Benchmarking Brand Discoverability: LLMs vs. Traditional Search Featuring JR Oakes

Garrett: Welcome back to The Next Chapter of Search produced by SEO Week. And iPullRank I’m Garrett Sussman. If you haven’t already gotten your tickets, New York City, April 28th to May 1st, we are going to have 35+ of the most amazing speakers in our industry across SEO, data, AI, content, you name it. And I am joined by a brilliant individual himself. I’m excited to be joined by JR Oakes, who is the VP of Strategy at Locomotive, this amazing agency that does everything across the board. Thanks for joining me today, JR, appreciate you, man.

JR: Glad to be here, man.

Garrett: What is going on with SEO? Like the whole industry has been in flux for the last 12 months, 24 months, 10 years, if you will. From your perspective, what is the current state of search?

JR: I think that Google has been trying to solve the content issue for the last couple of years, and I don’t think they’re there. I think content is 10xing right now with all of the generative tools. I can’t watch or read LinkedIn anymore because every tool there is how to 10x your content strategy. And at some point, I think that becomes a zero-sum game. I think we see Google moving away from a lot of the traditional 10 blue links and into like walls of content, walls of products, walls of images, walls of social and discussion boards and trying to go where kind of real conversation is happening instead of this 10x content. So, I think we’ll continue to see this.

I think Google has been lifting up like traditional brands, but I think that’s somewhat of a fallacy. It’s like Craftsman. Craftsman used to be known for like the tools that were like the best quality and everything. And then now it’s some brand bought them up and then it’s putting the Craftsman name on there and it’s not the same quality, right? But to Google, it’s like, I think they would see Craftsman and say, oh, this is a really good tool brand. So, let’s give them some quality for tools, but it’s really not. There’s a brand effect there, but I don’t think they’ve gotten, I think they’re still having those issues like they have with Wire– not Wirecutter, but some of the, I would say the kind of the add-on coupon code content sites that they have for-

Garrett: Reputation abuse.

JR: Yeah, reputation abuse. So, I think they’re still working on that. I think they’ve gone hard in the direction back to allowing AI content after kind of smashing a lot of that for a while. And that’s a little scary, but I don’t know, we’ll see. We’re all kind of spectators here from Google’s perspective.

Garrett: From your perspective, exactly. Do you think Google, do you think, because it’s so easy to both overestimate them and underestimate them. Do you think they know what they’re doing, or do you think they’ve lost the plot?

JR: I don’t, either Google is, Google just feels like they’re messing everything up. And I’ll give you an example. Like they’ve been building this Vertex AI thing for the last two years and I’ve watched it. I looked at it a couple of years ago and it’s this whole search and convert thing, right? On their platform where you can put in a website and it uses Google search to search it and pull snippets. And it’s a grounding thing for Google search. And then it also had things where it would like watch users through the site and just a lot of really cool tools. That thing has been renamed like three times in the last two years and new UX applied. Like you can’t even go back month over month and expect the same experience or the same tool or the same anything there. So, I feel like they’re just like trying to do everything and they’re not doing like anything like really well. The Gemini stuff is fine. It’s, I’d never use it for really anything. They still have search. What is it? The deep search on Gemini 1.5, which that doesn’t make sense. I don’t know. I don’t know what Google’s doing, but it’s not winning right now.

Garrett: And as more people use it, it kind of leads into your whole presentation SEO Week where you’re going to be talking about brand discoverability on these LLMs. You mentioned the brand, you mentioned the LLMs. Can you kind of give a preview of how you’re thinking about things? Why should people come and check out your presentation?

JR: Well, I’ve heard a lot and I’ve seen a lot of tools that are like getting, trying to apply like the same paradigm of search to language models. And I just don’t think that’s there. I think there’s a lot of reasons why language models need to be kind of treated differently, right? Because you have deep search, you have search, you have memory, you have people that have put direct, specific directions in the back of language models. And a lot of this information that traditional search doesn’t have, right? So, thinking about, oh, my brand came first versus second, like I think it’s just insane to kind of think that like that with language models, where I thought, what we tried to do is really look at a specific niche, which is a lot of kind of more B2B SaaS companies, but also look at it from a brand perspective and a brand discovery perspective.

So, if I’m a user and I go and I search “best CRM provider” in ChatGPT, Anthropic, Google versus– or Gemini versus Google search, like what is the experience of that? Like what brands am I shown, right? And we did that across, I think, thousands of keywords. And then we’re also looking at specific modifiers. So, we have, I think around 30 to 35 different modifiers, like best, top, enterprise, SME, a ton more, GDPR compliant, right? Bunch of those, like, yeah, different modifiers that are classed out. So, we’re going to be looking at it, not only across like distinct industries, but also looking at it across those different modifiers. And then also taking more of a temporal aspect to it of going back in time and saying, okay, what’s changed with the first GPT 3.5 versus GPT 4.5, right? So, we’re also looking at, I think, 17 or 18 different models and historically in search results to kind of look at what changed, but really putting that, thinking about I’m going and I’m searching for this type of software provider on a language model versus I’m going to Google. And then what is the brand discovery experience there, right? And looking at it just from that lens, not a ranking experience.

