Latest / Connor with Honor: AI, Real Estate and Real Life / 159: Why ChatGPT Won't Say Your Business Name (Answer Engine Optimization, Explained)

Transcript

159: Why ChatGPT Won't Say Your Business Name (Answer Engine Optimization, Explained) · Connor with Honor: AI, Real Estate and Real Life

• 0:00
  SPEAKER_00: I started programming in nineteen eighty-three before most even people called it that.
• 0:04
  SPEAKER_00: It was ninety-nine home computer copying basic out of magazines just to make something happen on the screen.
• 0:10
  SPEAKER_00: Saved on a cassette tape, of course, black and white TV, uh audio video cables to plug it in.
• 0:16
  SPEAKER_00: It was it was something.
• 0:17
  SPEAKER_00: And become you know, games back then weren't really much of a thing.
• 0:21
  SPEAKER_00: That machine was a Timex Sinclair 1000.
• 0:24
  SPEAKER_00: It was the American version of a British computer called the Sinclair ZX81.
• 0:29
  SPEAKER_00: Launched in July of 1982 as the first fully assembled computer ever sold for under a hundred bucks.
• 0:36
  SPEAKER_00: I was about 14.
• 0:38
  SPEAKER_00: I need to correct myself here on myself because correcting the record on people is the whole business I'm in.
• 0:45
  SPEAKER_00: So for years I told that story, and I called that machine at Texas Instruments TI-994A instead, and it was not.
• 0:52
  SPEAKER_00: The TI-994A was real, but it was my dad's sitting in the same house, and I wasn't allowed to touch it.
• 0:58
  SPEAKER_00: My computer cost 100 bucks, his quite a bit more.
• 1:02
  SPEAKER_00: The Sinclair was the one that was actually mine.
• 1:04
  SPEAKER_00: Two real machines under one roof, one I can use and one I could only look at.
• 1:09
  SPEAKER_00: And 40 years later, my own memory merged them into one.
• 1:12
  SPEAKER_00: And I can misremember my own first computer.
• 1:15
  SPEAKER_00: I have real sympathy for why a business address gets misremembered across 70 different websites as well.
• 1:22
  SPEAKER_00: So nobody does it on purpose.
• 1:24
  SPEAKER_00: It just happens quietly, the same way two machines in one house turned into a machine in my own head without me ever noticing the swap.
• 1:33
  SPEAKER_00: So there it was, 30-something years to artificial intelligence now being everywhere.
• 1:39
  SPEAKER_00: But the part in the middle matters most than some people think.
• 1:43
  SPEAKER_00: So stay with me for this one.
• 1:45
  SPEAKER_00: Around 2005 and 2006, I caught on to search engine optimization right at the point where it split into two camps.
• 1:52
  SPEAKER_00: Black Hat was the cowboy villain, like in the movies, the strategies that gamed Ask Jeeves and early Google and AOL search by loading a page with the same keyword hundreds of times or thousands of times until it climbed the rankings.
• 2:06
  SPEAKER_00: White Hat played it straight.
• 2:08
  SPEAKER_00: Google spent years closing that gap and mostly one.
• 2:11
  SPEAKER_00: But what it handed you instead was ten blue links you still had to sort through yourself.
• 2:15
  SPEAKER_00: Some were good, some were useless.
• 2:18
  SPEAKER_00: That's not how it works anymore.
• 2:20
  SPEAKER_00: AI answer engines for the large language models, they don't hand you ten links to sort through.
• 2:25
  SPEAKER_00: They pick a name.
• 2:26
  SPEAKER_00: Or two or three, and they say it out loud with confidence.
• 2:30
  SPEAKER_00: In one sense, that's a fairer playground than the old one because you cannot keyword stuff your way to a large language model's answer the way you could stuff a page back in 2005.
• 2:41
  SPEAKER_00: In another sense, it's a harsher one because the machine is not curating a list for you to sort through anymore.
• 2:47
  SPEAKER_00: It's vouching for a name.
• 2:49
  SPEAKER_00: Old SEO still has bearing here.
• 2:52
  SPEAKER_00: It just moved jobs from tricking a crawler to proving a fact that your name, your address, and your phone number are the same everywhere a machine that vouches for people goes looking.
• 3:02
  SPEAKER_00: This part is not a guess.
• 3:04
  SPEAKER_00: ChatGPT struck a formal deal with Yelp this year to read reviews and listings straight out of it.
• 3:10
  SPEAKER_00: Gemini reads Google's own map data directly across roughly 350,000 business locations studied in 2026.
