James Dooley (0:00): Hi, today I’m joined with Jason Barnard from CaliCube and today’s topic of conversation is the algorithmic trinity. So Jason, for anyone that doesn’t know what the algorithmic trinity is, please can you explain that? Jason Barnard (0:16): Oh, right in with the meat and potatoes, James. Thanks a lot. The algorithmic trinity is the trio of technologies that all of the AI assistive engines use — whether it’s Perplexity, ChatGPT, Google AI Mode, AI Overviews, Copilot — all of them use the same foundational technologies, and there are three of them: Search engines, knowledge graphs, LLM chatbots. And the key is to understand that even if you’re looking at ChatGPT and you’re having a conversation with it — you’re looking at the chatbot, the LLM chatbot — it’s using the knowledge it has from its training data to have a conversation. But for information it doesn’t have in its training data, it uses search engines for recent information or niche information it doesn’t already have. And it uses knowledge graphs to fact check. WordLift, Andrea Volpini just proved that within ChatGPT there is a knowledge graph. People were saying to me a few years ago “oh that’s not true” — it is true. Fundamentally, nobody is going to reinvent the wheel to do the same thing for the same audience. All of these engines are trying to solve your problem as efficiently as possible with the best possible solution — and their audience is users. And Bing, Google, ChatGPT, Perplexity — all of them are using the algorithmic trinity: LLM chatbots, knowledge graphs, and search results. James Dooley (2:05): So here’s a question for you then. If ChatGPT has got its own kind of knowledge base — is that the same as Bing’s or is it different? Jason Barnard (2:16): They don’t appear to be taking that from Bing. Microsoft have got a huge knowledge graph. Google have got a huge knowledge graph. ChatGPT’s knowledge graph is actually integrated — partially at least — into the LLM itself. And geeky talk: if you think about how an LLM works with parameters that have gradually strengthened, you could actually just look at a parameter in an LLM as a node in a knowledge graph. So essentially it’s a very similar process if you look at it from that perspective. But they obviously have a product knowledge graph — now you have a product feed with ChatGPT, so there is explicitly a knowledge graph there — plus WordLift proving there is a knowledge graph behind the answers ChatGPT is giving. So… to defer to your racehorsing career, we’re off to the races James — the algorithmic trinity is boss. James Dooley (3:17): So with regards to the trinity then, is there any that you think is more important or holistically you need to be doing all three — making certain that you’ve got the knowledge graph, you’re in the LLMs, and you’re also ranking in the search engines? Jason Barnard (3:30): You have to have all three. It can’t talk about you if you’re not in the LLM itself. It can’t talk about you confidently. If you’re not in the search results, you’re going to miss out on all the stuff it doesn’t know — the niche stuff or the new stuff. And if you’re not in the knowledge graph, it won’t be able to fact check. It won’t be confident in what it’s saying. And if you think about how the engineers are building these things, they’re going to be training the machine not to say something it’s not sure about because they don’t want hallucinations — that’s bad user experience. So the idea of: does it understand who you are, what you do, who you serve, and that you’re credible — that’s fundamentally important. And the confidence the machine has in that understanding is absolutely key. James Dooley (4:17): Yeah, that makes sense. So with regards to the algorithmic trinity, when you say the knowledge graph — I know about knowledge panels in Google… CaliCube is helping me a lot to go and find all trusted sources, repeat and do self-collaboration across everywhere. There’s a knowledge panel, there’s a knowledge vault, there’s a knowledge graph. You can strengthen your knowledge graph score. How does it work with regards to in the LLMs? Is this going to be completely different in trying to get that information there? Are they going to be using very similar bits of self-collaboration data? Jason Barnard (4:55): That’s a lovely question because in fact all of the algorithmic trinity use the same data source — and that’s the web. If you can get in the web index, you can potentially feed all three of the trinity. Then the question is: what kind of information are they looking at and what URLs do they trust. But when you look at it from an individual brand