Skip to main content

How Small Businesses Win AI Search Optimization and Citations

Logan Miller · November 15, 2026 · 14 min read

The shift toward generative engine optimization presents an unprecedented advantage for agile mid-market brands and small businesses. Recent independent research reveals that top citations in AI search overviews frequently belong to pages ranking on page two or beyond in traditional search results—pages with zero intentional legacy SEO. Large enterprise competitors struggle to maintain depth across broad product catalogs, leaving a critical gap for hyper-focused ICP content. Winning sustained AI visibility requires establishing multi-platform proof, specifically leveraging video content on channels like YouTube, which LLMs heavily prioritize for source verification. While traditional top-10 search rankings remain a key signal for engines like ChatGPT, Gemini, and Perplexity, AI models ultimately prioritize content consistency across trusted channels over site authority scores alone. By delivering deep, platform-verified expert content on high-buyer-intent queries, smaller teams can outpace larger competitors in AI recommendations.

Why Unoptimized Pages Beat Enterprise Brands in AI Search Overviews

Logan: How does it work for small businesses without a big marketing team? This is where it gets really interesting., there's been a lot of studies that have been put out there. I know AirOps has done some great studies. That same Princeton study has some really good information on it. But a lot of the most key players that are getting cited are not necessarily massive companies. I've even put out, independent research and information that shows even screen recordings of my searches for certain things that are in very, very competitive, search territory that when you look at what's in the AI overview versus what's on the page, sometimes the most cited source in the AI overview the top source, is not even., it's on the bottom of page two and has no intentional SEO done to the entire website, let alone that page.

Logan: It really it really is an interesting, an interesting case study that you could you could say that, honestly, small businesses might have one of the biggest opportunities to get more customers right now because of the newness of AI search and how quickly they can capitalize on getting traffic from really high buyer intent queries. The other thing is, I think the biz. The biggest upside that a small business has is strategy. A lot of big companies in competitive market spaces within the SEO side of things are targeting really, really big queries and really, really broad topics. And they're not niche companies, so they're not niching down on the content that their actual, every single one of their ICPs is really asking., if you have an ideal customer that you have a really specific offering of, you can answer the exact questions that they have likely in much more depth and with much more expertise than this big company really could.

Logan: A great example is. And we'll get back to this because I have a client that this is a great example for, but they're a they're a law group. And, if they were trying to win for being the best lawyer in all of San Diego, that'd be a really hard a hard task for them to achieve. However, they know that they have a very specific type of case that they know makes them really good money, that they're able to do a lot of good in the world in, and is something they have a really great win rate at. Those are all things that are just better for the world in that if more customers knew that they were a great person for that in the San Diego area, then their com. Their company would make more money, and those customers would likely be better served than at some big firm that doesn't have a specialty in this.

Logan: And they probably pay less money for it. I think that's a really great opportunity that AI search has created for small businesses. You don't need to go for these huge queries of being the best service provider of whatever you provide in the entire area. What you need to do is answer their exact questions and be the best provider of the actual thing that they're looking for. And you can decide now to target more specific things. If that you have products, if you if that you have products that have great margin or services that are really turnkey and you want to get more customers for specifically that, then you can just make content that is specifically targeting that. You don't have to go so broad anymore. You don't have to focus on broad keywords. You can go really, really hyper targeted and really just sniper zone in on one specific customer type and one specific product offering or service.

Logan: I think that's a huge opportunity for small businesses right now. And it's very specifically one that's very difficult for large businesses to do because you probably only have a few of those things that you really can hyper focus on. A huge broad company has hundreds of those offerings and customer types That if they were going to try and tackle every single one of them, they'd have to spend so much money and do so much upkeep on it so frequently that you would still have a chance of beating them. Moving on to the next question. A client who went from a client who went from never mentioned to recommended, what changed? Great example. We're gonna go back to that same law group. They are currently only, I wanna say, a couple months into publishing content consistently. And they had some videos that immediately got over a thousand views, but they also had some that got barely any at all.

Logan: And when you look for some of the very specific queries, again, for their specific, customer that they're trying to target and for specific cases, you get them popping up, which is great.

Another use case of another client, they, they're a builder of golf simulators, both for residential and commercial spaces. This was really interesting. Because they're only, I wanna say, three months, four months into publishing content, and their content is really fantastic. It's very specific. They're obviously very knowledgeable. They're experts in the space, so that helps. But we had a piece of content that there was a few short videos. I wanna say two or three short videos, one long video, and three written pieces of content on the website that were all about the same topic. And when I asked a really broad question about who the best golf simulator builder in their region was, I got them popping up as the number one source, which is obviously what we want.