• 3:18
  SPEAKER_00: Profile accuracy runs about 68% on the AI engines like ChatGPT and Perplexity and 100% on Gemini because Gemini is reading dead straight off Google's own map.
• 3:29
  SPEAKER_00: Businesses whose information matches everywhere get recommended roughly 10 times more than businesses whose information disagrees with itself.
• 3:37
  SPEAKER_00: And only 1.2% of businesses nationwide ever make it into one of these AI answers at all.
• 3:43
  SPEAKER_00: Out of every hundred businesses competing for that answer, roughly one gets picked.
• 3:48
  SPEAKER_00: And that's winner take most, not winner take some, and it's happening quietly every day, whether you ever notice it or not.
• 3:54
  SPEAKER_00: So people ask me sometimes if this matters for one AI, and it actually does it.
• 3:59
  SPEAKER_00: ChatGPT, Gemini Claude, Microsoft Copilot, Grok, Meta AI, every one of them gets asked local questions every single day, and every one of them makes the naming decision the same basic way.
• 4:12
  SPEAKER_00: It goes out and finds live current facts about you instead of trusting what it was trained on.
• 4:17
  SPEAKER_00: A large language model, GPT Claude, whichever one, is the part that writes the sentence.
• 4:23
  SPEAKER_00: It's very good at writing sentences, but its own knowledge was frozen at training time.
• 4:27
  SPEAKER_00: It doesn't know your phone number changed last month.
• 4:30
  SPEAKER_00: It doesn't know you moved offices in most cases.
• 4:33
  SPEAKER_00: It generally does not know your business exists at all.
• 4:36
  SPEAKER_00: So before it writes a word, something else has to go out and find who you actually are, right now, today.
• 4:42
  SPEAKER_00: And that's something is called answer engine optimization.
• 4:46
  SPEAKER_00: And it's not the AI, it's what feeds the AI.
• 4:49
  SPEAKER_00: So this is exactly why perplexity is worth naming specifically, and it's a sharper question than most people think to ask.
• 4:56
  SPEAKER_00: Perplexity is not just wrapping somebody else's model and repeating it back to you.
• 5:01
  SPEAKER_00: It runs its own crawler, its own index of hundreds of billions of pages, and its own ranking pipeline.
• 5:07
  SPEAKER_00: And only after all that narrows the field down does it hand the willing job to a model.
• 5:12
  SPEAKER_00: Sometimes its own, sometimes ChatGPT, sometimes Claude.
• 5:16
  SPEAKER_00: Depending on the mode you're in, the model that writes the sentence changes.
• 5:20
  SPEAKER_00: So Perplexity Zone Retrieval Ranking does not, which means getting found by Perplexity has almost nothing to do with which AI writes the sentence and everything to do with whether Perplexity Zone Crawler found one clean, confident record for your business.
• 5:36
  SPEAKER_00: And that's the whole business model right there.
• 5:38
  SPEAKER_00: Answer engine optimization is making sure that whichever retrieval system a machine leans on, Yelp and Foursquare for ChatGPT, Google's own map for Gemini, Bing Places for Copilot, Perplexity's own index for perplexity, finds one confident answer instead of a handful of conflicting ones.
• 5:57
  SPEAKER_00: And this is not something you have years to get ahead of.
• 6:00
  SPEAKER_00: Consumers using AI instead of a search bar to find a local business went from 6% to 45% in the last 12 months alone.
• 6:08
  SPEAKER_00: According to Bright Locals owned 2026 research, it's not a trend anymore.
• 6:13
  SPEAKER_00: This is most of your next customers' actual behavior already happening.
• 6:17
  SPEAKER_00: So whether your listings are ready for it or not.
• 6:20
  SPEAKER_00: Now here's where this hits real estate agents harder than almost anyone else.
• 6:24
  SPEAKER_00: A roofer or barbershop changes company once a decade, if that.
• 6:30
  SPEAKER_00: Real estate agents change brokerages every couple years of chasing whatever promises can level them up next.
• 6:37
  SPEAKER_00: I've watched it happen around me since 1998, and I'm most I'm mostly one of the ones that stayed put.
• 6:43
  SPEAKER_00: A little bit of moving, but not much.
• 6:45
  SPEAKER_00: Most agents don't.
• 6:46
  SPEAKER_00: Every move can change that phone number tied to your name, the office address, how your license displays.
• 6:53
  SPEAKER_00: Directories that had you correctly filed, under one brokerage, now disagree with the ones that have not caught up yet.