perspective, they’re all looking at your digital footprint. So immediately you’ve reduced your playing field to: what is there out there on your website, on third-party websites, on your own social media channels — all of that stuff. All of that needs to be under your control. You need to control it, manage it, and optimize it — because that’s what’s feeding them all. Get yourself in the web index. Make your content friction-free to get into the web index, well organized so it’s well indexed, and then tasty for the algorithms so they want to use your content — whether it’s to build their understanding for a knowledge graph, add it to their training data for the LLM, or surface it in search. James Dooley (6:08): So, I’ve known you now for over six years and I’ve always known you as being the brand SER guy — that’s where I initially knew you from… But now with the algorithmic trinity, it’s more than just brand SER. You’ve got LLMs, AI overviews, and the knowledge graph… you need to be multifaceted — an AI engine expert. So what are you doing there to get that trinity to expand on who you are and what you do? Jason Barnard (7:04): What’s really interesting is the brand SER remains the focal point. I started with brand SERs — the left-hand side on a search result for your name as a person or a corporation. A lot of people were saying “it doesn’t really matter, I rank number one.” Craig Campbell said that to me once and I said, “But look underneath.” All that content underneath — Facebook, LinkedIn, review sites, right down to page 10 — that’s what Google is paying attention to for your brand. And by association: how it represents your brand, but also how it understands your brand and where it’s looking for information about your brand. That gives you huge insight into what you should be doing in SEO. So I wrote a book — The Fundamentals of Brand Search for Business — boring title, but it explains how you can build your entire digital marketing and SEO strategy just from looking at your brand search. Then I added knowledge panels — which is their understanding of who you are, what you do, and who you serve. Then, luckily for me, AI and LLMs are all focused on brand. So now at CaliCube, we’re working on: how does Google represent you in search, how does it represent you in its knowledge graph, and how does it talk about you in AI mode? What AI résumé do you get when you search your name? For me it’s been an evolution: brand SERs → knowledge panels → AI résumés… and that’s made it really easy to segue into optimization for AI. James Dooley (9:03): I think the massive one for CaliCube is the AI résumé… Now it’s made it 10 times more difficult because of query fan-out — multiple searches like “can I trust…”, “reviews…”, “who is…” Is that where you’re going to move more into the AI résumé? Because AI reputation management has skyrocketed. Jason Barnard (10:15): Yeah. Number one: we’ve got 25 billion data points, 74 million brands, and a million entrepreneurs in our system. We can tell you what needs to be done to manage that AI résumé — and AI résumés are much more interesting than brand SERs ever were. They can list multiple people with the same name — Craig Campbell, for example, there’s a country music star called Craig Campbell — so they’ll ask which one you mean, and they invite you to start a conversation. Then you get into the due diligence rabbit hole because they suggest follow-up questions — and not only do you need to control the initial response, you need to control the answers to all the follow-up questions. And if you’re really smart, you manage the follow-up questions so it’s easy for you. For me, when you search for me, it always pushes: “Tell me more about brands,” “Tell me more about CaliCube,” “Can Jason help me with my knowledge panel?” Because I’ve been so clear and consistent over the last 10 years. Consistency is key — consistency over time and across the web — and as human beings, we’re really bad at being consistent over time and over space. James Dooley (12:04): Yeah, for sure. The light bulb moment for me with the algorithmic trinity was… People argue “is SEO and GEO the same thing?” But it’s not just SEO and GEO — you’ve got your knowledge graph and knowledge panels. It boils down to: who are you, what do you do, why should you be served, what the LLMs say, and owning the brand SER positively. I genuinely think in 2026 people are going to say the number one ranking factor is the algorithmic trinity. Jason Barnard (14:30): Oh, I hope so. When I figured it out, I thought “Eureka!” This is it — and it’s simple. And if it’s really simple for me to explain, and it makes total sense as soon as I