, Twitter, Apple, Amazon et al. have one driving purpose: to best serve the needs of their clients and users. To do that, they are all building machines that understand the world and can use that understanding to best serve their users and clients.

Essentially, they are all building a knowledge-thirsty child that is learning everything it can about who we are, what we do, and who our audience is and training it to use that information to serve their users, clients, and business interests.

What strategy do we need to employ? We are the responsible adult in the room and we need to educate each BigTech child about who we are, what we do/offer and who our audience is. It is our responsibility to educate these Child-machines about our small corner of the Internet. We need to start now before it is too late. 

In this video, The Brand SERP Guy (Jason Barnard) expands on the concept of Google as a Child and explains how we can educate it and the other BigTech child machines.

What are Brand SERPs and Knowledge Panels?

[00:00:00] The Brand SERP Guy (Jason Barnard): Hi, I’m Jason Barnard. I’m The Brand SERP Guy. And today, I’m going to be talking about Big Tech Algorithms. They’re children that want to understand and you need to learn to educate them. So, like any business, Google, Bing, Facebook, Twitter, Apple, Amazon, and all of the other Big Tech companies have one purpose and that’s to best serve the needs of their clients and users. To do that, they are all building machines that understand the world and that can use that understanding to best serve their users and their clients.

[00:00:35] The Brand SERP Guy (Jason Barnard): Now, a little bit off topic, but you’ll see it comes back on topic later on. I’m going to start with the Brand SERP which is what your audience sees when they Google your personal or brand name. Now, this is my specialty. I’ve been working on this since 2013, and it’s phenomenally interesting. It might not seem very important or incredibly interesting right now, but I’m pretty sure by the end of this talk, you’ll find that Brand SERPs are possibly the most interesting thing you’ve never thought about before.

[00:01:06] The Brand SERP Guy (Jason Barnard): Now, here’s my Brand SERP. I’m The Brand SERP Guy. I work on this, and I’ve been working on it for years. And this is what a great Brand SERP looks like. Remember, this is what your audience sees when they google your brand name or your personal name in this case. I’m also Mr. Knowledge Panel. John Mueller from Google has called me Mr. Knowledge Panel in the past.

[00:01:30] The Brand SERP Guy (Jason Barnard): Knowledge Panels are the information box on the right-hand side on desktop when you search for a brand, a person, a film, a podcast, or any other thing that Google can identify by name. And that contains a summary of the information that Google has understood that it feels is important, helpful, valuable, useful to your audience when they search your brand name or your personal name. And through my Brand SERP and the Knowledge Panel, Google tells my audience who I am, what I do, and who that audience is, who I’m addressing myself to and how they might want to engage with me.

A Background on Jason Barnard (The Brand SERP Guy) and What You Can See On His Own Personal Brand SERP 

[00:02:11] The Brand SERP Guy (Jason Barnard): So, I can tell my story through the Brand SERP. I’m currently living in Paris as you can see from my LinkedIn Profile. I was a voice actor, a cartoon blue dog in the noughties, as you can see from that video, from the songs in that Knowledge Panel, and from the IMDb listing that ranks up there on my Brand SERP. I was also a punk folk musician back in the 90’s. As you can see from my home page, I’ve mentioned The Barking Dogs, that was the punk folk musician. And down there, once again, that video shows my life story, including that episode as a punk folk musician.

The Brand SERP Guy (Jason Barnard)  is an Author, Podcaster, and CEO

[00:02:48] The Brand SERP Guy (Jason Barnard): I also have a groovy podcast, comes out weekly. It’s intelligent, interesting, and fun at least I hope so or I think it is. And that comes on the right-hand side with the See Results About which is additional entities related to me that Google thinks might be interesting to my audience. And also the With Jason Barnard Podcast dot com. I also have a groovy podcast that’s intelligent, interesting, and fun. I hope. And you can see that with the See Results About on the right-hand side which is an additional entity that Google thinks my audience might be interested in. And the With Podcast site which appears on the left-hand side to allow my audience to interact with me through my groovy podcast.