Logan: You wanna go for those niche questions first, and then once you can start winning the really broad questions, and that's even better. But it cited a really specific piece of information that was only, I think, two weeks old in the actual AI response. It said a bunch of other reasons, but it was something that I clocked that just as I was doing a native search on ChatGPT, and I have a screen recording of this. I did a native search on ChatGPT because we're running thousands of query tests for them every single week. But, that's going straight into our system and our database for us to analyze the broader metrics of and for our system to optimize the really small details of all their content on, but I'm not seeing the exact searches play out. Sometimes for my clients, I go on and I search exactly how I think a customer would search.

Logan: I utilize the actual tonality that our system is telling us that their customers are using. I will, a lot of times, even do voice to text into the into the platform, whether it's ChatGPT or Clawd or Perplexity or Gemini. And I will look at how it responds to not just a clean question of who's the best golf simulator builder in the Southeast. I say it how a customer would say it. Hey. I have a house that is in the Southeast, and I'm looking for a builder who can make me a custom golf simulator. But it's specifically a new build, and, and I need it to be able to match with the rest of my contractor. Should I go with a regular contractor, or should I farm this part of it out to a specific contractor that builds golf simulators?

Logan: I would like to get, just a few sources that are gonna be able to give me a bid on this, and I want an explanation of who the best is and maybe why. That's how your customers are speaking. They're not so eloquent. They're not so direct. They're not so to the point, and they're honestly. They're not they're not even typing half the time anymore. They're just talking to ChatGPT like a normal person. And so when you start asking questions that way, you can start finding the results much easier. But what really changed for them was, well, a, they started getting a bunch more referrals from AI search and a bunch more leads. They've even gotten multiple customers from it now. And so that's been great. We're really, really happy about that because the ROI on that for them is massive, without naming it.

Logan: We're in the 10 to 20 x range, but, I have to verify that. And, and on the technical side, what changed is they have a lot more written content on their website, which we can talk about the importance or not importance of that later. But they also have social content going out in short forms very consistently. They also have long form video content going out very consistently. They have other, partners that they're even, making this content with that are being mentioned by name, that are experts in their field, that are also being, cited over to frequently. And so the frequency of content being published, the different places both on and off-site is very important. And why real. Why we really like video content is not just purely from the data that shows that video content is getting cited significantly more, specifically in certain LLMs, and specifically YouTube being one of the most cited sources for some of these LLMs now.

Logan: We really like it because it's very easy for platforms to tell if a video is AI generated or not. It's not as easy for platforms to decipher whether or not content is being made by AI if it's written.

It's still very possible, but it's not as sure of an answer. It's very difficult to perfectly mimic human on camera content that no computer or no system or no search engine or generative engine can decipher whether or not it's AI. And. Because there's just so many more data points. However, when you're on camera and you're speaking expert knowledge and putting out non commodity, expertise and experiences and case studies onto the Internet and you're putting them on video and you're putting them out there, that's really hard to. That's really, really hard to, to fake. And AI knows that, and that's why it likes it.

Logan: And then when it finds it in multiple places on the Internet, then your trust from all of these different platforms, everyone was talking about Google KVT scores and your different, site authority scores for a long time, but LLMs have something very similar., they wanna know that it's coming from a trusted source, and then they wanna be able to verify that not just because they've trusted them in the past, but because it can be verified on a bunch of different platforms and in a bunch of different places. Going on to the next question, do I need to rank on Google before AI will cite me? Yes and no. It really depends on which AI platform you're going for in terms of getting citations.

Logan: , you can go to a few really interesting studies. Ahrefs did a really interesting study that cited that 60. Sorry. 76. I'll go back to that. Ahrefs did an interesting study earlier this year that stated that 76.1% of AI overview citations came from the top 10 pages, and that was in, July 2025 not this year. That changes really often. Now there are certain things in place, Google really only wanting ChatGPT to be able to cite from the top 10 sources on Google's search engine. But ChatGPT doesn't just pull from Google. ChatGPT has a bunch of other outside sources, a bunch of other outside sites. It also. It really likes pulling from Bing. It really likes pulling from even similar, results to some other LLMs. And, also, it matters whether you're talking about ChatGPT, on a browser, on an, actual computer app, on a phone app, whether you're.