say it, then it means I found a universal truth that is useful. If all of them use the web index and all of them are using my digital footprint — you simplify your job to: how do I best manage my digital footprint offsite and onsite to feed the web index with the right content for each part of the trinity? The job isn’t difficult — it’s just a lot of moving parts. CaliCube Pro is designed to do that. It’s now tracking each of the three for any brand, organizing the info so you know what to focus on. And we’re creating smart AI assistance — not just wrappers around ChatGPT — multi-layered AI that understands your brand persona, KPI, business model, goals, USP, lexicon, claims, how you frame them, and it will find the proof. We’ve talked about this: claim, frame, proof. Jason Barnard (16:37): Online reputation management is essentially managing knowledge panels and brand SERs. Therefore I’m a world-leading expert in online reputation management. Machines love that logic. Then all that’s missing is the proof — and the proof is out there. Claim, frame, proof. Bingo. James Dooley (17:50): For anyone watching — Jason sometimes plays this down saying it’s simple. Cutting down a tree is simple with a chainsaw, but not with a blunt saw. That’s where CaliCube comes in — it shows what you’re missing. Without the right tools, you’re cutting across without direction. Now I understand: knowledge panels, knowledge graph score, claim frame proof for different facets… it starts showing up in AI overviews and you get that trinity in place. In my opinion the number one ranking factor in 2026 is the algorithmic trinity. Jason Barnard (19:42): To add to the tree analogy — if you chop it straight across, you don’t know where it will fall. It might fall on your house. You need someone who tells you where to cut so it falls the right way. I’ve seen people take a simplistic view — they do it themselves — and they come back a year later: knowledge panel lost, brand SER terrible, AI résumé mixing them up with someone else. Recovery takes longer than building it right in the first place. James Dooley (20:49): Yeah — exactly that happened to me. I did things at 100 miles an hour. I thought Google Books would help a knowledge panel, did eight Google books, and it created multiple KGM IDs. Now I’ve got different KGM IDs fighting for who I am. You came in and said: take a step back, fix the foundations, then build. Without the right tools, I didn’t have the chainsaw — and I didn’t have the guidance either. Branding is becoming more important for local SEO and lead gen across the board — who you are, what you do, who you serve. It’s going to become even more important. Every update pushes toward brand and trust. And the cost of information retrieval becomes cheaper for Google/LLMs if they can trust you — because your third-party sources repeat and corroborate who you are and what you do. Jason Barnard (23:30): Trust and confidence are overlooked. People say “it’s just using your website,” but it doesn’t work like that — by default they don’t want to trust you. What we did with your website was organize it, create the pages that make the difference, because we want machines to systematically come back to your site to learn who you are and how you want to be represented. They’ll believe you if you claim and frame on your website, and you can prove it off your website. If machines use your website, it’s easy to pivot your career or business. That’s why the “entity home” matters. James Dooley (24:55): Yeah — having an entity home is crucial. If I pivot, I can change my website — I can’t do that properly on LinkedIn or Twitter. I can’t use schema in the same way. That becomes part of the algorithmic trinity. Jason Barnard (25:38): Exactly. If you master search but not the LLM and not the knowledge graph, the message is never straight. You need all three. We’ve pivoted CaliCube’s messaging in less than a year — from SEO-only to serving entrepreneurs and corporations. And I like this term: top of algorithmic mind. If you get your brand into the top of that algorithmic mind, it’s like being top of human mind. And there’s value in associating with authoritative entities — like Semrush — and letting that authority rub off. James Dooley (28:38): Yeah, that comes back to the trinity again — improving connected entities strengthens you overall. For anyone watching: let us know if you agree or disagree — is the algorithmic trinity the number one ranking factor? If you didn’t know what it was, reach out to Jason Barnard at CaliCube. Jason, it’s been an absolute pleasure. Jason Barnard (29:28): Yep. Brilliant. Wonderful. Thanks, James. That was fun.