[00:03:44] The Brand SERP Guy (Jason Barnard): And also, since the beginning of 2022, I am an author, and that appears in my Twitter Boxes. I announced that with the book at 61.99 euros. I don’t think anybody would buy it at that price. The actual book costs about $16. And I’m the CEO and Founder of Kalicube that offers SaaS and a SaaS platform and courses for people interested in optimising their Brand SERP and managing their Knowledge Panel.

The Brand SERP Guy (Jason Barnard) on Optimising His Own Personal Brand SERP and The Brand SERP of Kalicube (His Company)

[00:04:10] The Brand SERP Guy (Jason Barnard): As you can see, it’s showing my business website and my page. In fact, on that business website up there, my Brand SERP so people googling my name could decide to interact with me commercially through business, through my website, my company website. 

[00:04:27] The Brand SERP Guy (Jason Barnard): And Google even understands that I am The Brand SERP Guy. The People Also Ask is Google’s suggestions of other questions people might want to ask around what they’re searching. And in this case, it’s “who is The Brand SERP Guy” and Google has understood that The Brand SERP Guy is Jason Barnard.

[00:04:44] The Brand SERP Guy (Jason Barnard): And then, my company (Kalicube). We’ve worked on this as well a lot less time, perhaps a year and a half, but it still looks pretty good. We have on that Brand SERP, not only the company slogan, but we have our online courses. We have our visual identity. We have our social voice. And Google mentions or shows our SaaS platform including some great reviews there. Google shows Kalicube’s brand message as we intended it to.

How Do You Ensure Google Presents Your Brand Message Correctly to Your Audience

[00:05:14] The Brand SERP Guy (Jason Barnard): So one question you might be asking is how do we do that? You might want to look up your own brand name or your own personal name or even a product name and see if it shows your brand message in the way that you intended to. Is Google showing your audience the brand message that you want it to show? If not, then you can indeed optimise, manage, and help Google to understand.

[00:05:39] The Brand SERP Guy (Jason Barnard): How can you do that? You educate Google. We’ve educated Google who we are, what we do, and who our audience is. And it understands. And we can see that it understands because it’s reflecting our brand message, our company slogan, our online courses, our visual identity, our social voice, and our SaaS platform through our Brand SERP.

[00:06:04] The Brand SERP Guy (Jason Barnard): So when our audience searches our brand name, not only can they go to our site which is ranking number one, but they can also see our entire Brand message through that Brand SERP and decide how they want to engage with us and also understand what it is we have to offer them.

Google Wants to Get its Users to the Solution to Their Problem as Efficiently as Possible 

[00:06:25] The Brand SERP Guy (Jason Barnard): Now, why is that so important? Google wants to get its users to the solution to their problem as efficiently as it possibly can. When somebody searches on Google, they’re expressing a problem to which they’re looking for a solution or a question to which they’re looking for an answer. And Google wants to get them to that solution or to that answer as efficiently as it possibly can. I’ve interviewed a couple of people from Bing, and this is exactly the term they use. Efficiency to get their user to the solution. Google is exactly the same as Bing in that respect, but that requires Google to understand its users’ needs.

The Solution Big Tech Companies Use in Optimising Their Platforms

[00:07:05] The Brand SERP Guy (Jason Barnard): Number one, what they’re searching for when they type something into the search box, but also what the available solutions are. And that’s where educating Google comes in. Now, I’m a Google-geek. I’ve been optimising websites for Google since 1998 which is the year it was incorporated. That’s a long time, almost a quarter of a century now in 2022. But I realise now that all Big Tech platforms need to understand their users’ needs and the available solutions to them because like any business, Google, Bing, Facebook, Twitter, Apple, Amazon, and the other Big Tech companies have one purpose and that’s to best serve the needs of their clients and users. 

[00:07:47] The Brand SERP Guy (Jason Barnard): And the solution they have is Knowledge Graphs. Here, we have Google’s representation of a Knowledge Graph using Da Vinci with his relationship to Mona Lisa. He painted her. To Italy, born there, lived there, and then Michelangelo as well. Those are entities, Italy, Da Vinci, Mona Lisa, Michelangelo, and the relationships between them and also the attributes such as his date of birth and his date of death.