Logan: Whether or not you're using one of the newer, models or one of the older models, all those things are going to differ. What I would say is that if you really want to be cited on the most amount of LLMs, you don't just wanna get cited on Gemini or ChatGPT or Claw or Perplexity or one of those, but you wanna be cited on all of them, it is probably a safe bet for you to try and get a high ranking on Google while also putting out a lot of content that AI search specifically likes to pull from and can easily trust and verify. If you're doing both at the same time, you probably have a lot of overlap, and you'll probably succeed in both. Our clients usually are getting to the top 10 of Google within the first few months on their most valuable keywords and queries.

Logan: That's just what we've seen with pretty much all of our clients now. Within the first three months, they're seeing top 10 average SERP. And, specifically, it's oftentimes happening even quicker than three months when you narrow that down to just their most valuable long tail keywords and queries that we're targeting. You don't need to win for all the broad keywords anymore. You can win for just the very specific ones. When AI cites you, it's not just citing you for that one question. It's citing you for the entire fan out of questions that it's asking around the question that you asked. When it does that, it's likely running five to 30 different queries all at the same time. They're all just different versions of the same question so that it can get a good context of it. If somebody asks a really broad question, it's still going out and finding the really niche, more specific versions of that question that that person could ask.

Watch the full interview on YouTube

How Buyers Use ChatGPT Before Choosing B2B Vendors

Logan: And if you're being cited on Google for one of those, then you might even still be winning a citation or a recommendation whether or not you're getting recommended as the first recommendation or the second, third, fourth, fifth, tenth, whatever. Who knows? But you still might get cited by AI if you go for some of those more niche versions of the same queries, which is again why it's really cool to be doing this in 2026 and beyond as a small business. Moving on to another source for that same that same question of do I need to rank on Google before AI will cite me? Just to go into more of the specifics, ChatGPT's cited URLs overlap with Google's top 10 are really only about a six and a half percent, overlap. If you wanna get a 100% of the citations, then, yeah, you probably need it.

Watch this moment on YouTube (27:55)

Building Trust Scores Across LLMs via Multi-Platform Video

Logan: But ChatGPT is really at a 6.5. Perplexity is at a 43.5. Again, those numbers change all the time, and everyone is changing where they're citing from and what they're trusting, which is why it's important to put out trustworthy non AI content on multiple platforms and in multiple different types of media. And that citation is from UpGrowth. They've done some really great studies as well. I think that I think that answers the question in a pretty robust way.

Comparison at a glance

How much each AI platform depends on Google's top 10 results
PlatformOverlap with Google top 10What it means for your strategy
Google AI Overview76.1% of citations from top 10 pages (Ahrefs, July 2025)Google ranking is close to a prerequisite
PerplexityRoughly 43.5% (UpGrowth)Google rank helps but is not decisive
ChatGPTRoughly 6.5% (UpGrowth)Pulls heavily from Bing and outside sources
Practical takeawayNumbers shift constantlyPursue Google rank and multi-platform trust signals together

Frequently Asked Questions

How do I get cited by ChatGPT and other LLM search engines?

To secure citations across LLMs like ChatGPT, Perplexity, and Gemini, focus on delivering granular, highly specific answers to exact buyer questions on your site while validating those claims on third-party platforms. LLMs cross-reference data across multiple channels to establish trust scores, so publishing non-AI content and leveraging heavily cited platforms like YouTube increases your citation probability across all major engines.

How do buyers use ChatGPT before choosing B2B vendors?

B2B buyers use ChatGPT to evaluate hyper-specific intent queries and compare vendor solutions to exact enterprise requirements. Rather than browsing generic corporate homepages, AI models parse deeply detailed expert content across the web to synthesize clear vendor recommendations for decision-makers.

Why does video content perform better for generative engine optimization?

Video platforms, particularly YouTube, are among the most frequently cited sources for leading LLM algorithms. Video creates a verifiable footprint of original, primary-source expertise that AI engines easily recognize as distinct, trusted content rather than rehashed text.

Do you need a high domain authority to achieve AI visibility?

No, high domain authority is not strictly required for AI visibility. Independent research demonstrates that AI overviews often cite sources positioned on page two of Google search results or pages with no intentional SEO optimizations, provided the content accurately and directly answers the user's specific query.

How does LLM trust verification differ from traditional Google SEO rankings?

Traditional SEO relied heavily on site authority scores and link equity, whereas LLMs evaluate trust through cross-platform verification and factual alignment across the web. While traditional top-10 search rankings correlate strongly with citations, AI engines prioritize content that can be validated in multiple distinct formats and on reliable platforms.

Get your free AI visibility scan

Find out where you stand in ChatGPT, Perplexity, Gemini and Google's AI answers.

Run My Free Scan

← Back to the blog