• 6:59
  SPEAKER_00: And a machine reading fragments about you instead of one confident record about you treats that exactly the same way it treats a plumber with three different suite numbers listed in three different places.
• 7:11
  SPEAKER_00: Not sure enough to say your name out loud.
• 7:14
  SPEAKER_00: I found out how bad my own fragments were the same way.
• 7:17
  SPEAKER_00: I want you to find out about yours.
• 7:19
  SPEAKER_00: I ran the scan that I'm going to talk about in a moment on myself first.
• 7:22
  SPEAKER_00: My own listing showed a brokerage I had not worked at in years.
• 7:26
  SPEAKER_00: A wrong address attached to the name on one of the biggest review sites there is, and that same site just signed a deal with Chat GPT.
• 7:34
  SPEAKER_00: My own name spelled three different ways across major platforms.
• 7:38
  SPEAKER_00: Like there were three different guys running around Santa Clarita, all claiming to be me.
• 7:43
  SPEAKER_00: The computer story and the listing story, they are the same story.
• 7:47
  SPEAKER_00: One is embarrassing and harmless, two real machines in one house turned into one machine in my own head, and I never caught it for 40 years.
• 7:55
  SPEAKER_00: The other one costs real money every single day it goes uncorrected, because the machine deciding whether to recommend me doesn't have 40 years of patience.
• 8:04
  SPEAKER_00: It makes the call in a fraction of a second and then moves on.
• 8:08
  SPEAKER_00: I couldn't fix 70 directory listings by hand for every business in Santa Clarita, every real estate agent, every plumber, every roofer, every pool inspector, everybody.
• 8:17
  SPEAKER_00: If I tried to do it at scale from day one, and I'm not trying to, the founding block, I capped it at 50 businesses on purpose because I run that audit and fix myself personally, not my team, not anybody else, not some offshore O desk or anything else like that.
• 8:33
  SPEAKER_00: I hand correct the core listings on Google, Yelp, and Bing during the onboarding sequence.
• 8:38
  SPEAKER_00: I push one clean record out everywhere else and send a monthly report on what ChatGPT Perplexity and Gemini Claude actually say about you before and after.
• 8:48
  SPEAKER_00: So you're never wondering what you're paying for.
• 8:51
  SPEAKER_00: That free scan costs you nothing.
• 8:53
  SPEAKER_00: And that report, well, it's yours to keep either way.
• 8:57
  SPEAKER_00: Whether you become a client or not, if it comes back clean, I'll hand you that clean report, shake your hand, and that is a good day, even if I don't get paid for it.
• 9:07
  SPEAKER_00: But that founding 50, the usual$3.99 setup is waived completely, and it runs$79 a month after that.
• 9:13
  SPEAKER_00: No contract, cancel anytime you want.
• 9:16
  SPEAKER_00: That's the whole offer.
• 9:17
  SPEAKER_00: No hidden number in between, no fine print doing something like the spoken price did already, did not already tell you.
• 9:25
  SPEAKER_00: And I do it this way because I'm a real person running this in the same valley I still real estate in, Santa Clarita, not a call center somewhere.
• 9:33
  SPEAKER_00: I will never see you again.
• 9:35
  SPEAKER_00: Well, I will.
• 9:36
  SPEAKER_00: If I don't do your setup or get it wrong, I can't hide from you on that.
• 9:42
  SPEAKER_00: I'm gonna run into you eventually at Costco one of these days, and I would like a whole lot rather you to be glad with me than have you save up for a tomato and throw it at me in the in the store for that occasion.
• 9:54
  SPEAKER_00: This is the actual quality control here, not a policy on a page, just a guy who has to live in this valley with whatever he did to your listing.
• 10:02
  SPEAKER_00: If any of this made a light come on, if you have wondered for a while why the phone doesn't ring the way it used to, go look for yourself before you take my word for any of it.
• 10:12
  SPEAKER_00: Machinefound.com or get machinefound.com.
• 10:16
  SPEAKER_00: Either one is mine, either one gets you there.
• 10:18
  SPEAKER_00: Run that free scan, read exactly what it says about you, and if you want to walk it through after that, the number's right there on the site.
• 10:25
  SPEAKER_00: 661 476 2217.
• 10:29
  SPEAKER_00: And it's actually me who picks up.
• 10:31
  SPEAKER_00: Be well.
• 10:32
  SPEAKER_00: I'll see you in the next one.
• 10:33
  SPEAKER_00: Thanks for watching.