[00:08:15] The Brand SERP Guy (Jason Barnard): This is Google’s understanding of the world. In fact, all of these Big Tech companies have Knowledge Graphs that attempt to understand the world we live in so they can best serve their user’s needs using their understanding of the world.

Understanding What is a Knowledge Graph by Comparing it to Wikipedia

[00:08:33] The Brand SERP Guy (Jason Barnard): So what is a Knowledge Graph? And I’ll focus on Google here because it’s really my speciality. If we look at Wikipedia, that’s an encyclopedia for humans where we look up things that we’re interested in finding more about. It’s a fact filled encyclopedia for humans. The Knowledge Graph is a fact filled encyclopedia for machines.

[00:08:58] The Brand SERP Guy (Jason Barnard): Now, initially Google and most other machines trained their Knowledge Graphs using Wikipedia. So, a great deal of information from Wikipedia is in their Knowledge Graph. But they’ve gone much, much, much, much further than that.

[00:09:13] The Brand SERP Guy (Jason Barnard): In Wikipedia, there are 55 million articles. That’s a lot. In Google, there are 500 billion facts, about 5 billion entities, and that was in May 2020. We can imagine that now there must be well past a thousand billion facts and probably well over 10 billion entities. That is massive, and it’s growing exponentially. It’s growing incredibly fast. Google is learning to learn, and it’s filling up this Knowledge Graph with facts that it is learning from around the web.

[00:09:52] The Brand SERP Guy (Jason Barnard): Now importantly, Wikipedia has an idea of notability. In order to have an article in Wikipedia, the entity needs to be notable. Somebody would need to search for that entity spontaneously because it’s for humans. We don’t want to fill up Wikipedia with lots of useless information about entities, people, things, companies, films that nobody is interested in.

How Does Google Build its Knowledge Graph?

[00:10:24] The Brand SERP Guy (Jason Barnard): Google doesn’t have that concept of notability. It’s simply looking to understand the entire world. A quasi human understanding of the world is what Google is trying to build. And with, let’s say, 10 billion entities and a thousand billion facts. Google is moving towards that, but it’s still at its early stages.

[00:10:50] The Brand SERP Guy (Jason Barnard): Now, how does it build that Knowledge Graph? It uses machine learning. And one thing we need to understand about machine learning is that it improves exponentially. Every year that goes by, it gets better and better and better, faster and faster and faster. Its capacity to fill this Knowledge Graph with facts is accelerating.

A Brief Explanation of How Machine Learning Works 

[00:11:12] The Brand SERP Guy (Jason Barnard): And underneath, we got a little diagram that explains a little bit how machine learning works. You have over there on the left-hand side an engineer who creates a mathematical formula that applies to data that gives results that the engineer is looking to achieve, in this case, understanding the world. It then feeds into the Google algorithms that then try to emulate what the engineer managed to do with his or her data. I achieve the same goal finding, understanding facts. It tries to do this under its own steam as it were with its own methodology. So, it adapts the mathematical formula that the person has given and tries to figure out new ways to understand these facts.

[00:12:02] The Brand SERP Guy (Jason Barnard): Now, that seems dangerous because the machine might just go wild and we might lose control of it. So, what they then do is push out the results that the machine has found or the facts it’s found, and they get human beings to check whether those facts are true or not. They then feed those facts back into the machine as corrective data saying, this was a good fact, this wasn’t a fact, this was a mistake, you got it wrong. And the engineers can tweak the algorithm to help the machine move forwards faster and better and achieve its goal of finding those facts more efficiently and more accurately which is possibly the most important thing here is getting the facts right.

Google is Building a Quasi Human Understanding of the World Using Manual Curation, Crowdsourcing, and Data From The World Wide Web 

[00:12:44] The Brand SERP Guy (Jason Barnard): So when building this quasi human understanding of the world, Google started off by using manual curation. And here I’ve got the logos for the Google Podcast and for Google My Business. Those are human curated. It means that human beings have put the information together. So, we’re pretty sure that it’s true.

