Building Third-Party Trust Signals to Boost AI Recommendations for Local Service Businesses in 2026

**James Dooley** (0:00)
How to get AI to recommend your brand? The amount of people in the community that are asking, how do you get cited in ChatGPT? How do you get cited in Claude? How do you get cited in places like Gemini and these other LLMs?
When people are basically turning around and they're looking to get cited, what is the best way to get recommended into these LLMs?

**Kyle** (0:26)
It's a huge load of question there. So there's a huge amount of nuances involved. But one of the easiest ways is you have to be able to clearly define what your brand is and what you do. So you have to be able to answer a big list of questions of what your brand is, what services it offers, why you're bloody brilliant at those services, and then also competitors. Who else offers these services and how good are you versus these competitors as well. So it's very important when trying to get AI to recommend your brand, that you go after all these different signals and make sure that the information is concise, it's filled with semantic triples. Anyone who doesn't know what that is, it's subject, predicates, and objects. It's the typical way that a machine understands black and white. It does not understand the gray area. You have to feed it almost like a child. You have to say, this is the way. There's no alternative, this is the way.
And you feed it the information like that and it kind of understands it a little bit better. Basically remove any ambiguity out of your brand and get those articles to be, or the mentions online to all be the same. It's almost like you can imagine like back in the day, we used to spam the internet when everyone was like, you can't spam anymore. Penguin came out, Panda came out, it was the content morning. You can't spam anymore, but you're not spamming. What you're actually doing is, LLM understands based off machine learning that the more in-depth you go for entity and the more you mention the entity connected to the predicates and objects, it's building its knowledge base of what this is. And that's the key. Spamming is me just saying, I am James Dooley, I am James Dooley, I am James Dooley, I am James Dooley, but you're not. You're going out there saying, I'm James Dooley, I am 26-year-old, obviously you're not going to say, I am. You say James Dooley is 26-year-old, James Dooley is.
You define it in a way like that. James Dooley is a successful business owner. James Dooley owns 700 businesses. That's how you define it. So you would skip the I am and that's the semantic triples coming in there.
That's what you'd be looking to do when it comes to getting to AI to recommend your business and trying to get more trust signals.

**James Dooley** (2:50)
So with regards to branding, so people are looking to go, how do I get recommended by AI for my brand? Let's talk about personal branding. So now when you're being recommended, it could just be someone's looking to, an employer's looking to employ an employee and you're an employee and you want to have a positive sentiment online about what you do. It could be that somebody like yourself that's got a knowledge panel, that your next business partner could be coming along to then basically turn around and say, okay, I want to JV with yourself. Or it could be looking to sell the business and you're looking to get, not just recommended, but a positive sentiment about your personal brand.
How important is the third party sources in manipulating or LLM seeding or getting that brand to have a positive sentiment? How important is the third party sources in that?

**Kyle** (3:51)
I like the word seeding. It's not manipulation. We're not manipulating anything here. We're just spouting the facts and we're spouting it loud. So we're not trying to manipulate anything, but we are making sure that the facts are out there.
It's vital. I'll say it's the most important pillar when it comes to LLM optimization. You need to be being clear, concise, and confident in everything that you mention online, and not publishing anything that's ambiguous. Because if you publish something that's ambiguous, we like to look at it as an off-page topical map. If you're publishing any bit of content or there's any bit of content online that's being published about you that is ambiguous, it lowers the confidence score in the entity itself. So it lowers, it might not understand, it might confuse you with a music composer, it might confuse you with something else because it's not confident enough that your personal brand is James Dooley.