David Hunter, CEO of Local Falcon, has been tracking local rankings for 125,000 users since 2018. Now he’s tracking how ChatGPT, Claude and Google’s AI decide which local businesses to recommend, and the rules are different.
While David Hunter was talking on our show, I opened Claude and asked for the best pizza and the best breakfast on Sanibel, where I live. It did a pretty good job. Got one of the pizza places right, got both breakfast spots right. It also listed a breakfast place that doesn’t exist anymore.
That little test sums up where AI local search is today. It’s surprisingly useful, it’s only as good as its data, and most businesses have no idea how it decides who shows up.
David Hunter has been in digital marketing and SEO for 17 years. He runs an agency in Erie, Pennsylvania, and he built Local Falcon after realizing clients were confusing their Google Business Profile rankings with organic search. Local Falcon puts a grid over your area and shows where you rank as you move around town. It launched in December 2018 and just crossed 125,000 users.
Does AI search actually use local data?
Not well, at least not until recently. Google has a massive knowledge graph with a profile for almost every physical business on earth. OpenAI, Anthropic and the rest don’t. So for a long time, ask ChatGPT for a pizza place and it would pull from Yelp listings or random Trip Advisor lists.
That’s changed a lot in the last year. Google’s AI Overviews, AI Mode and Gemini pull from the knowledge graph and the map pack, so those results are stronger. And the others are catching up.
It also feels like the early days of Google again, spam tactics and all.
You could just put a bunch of keywords on your homepage as hidden white text for a prompt, and it works. You’re actually showing up in AI results.
— David Hunter, CEO, Local Falcon
That made me laugh, and also a little nervous. David Hunter also pointed out that somewhere around half or more of Google searches carry local intent. So of course people are asking AI where to eat when they’re in a new city.
What ranking factors matter most for AI local search?
First, a warning from David Hunter that I really appreciated.
Anyone that is out there telling you, oh, it’s these pillars, this is definitely how you do it, is a snake oil salesman. Nobody actually really knows what the exact steps are.
— David Hunter, CEO, Local Falcon
That said, Local Falcon has a lot of data. They ran a study across roughly 60,000 businesses, plotting local intent queries on a 7 by 7 grid of 49 points, to see whether AI results were truly localized. The answer: not really, at least outside of Google.
The proximity is less important. It’s the prominence and the relevance that do matter.
— David Hunter, CEO, Local Falcon
In traditional local search, proximity was the number one factor by a mile. In AI search it’s authority. Content that proves your expertise. Before and after photos. Testimonials. Mentions on places like Reddit and Quora, which show up constantly in AI citations. David Hunter even found a random TikTok video being cited in a result that mentioned Local Falcon.
And here’s where it gets interesting. Classic backlink thinking doesn’t fully carry over.
If you got a link from CNN.com about your business, that’s gold on Google. That doesn’t necessarily matter when it comes to AI search.
— David Hunter, CEO, Local Falcon
AI cares more about the context of the question. That’s actually good news for smaller businesses. There’s a lot of open territory if you’re willing to test it.
How should you write content for AI search?
Two things stood out from David Hunter’s advice.
First, test with natural language. Local Falcon’s top keyword is “local rank tracking,” but nobody types that into ChatGPT. Search the way real people talk and see where you show up.
Second, accept that clicks are going away.
I’ve got tons of impressions now but I’ve got no clicks, and the reason for that is because it’s answering the question right there.
— David Hunter, CEO, Local Falcon
So write in passages, not walls of text. AI retrieves content in chunks. FAQs help a lot. So do testimonials, because they’re passage level text the AI can pick up. Use schema markup so crawlers understand what each section is about.
Are AI crawlers actually able to reach your website?
This was the most practical tip of the whole episode, and it’s a little scary.
David Hunter’s agency used a firewall service that blocked unknown bots. Pretty standard. Except it turned out to be blocking OpenAI and Anthropic completely.
Our firewall provider was actually blocking all of OpenAI and all of Anthropic from even coming and getting the content on the website.
— David Hunter, CEO, Local Falcon
What makes it worse is that they had checked. They emulated the OpenAI crawler from their office and everything looked fine. But their office IP was whitelisted, so the test passed while the real crawlers were blocked. It stayed that way for about a year and a half.
And the AI might still mention you, because it’s falling back on a Bing or Google search. So you’d never know. That’s a big deal. Go check your firewall settings today.
How can local businesses use automation without sounding like robots?
Google rewards businesses that respond to reviews, quickly, whether they’re five stars or one star. For multi-location brands that’s a lot of work, so automation makes sense. But David Hunter was clear: no generic “thanks for your review” replies. Use AI to actually read the review and respond to the details. If someone had a bad experience with their server on Monday, say so.
Jessie asked the question every CMO is thinking: what if the bot gets it wrong? David Hunter’s answer was simple guardrails.
Don’t allow a response on one or two stars. Don’t do it. Just make sure those get flagged and you have to go do that yourself.
— David Hunter, CEO, Local Falcon
Local Falcon also has a reviews analysis tool that reads every review for you and your competitors and builds a strengths and weaknesses report. David Hunter suggested using that as part of your guardrails for automated replies. One more tip: your hours affect visibility, so scan your rankings when you’re open and when you’re closed and compare.
And as Jessie said on the show, garbage in, garbage out. If the first name in your merge field is wrong, your “personalized” reply is a disaster waiting to happen.
The bigger picture
The tech stack conversation told its own story. David Hunter uses ChatGPT, Gemini and Claude every day, lives in Slack and Google Workspace, and replaced HubSpot with Intercom and SendGrid. He went from over $5,000 a month to roughly $500. Product led growth, no big sales team.
What I took away is that AI search rewards the same thing good marketing always has: real authority, real customer proof, and clean data. The difference is the rules are being rewritten in real time and nobody has the full playbook yet. The businesses that test, check their crawlers and keep their data clean are going to own this space while everyone else is still guessing.
Connect with David Hunter and Local Falcon: Visit localfalcon.com or find David Hunter on LinkedIn.
Want your personalization and automation running on accurate identity data? Visit bdex.com and click “Talk to an Expert.”
Watch the full episode: Unlocking Local Intelligence: AI Strategies for Hyperlocal Marketing
This article was adapted from an episode of Deconstructing Data, BDEX’s weekly podcast on data-driven marketing. Tune in live every Thursday at 4:15 PM Eastern on LinkedIn.
About BDEX: For companies that need clean identity data to power their products, BDEX offers unmatched quality and execution. Visit bdex.com and click “Talk to an Expert” to get started.