[00:13:03] The Brand SERP Guy (Jason Barnard): And then crowdsourcing, those are things like IMDb or Wikipedia where multiple people come together to create databases, create lists of information that are very well organised. And what Google then does is use that as the data that they feed into the machine at this stage. Over there on the left, that little green icon there is the data they’re feeding in of manually curated data and crowdsource data. And they tell the machine, this is almost certainly true, you can rely on this, and this is the data we’re using to train you to figure out how to understand and evaluate data from the World Wide Web. Remembering that data from the World Wide Web is not human curated in the sense that everybody is putting anything they want on the web. So, there is no inbuilt reliability to that information that Google can rely on.

Kalicube Pro Data Provides a Breakdown of Where Google is Getting the Facts

[00:14:01] The Brand SERP Guy (Jason Barnard): And if we look here, we can see from Kalicube Pro data the breakdown of where Google is currently getting its facts. In June 20, 2022. 70% is from third-party human curated sources like Wikipedia, Wikidata, IMDb, MusicBrainz. 17% is from Google sources that have been human curated: things like Google Podcasts, Google My Business, Google Books, Google Scholar, where human beings have curated the data and that data belongs to Google. Whereas the third-party human curated is data they’re using from sources that they do not control directly.

[00:14:44] The Brand SERP Guy (Jason Barnard): The third one is interesting, the yellow one. 11% comes from web facts. Now, web facts are data that Google have collected for the Knowledge Graph and have put into the Knowledge Graph. The algorithms have assimilated data from around the web, decided that it is true, and put it in the Knowledge Graph. But those web facts come from a seed set (of websites) that the engineers gave the machine. We don’t know exactly what that seed set is, but it’s going to include things like Crunchbase, LinkedIn, and so on and so forth. So, that seed set is being used by the machine to understand data from around the web that isn’t fed in directly and verified by human beings.

[00:15:30] The Brand SERP Guy (Jason Barnard): The green section of 2% is the really interesting one because this is what I’m calling web facts second generation open knowledge. This is the web facts, the 11%, without the limitation of the seed set. At the moment, it’s only 2%, but it’s growing. And that 2% represents Google, the algorithm for Google, pushing into the Knowledge Graph facts and understanding that it has found from around the web on any source including potentially your site

The Brand SERP Guy (Jason Barnard) Being a Qualified and Authoritative Source That Google Can Trust For His Sites

[00:16:08] The Brand SERP Guy (Jason Barnard): Some web facts, for example, have come from my own site because Google trusts my sites. It trusts my sites so the entities for which I am authoritative, the entities that it knows that I am qualified and authoritative to talk about. And in this case, it would be me, it would be the music group I played in, The Barking Dogs, and it would be the blue dog and yellow koala cartoon that you saw right at the beginning, Boowa and Kwala.

[00:16:37] The Brand SERP Guy (Jason Barnard): These are entities that I can talk about and that the machine will tend to believe me. I am an authoritative site on those particular entities, and it is using my site to create web facts second generation for the Knowledge Graph. And that is the aim that we all need to have because that 2% is slowly but surely going to overtake the web facts, the Google heat sources, and the third-party human curated content as the Knowledge Graph grows from a thousand billion facts to a hundred billion trillion zillion gazillion facts in, let’s say, 10 years time.

[00:17:21] The Brand SERP Guy (Jason Barnard): But right now, we’re not 10 years in the future. Google is just a child. It’s a child that’s thirsty for knowledge. It wants to build up its understanding and we need to learn to teach it. And that is the key for the coming 10 years or more.

Google is a Child That Learns Through Clear Explanations and Repetition From Authoritative Sources 

[00:17:41] The Brand SERP Guy (Jason Barnard): But here’s the catch, how do we teach Google? This child learns through clear explanations by repetition from authoritative sources. It needs that clear explanation from you. It needs repetition of that explanation from multiple sources and those sources need to be considered to be authoritative.