Garrett: I love that because it’s really hard. I think a lot of people are trying to understand the complexity. And we always, one thing I keep coming back to is this idea of like, we want precision when we just really have to focus on general accuracy at this point. Like precision’s kind of out the door, not that we ever really had it, but with these, it’s gone.

JR: Yeah, I mean, I think it’s interesting though. We’ve already seen some interesting things like just how much more brands are being exposed now versus in some of the earlier models. It’s like in some cases, like two to three times, more brands coming back from the same type of search results and still like some sites that just dominate all models. And one of the cool things that I think, and this is Robert Paget on our team said is that, I think one of the most interesting thing is the AI categories, right? Because AI categories have been around for the last like year or two, right? So, we’re actually seeing them from infancy to growth and it’s not these like established brands like Red Hat, IBM, these companies have been around forever. We’re seeing their like growth and their content growth and then coming about. So, we can kind of use language models and use some of this data to kind of see what they’re doing, right? What are companies that have been around like less than a year doing and showing language models and Google and what are the kind of the differences there?

And I think that’s really cool because I think that gets you to more kind of tactical actions that we can say this person did this and they’re winning in this category for this reason, right? So, I think that’s cool.

Garrett: I think, and it leads right into, I’m curious, do you have one, speaking of tactical, like one tactical thing that people can do right now, that you’re going to expand upon in your presentation, but what’s like based on your research, what’s something that folks can do this moment?

JR: I would create a brand reviews page. I think that is the lowest of lowest hanging fruit because on Google you have, I did a thing a few years ago about like, is like Google, like the PAA questions, like is Google legit or is Apple legit? And Apple, I mean, Google had like, “why is Apple a bad company” showing up in those? And I was like, you’re kind of forcing people into an opinion there based on some of these, but you have so many of those, like I hear about this product or service and then I’m going to go and say, all right, let me search them and see kind of what they’re all about. Right? And then you have is this service any good or reviews or reputation type queries about companies. And that’s big on search. It’s also big on language models, right? And the thing about it is, is that you can, if you have a, I don’t know, a, a traffic reviews page and somebody says, is a trust any good, they’re RAG models and search is going to pull out that page. So, you must assuredly have a voice in how you’re perceived there.

And Google does, I mean, the language model doesn’t know whether it’s real. I mean, I guess they would know what’s your site, right? But I don’t know whether they’ll have that presence of mind. They’ll just be like, oh, this company is like really amazing because they have, you know, five stars on Yelp and this and this and this and all these testimonials. Right? It’s the easiest of easy, have low hanging fruit there, I would say.

Garrett: I love it. Because it does very much, especially with LLMs as search engines feels like a wild, wild West where so many of these older techniques currently work and you have like a ChatGPT that’s not thinking about it with the same sort of algorithmic parameters and ranking signals and all that you’re saying about like authority and trust. It just doesn’t necessarily apply. I guess it’s somewhat dependent on Bing’s index but that’s still volatile. I feel like we don’t know what ChatGPT is doing.

Dude, so what’s next JR? Like, what do you anticipate is the next indication of search? You’re talking about a little bit but where do you think we’re going?

JR: I think the packaging and the accessibility of search gets better. You have the demos teased with ChatGPT or with open AI about having that model where you have your phone with you and they’re kind of interacting with the environment. You have Project Astra from Google, you have Lens, you have this multimodal experience, right? So obviously at some point, you know you’re going to have an experience where you’re like putting up your phone and it’s like I want to redo my kitchen, right? And then there’s this interactive experience. I feel like the packaging and accessibility of search gets a lot more really interesting over time.

I think that the value proposition between content creators, written content creators and search engines and LLMs will be kind of refactored because you can’t have people 10Xing or creating 12,000 pieces of content a year. And you can’t have agents that are visiting websites and kind of not seeing ads or not giving the creators back anything. You can’t have those and that exists for a long period of time without some kind of really big change. So those are the barriers I would say I would say are pretty ripe for change.

Garrett: I, and who knows, maybe that’s a good thing. It’s like once content creation is not just about visibility in organic search channels and more in other places, the quality improves. And like, then it’s on the LLMs to figure out like the AI generated crap versus like actually authentic content that wasn’t necessarily designed for them.

JR: Yeah, well, I think the detail has to get really good. Right, like the 99% will be like covered well by language models and search engines. And it’s that 1% that I think has to get really great.

Garrett: Well, there you go. I have so many brilliant things. I can’t wait to have more conversations with you. But for those of you who have not already bought your tickets for SEO Week, April 28th to May 1st in New York City, come chat about the future and the ridiculousness of LLMs with JR, myself and all the other brilliant SEOs and developers and data scientists and content creators that are going to be there. Thanks so much for joining me, JR. I really appreciate your time, man.

JR: Yeah, looking forward to this conference.

Garrett: It’s going to be huge. See you then.

JR: Yeah