How to Pick an AI SEO Agency (4 Factors Most People Miss)

Our AI SEO or GEO strategy is something we call prioritize GEO. And we built this specifically to help brands avoid this mistake of chasing these tactics that we don't think really matters. So all of these on-site tactics like you got to change your schema, you need to add FAQs to your site, you got to have question and answer format for your headers and things like that. We are grouping all of them into tier three, which we think is the least important tier. The only queries that matter when you're doing AI SEO are queries where the user is asking the AI bot about product recommendations or service recommendations that are in your space because the old-fashioned SEO where you do top offunnel content around tips for this and how to do that chache BT and all will just answer that by themselves. So they're not going to cite you, mention brands. This is the way brands and marketing teams can get themselves in the discussion and influence LLMs.

If you're looking to hire an AI SEO agency, how do you even pick? AI SEO is so new, people can't even agree on what to call it. Call it GEO, AEO, AI SEO. And so, what do you look for? We are one of those agencies and we've been doing regular SEO for about a decade and recently like a lot of other a SEO agencies we're also doing AI SEO and we have thought carefully about how we should do AI SEO and therefore what makes a good AI SEO process and as a result we've seen what agencies are doing things in what we think is the right and responsible way and what other ones are recommending tactics to companies we end up talking to that we think are just not doing anything. And so these companies are being sold kind of nonsense tactics. So we're going to go over in this video four factors that we in our experience actually doing AI SEO for dozens of clients think differentiates someone who knows what they're doing in AI SEO and not. And so you can decide to use these factors to evaluate an AI SEO agency you're thinking of hiring. Factor number one, does the agency understand the connection between traditional SEO and AI visibility or GEO or AEO? Specifically, what we've noticed is the vast majority of content around AI SEO and getting visibility in LLMs and search engines that are driven by AI is around content tweaks and things you tweak on your site. But the problem with those is they don't actually increase

how often your brand is exposed to LLMs or discover. All these tactics are supposed to do is make the content on your site slightly more digestible for AI bots. And in our experience, it's just been completely unnecessary. And we think the actual real important thing that you need to do is be ranking in traditional SEO for the searches that AI bots like Chat GBT do when they're trying to answer a user's question. So, let's go into this in detail. Our AI SEO or GEO strategy is something we call prioritize GEO. And we built this specifically to help brands avoid this mistake of chasing these tactics that we don't think really matter. So, all of these on-site tactics like you got to change your schema, you need to add FAQs to your site, you got to have question and answer format for your headers and things like that. We are grouping all of them into tier three, which we think is the least important tier. And there's just on-site tactics. Most of the evidence that we've seen and in our own research, this doesn't do much. Instead, we think the place to focus is on tier one and two, which is to expose your brand to these LLMs like chat GPT via traditional web search, meaning SEO. Why? Because the only queries that matter when you're doing AI SEO are queries where the user is asking the AI bot like Chat GPT, Gemini, Claude about product recommendations or service recommendations that are in your space because the old-fashioned SEO where you do top offunnel content around tips for this and how to do that, Chachi BT and all will just answer that by themselves. So they're not going to site you. mention brands. If you sell project management software and someone's asking chatbt for project management tips or strategies, it will just have a conversation and answer that and give it tips and strategies without

mentioning your software once unless the conversation naturally turns to how do I implement this and use software. So that also represents the user moving the conversation towards product recommendations. So the number one thing that you have to understand is AI SEO is about getting these AI tools to recommend your product or service when the user is asking for that or the AI tool has determined the user needs that. So that we call bottom offunnel content. So for bottom offunnel content LLMs like chat GPT essentially will use two different buckets of sources blended together to answer the question. They're training data and retrieving recent sources aka web search like googling it. It's not strictly googling it. They also have some other sites that they have partnerships with like Reddit and all, but it's mostly googling it. And that's an easy way to think about it. Training data and googling it. And so the training data bucket, of course, it would be great to get into that, but that's not immediately controllable by you. Why? because we don't know how often OpenAI and these other AI companies retrain their models. They don't reveal exactly what goes into the training data. We know roughly it's sort of like all written works of humans. And so getting into the training data is a long-term brand awareness project aka all of your marketing eventually helps associate your brand name with your product space. So big brands like Salesforce will always kind of be associated with CRM, for example, and they're going to just naturally be in the training data. Or a consumer brand like Nike will be associated with athletic shoes and they're just going to be in the training data. But if for most upandcoming brands that aren't the incumbent in their space, you're probably not going to be in the training data. So then the other thing that chatbt and these LLM use to answer users

questions around products is to search the web. Why do they do this? They do this naturally. Even if the user doesn't click the web search button, it will determine by itself this question from the user is going to require web search. So I'm going to search the web. This is the way brands and marketing teams can get themselves in the discussion and influence LLMs is via web search because that's controllable. You can show up in those searches. So the fundamental strategy that an good AI SEO agency needs to know is how does that work? And they need to have a process of how are they getting you into these searches. So you need to be asking questions like do they understand this much? If you've watched this much of the video, does it seem like in your discussion when you ask them what are you actually doing? Are they just mentioning on-site tactics or do they understand this connection between traditional SEO and AI SEO? And then do they have a process of getting you in there? In terms of our process of getting our clients in there, this is where tier one and tier two come from. Tier one says, well, the first most controllable thing that a brand can do and that we can do for our clients is get their own content living on their site to rank in Google for the what queries these bottom offunnel productcentric queries. So in our CRM example would be like best CRM for small business, best CRM for dentists or lawyers or whatever you're selling is know those specific queries and have a process to rank your own content for those queries so that then when the LLM's and search some version of those queries, you have some chance of being in there. You're in that discussion. And tier two is do outreach to the other sites that are already ranking list post, best software post, and it doesn't have to be software. It could be best e-commerce product lists, service businesses, and get in those and reach out to them and say, "Hey, can you mention our brand? Here's why, whatever. " This is how you actually control whether your brand is exposed to these LLMs. But everyone else is just talking about tier three. Oh, we're going to change the schema on your website and do this, that, and the other. If your agency is only thinking