[00:18:06] The Brand SERP Guy (Jason Barnard): If we take the example of a child, that clear explanation, you are the parent. You can then say, go and ask grandma, grandma will repeat hopefully the same facts that you gave the child. Then the child might ask the brother or the sister who will then also repeat the same thing. If it’s a question about history, an authoritative source of the child might want to go to is the history teacher at school. If it’s about baking bread, the authoritative source might be the baker.

[00:18:39] The Brand SERP Guy (Jason Barnard): The point here is that the child has the information, understands the information from that initial clear explanation and by the repetition, but it won’t build confidence in that understanding until it hears it from multiple authoritative trusted sources. And those authoritative sources will vary according to the entity type, person, business, podcast, film, whatever that might be, the georegion, the country, and the industry.

Google is Faced With a Huge Problem: Fragmented, Contradictory, and Badly Organised Information

[00:19:13] The Brand SERP Guy (Jason Barnard): So, it faces a huge problem. Google is faced with fragmented information. It’s contradictory often because we’re not consistent when we post things online over time either as people or as companies. Every person who works for a company will express themselves slightly differently. What then happens is over time, not only does the information become fragmented across multiple different sources, LinkedIn, Facebook, Crunchbase, MusicBrainz, IMDb, Wikidata, and even your own site, the contradictory information that says slightly different things will build over time. Something that was said on LinkedIn five years ago might not have been updated. It hasn’t been updated on Crunchbase either, but it was updated on your site and on Facebook. And there’s a problem of contradiction which is going to confuse the child. 

[00:20:09] The Brand SERP Guy (Jason Barnard): The other problem is that the information on the web is badly organised. As human beings, we tend to look at it and think, well, yeah, that is quite organised. We click through links. We read pages. We can understand them. Google doesn’t have that advantage. The web is absolutely massive, and it’s trying to look at the entire web. It’s badly organised because we all organise things differently. Our brains work differently. We put things in different places. Google is faced with a massive problem of trying to sort through everybody’s junk.

[00:20:44] The Brand SERP Guy (Jason Barnard): The other thing is that we tend to think it’s quite organised because we search on Google and we find things, the solution to our problems, the answers to our questions relatively easily. And that’s because Google has taken the big mess that is the World Wide Web and pretty much organised it in a manner where we can generally find what we’re looking for. So, that idea that we tend to have as human beings that the web is reasonably well-organised is absolutely false. We just get that impression because the interfaces we use to interact with the web are the ones that are organising it for us.

Child Faced With Broken Plate Analogy 

[00:21:19] The Brand SERP Guy (Jason Barnard): The child in short is faced with a broken plate. It has the pieces of the puzzle. The pieces of the puzzle are spread all over the internet in tiny little fragments. And all of these tiny little fragments, it’s trying to put them together. And generally speaking, it probably can unless its a big mess. It will put it together and it will think, yeah, maybe and I have understood, maybe I haven’t, but it’s probably about right. Google tends to get it reasonably right. Sometimes, it gets it terribly, terribly wrong, and that’s when it’s got a spectacularly broken plate or it’s put the wrong pieces into the wrong plate.

[00:21:56] The Brand SERP Guy (Jason Barnard): But what do we need to do to help it with that broken plate so they can bring it together into the correct fully-formed plate, the completed puzzle? We’re responsible adults. We need to teach this child about our little corner of the internet. We need to give the child a reference. That’s called reconciliation. It’s a reference to which the child can then compare its plate to the completed plate puzzle. We then need to guide the child to show it what’s true, what isn’t true, and we need to point it to consistent corroboration on authoritative sources.

[00:22:42] The Brand SERP Guy (Jason Barnard): Once again, back to the child analogy, telling it to go and ask the teacher, the history teacher for something about history, asking the baker for some information about baking bread. That authoritative source is incredibly important. And it needs to be consistent. And it needs to be multiple corroborative trusted sources for the child to build confidence in the understanding you have given it by providing it with the reconciled, the defragmented plate, the complete plate puzzle. And it’s really, really simple.

The Process of Entity Reconciliation for Google

[00:23:17] The Brand SERP Guy (Jason Barnard): John Mueller from Google talks about a process of reconciliation. And the reconciliation is taking those pieces of the puzzle from around the web as we can see here and putting it together as a single plate. Now, the child will have different pieces of the puzzle in different places. It will need to bring them together. It will need to find all of the pieces. And it will need to find the correct pieces for the correct plate. And if you imagine there are 5 billion entities, probably 10 billion entities now in the Knowledge Graph, it’s juggling 10 billion broken plates. That’s a lot. It’s a big, big, big, huge problem.

Entity Home As The Point of Reconciliation For Google

[00:24:01] The Brand SERP Guy (Jason Barnard): So for Google, I figure it out. We have a concept called Entity Home which is one page on your website where you explain clearly to the child who you are, what you do, who your audience is. That is the point of reconciliation. And Google is actively looking for that. The Google child is looking for a point of reconciliation where the entity itself, you, explains who you are, what you do, who your audience is in detail clearly the adults explanation, the parents explanation.

[00:24:37] The Brand SERP Guy (Jason Barnard): Then, those are the lines of you can see in the middle there is the Entity Home with the plate which is the complete puzzle that you are presenting to the child to indicate to the child, this is what the facts are. This is what it should look like. And those arrows are you guiding it to all of those different sources that corroborate what it is you say, more or less completely, more or less accurately.

[00:25:03] The Brand SERP Guy (Jason Barnard): Now, more or less completely, naturally all of these places will have partial information. And the trick here is to add as much of the information as possible to each of them without ruining the user experience on any of those other sites. And the other trick is to correct any inaccurate information so that when the child gets to the site, finds the piece of the puzzle, the piece of the puzzle corresponds to the place it’s supposed to go in the plate. So, it is correct information and that the piece of the puzzle is the correct shape to fit in to the plate.

[00:25:40] The Brand SERP Guy (Jason Barnard): So, very simply, Entity Home, a page on your website where you explain everything, you provide a point of reconciliation for the child, you then with a completed puzzle of that plate, and you then point it to all of the sources that will corroborate. They need to be authoritative and trustworthy. They need to be consistent. They need to be accurate. And they need to support what it is you’re saying. Ideally also, those corroborative sources point back to the Entity Home, the point of reconciliation, so that the child goes backwards and forwards to all of these different places and sees the same information. So, not only has it understood the full plate that you’ve presented it with, but it is incredibly confident because everybody else confirms what you have been saying.

The Kalicube Process and How It Can Help You 

[00:26:33] The Brand SERP Guy (Jason Barnard): So, back to the Kalicube Brand SERP. As you can see when you do this well, you do get Google to represent your brand the way you want it to. You do get Google to represent your brand message. And you do get Google to present your audience with a great representation of your brand where you have the company slogan, the online courses, in our case, the visual identity, the social voice, and your product, in our case, the SaaS Platform.

[00:27:03] The Brand SERP Guy (Jason Barnard): Google has understood. But Google understands because we educate it. And what we’ve done to educate it, obviously we’ve been doing this for years and years. We’ve built this SaaS Platform, the Kalicube Process Platform that simplifies the whole process of providing that completed plate puzzle but also guiding the child to the corroborative sources and correcting those corroborative sources. So, that whole process of reconciliation is all packaged up into the Kalicube Process Platform to make it simple, efficient, and effective for you.

Applying to Other BigTech Algorithms What We Have Done With Google 

[00:27:47] The Brand SERP Guy (Jason Barnard): Now, I’ve done Google. Next, we’re going to have to learn to teach the others too (Apple, Microsoft, Amazon, Facebook…). They’re all going to have slightly different systems, but that concept of Entity Home and point of reconciliation with guiding the child and corroboration on authoritative sources is likely to work for all of them. The slight disparities and differences between them all is something we’re going to need to discover over the coming years. Thank you very much. That was Big Tech Algorithms are Children That Want to Understand and You Need to Learn to Educate Them. I’m Jason Barnard. I’m The Brand SERP Guy, and I’ll see you soon.

Similar Posts