about tier three, and otherwise they just do content as normal, in our opinion, they don't understand AI SEO. So, that's factor number one. Factor number two, do they prioritize bottom offunnel content over top ofunnel? Why is that important? Because of what I said earlier in factor number one, LLMs and AI SEO doesn't need to even site sources for a lot of top offunnel queries. I urge you to test this in chatbt yourself. Ask some basic how-to type query in the field that you play in that your customers may ask. It may site sources, but the percentage of times that you click on a source when you're just asking chat GPT some basic how-to query. Think about that yourself. How often do you click on that source? It's not very often. It's not like old-fashioned SEO in Google where Google didn't used to. Now, Google has their own AI summary that applies. That's part of AI SEO. And what I'm saying using chatbt as an example applies to Google AI overview mode, Gemini, Claude, and all of them as well. Their details of how they work are a little bit different, but the core principles apply. But in traditional SEO, Google didn't used to answer the question. You would just present websites that you had to click into. So at least for these top offunnel queries, they didn't convert to actual customers very well as it is, but at least they got you traffic. In AI and chatpt world and AI SEO, they don't even get you traffic or they get you very little traffic. And so you need to be producing bottom offunnel content to go after these product related queries. people literally googling and asking for your product because the LLMs and the AI bots will also search the web as we just concluded for their product. That's factor two. If they don't have a good understanding of that, if their strategy when you ask them what is your content strategy, what will you be producing? What keywords will you be targeting? Are you doing traditional SEO as well? What are you doing for AI? If they don't understand that you need to go after and rank for these bottom of funnel queries, if they don't understand that these AI tools will be searching that when they're answering the user's question, we think very strongly, they don't

really understand how AI SEO works. Factor three, the actual production of the content to rank well for this content and to have the level of detail you need to have these AI tools decide to use your content when they're searching the web. You need to have very high quality content that discusses the nuances of your product or service and exact features very well. And to do that, if it's a full AIEL content production process where they're just using AI to research these topics and produce the content, it's not going to work well. rank in Google very well, and it's not going to have the level of depth needed to surface the right answers for these LLMs. Most agencies these days use some form of AI assistance in generating content, but ask them the details. Who is coming up with these ideas? Can you show me an outline? How much is the human editing this? What about the final draft? How much is the human drafting that or editing the actual final draft? You need to ask this and understand this at a deep level. If you have an agency that's producing basically almost untouched AI generated writing, it's extremely precarious. That kind of content can often get you quick rankings and traffic. It'll spike up because they can produce a ton of it really fast. And then we've seen graph after graph posted of folks in our industry where that traffic eventually tanks. So be careful. Last factor, factor four, how do they measure AI visibility and decide on success? So there's a ton of AI visibility tools out there. Number one, do they use them or

do they just use your basic analytics and measure chatpt traffic? That's not enough. Here's why. the traffic that your normal Google Analytics uses that just shows the traffic from Chatypt, Perplexity, Gemini, and all that's not enough because of the reasons I mentioned before. All that's showing is people who actually choose to click into the sources. But if you're mentioned as a brand for one of your most basic kind of topics in the CRM example, if you're mentioned for best small business CRM and that's exactly what you do, that's a great win. And you may not even have your own article there. You may have done tier 2 of our pyramid and done outreach and be mentioned in multiple people's sources and you're mentioned there, for example. The actual traffic from ChatBT doesn't matter. The fact that you are a brand mentioned in those prompts is what matters. So, they need to track that and you're not going to see that inside Google Analytics. So what you need to look for when they're actually tracking in these other tools is do they understand that measuring AI visibility on a single prompt basis is actually not enough. Here's why. There's a massive variance in the actual results of these AI tools. So for example, if you ask best small business CRM today and then you ask the same thing again tomorrow, it'll give you different results. If you ask and your colleague asks, then just open a new incognito window and ask again, it'll give you different results. There's a randomness built into it. So, just whether or not you're showing up for an one prompt is not enough. You're going

to be measuring random noise. In addition, you don't know, unlike traditional Google where there's known search volumes because everyone searches these little specific phrases. You don't know in chatbt for example exactly how the user is going to talk about it. We've asked our clients that said they found us on chatbt send us the chats and they were long chats about growing their business and at some point organically the concept of content SEO AI came up and then they asked for help and then eventually it recommended agencies including ours. That's not reproducible. And so what we think is you need to be measuring this at a topic basis. For us, we've built our own tool called tracker. ai where we measure the our clients visibility for different topics. Under each topic is multiple different prompts. That way, in this particular training organization, if you asked about CompTIA A+ training in a certain way, you're mentioned. But if you ask about the same thing in a slightly different phrasing, you're not mentioned. Which one is right? The answer is neither of them are right. There's a probability of whether you are mentioned. an overall visibility percentage within a given topic. And so if all you're doing is looking at a single percentage of your whole brand, but you're not for each topic, there's just one prompt, you're not doing it properly. Ask the agency in detail exactly how are you measuring. Show me an example. And that way you know whether they're just going to be chasing ghosts in the cloud or not. So

take these four factors, do a rigorous interview of your agency, and we hope that you're able to find a good one. And if you're interested in working with us and you like the way we're thinking about it based on these factors and what we've explained how we're thinking about it, you can reach out to us at growandconvert. com. If you like this video, don't forget to subscribe. You can also get the audio only versions of these shows wherever you get your podcasts. And you can follow us at growandconvert. com/newsletter for any articles and updates for when these videos come