Why AI Is Recommending Some Builders and Ignoring Others: The Case Study-186

Why AI Is Recommending Some Builders and Ignoring Others: The Case Study

A regional home builder in a competitive Florida market was being cited by ChatGPT 16% of the time when buyers asked real questions like "who are the best new home builders in this area?" A few months later, after a focused initiative with Anewgo, that number was 72%. Same builder. Same market. Same competitors.

In this solo episode, Anya Chrisanthon - CCO at Anewgo and host of the Anewgo of New Home Sales podcast - breaks down exactly what was measured, what changed, and what it means for builders and marketers who are paying attention to where buyer search behavior is heading.

What Anewgo Tracked Anewgo ran 50 real buyer-intent questions through ChatGPT and Gemini and measured three things: mentions (how often the builder was named at all), citations (how often their website was used as a trusted source), and top recommendations (how often they appeared in the top spots). Citations matter most. A mention means AI knows you exist. A citation means AI trusts you enough to point a buyer directly to your content. Those are not the same thing.

The Results On ChatGPT, the citation rate jumped from 16% to 72% - a 56 percentage point increase. The number one ranking rate went from 7% to 36%. Top three ranking rate went from 16% to 59%. On Gemini, citation rate went from 25% to 63%. And during the same period, average website engagement time nearly doubled - from 51 seconds to just under two minutes - because the content wasn't just AI-friendly. It was genuinely useful to real buyers.

The Philosophy The goal was never to trick AI or game the system. Anyone telling you they can game AI recommendation engines is not someone you should trust with your marketing budget. The goal was to make this builder genuinely easier for AI to understand, verify, and confidently recommend - through content that answers real buyer questions and clarity about what the builder actually offers and who they build for.

Why Most Builder Websites Are Invisible to AI Right Now Most builder websites were built for two audiences: human visitors and Google. AI reads your website differently than either. It is trying to build a mental model of your business - who you are, who you build for, where you build, what makes you different. Vague taglines, generic community descriptions, and inconsistent information across your digital presence create uncertainty for AI. And uncertain AI does not recommend you. It moves on.

Three Things to Do This Week:

(1) Search for yourself the way a buyer would - open ChatGPT, Perplexity, or Gemini and ask for a builder like you in your market with specific criteria. See what comes up. 

(2) Read your own website like a stranger - is your brand story specific and consistent? Could AI accurately summarize your communities and differentiators to a buyer? 

(3) Get an AEO assessment from Anewgo - email your website URL to beth@anewgo.com with subject line "AEO Assessment." No obligation. Beth will reach out with the report and a time to walk through the results together.

Next episode: a deeper dive into AEO itself - what it is, what good AEO looks like in practice, and the broader framework for thinking about this for your business.

About Anewgo Anewgo is an all-in-one new home sales and marketing platform. We equip builders with AI-ready homebuilder websites, interactive design tools, floorplans, sitemaps, AI Sales Assistants, and data analytics to create personalized buyer journeys. Learn more at anewgo.com or listen t

Transcript

0:00

Hello, and welcome to another episode of Anewgo of New Home Sales. I'm your host, Anya Chrisanthon, and today it's a solo podcast. I'm actually sitting outside on my back porch, so if you hear some birds chirping, that's why. I wanted to share a case study that Anewgo recently conducted with one of our clients. But before we jump into it, I wanted to share a number with you today that genuinely surprised me, and I don't get surprised by data very often anymore. A home builder we work with, a regional builder in a competitive Florida market, was being cited by ChatGPT in response to real buyer questions 16% of the time at the beginning of this year. Meaning when buyers asked ChatGPT questions like, you know, "Who is the best new home builder in this area?" Or, "Which builder gives you the most value?" This builder was showing up as a source roughly one out of every six times. Not great, but not unusual. Most builders are in that range or lower. By the time we finished a focused initiative with them a few months later, that number was 72%. Same builder, same market, same competitors, from 16 to 72%. That is what this episode is about. Not theory, not trend to watch. A real builder, real numbers, and what actually changed to make this happen. So let me, back up for a second and make sure you're all starting this from the same place, because I know that not everyone listening has been, deep in this conversation yet. So buyers are changing how they search for homes, and I don't just mean they're spending more time online. We've known that for years. What I mean is the actual tools they're using to search is changing. More and more buyers are opening ChatGPT, Gemini, or Perplexity and just asking. Not typing "new home builders near me" into Google and scrolling through results, Actually asking the way you'd ask knowledgeable friend, "Who are the best new home builders in my market? What's the difference between a production builder or a semi-custom builder? Which builder actually gives you the most for your money?" And AI answers confidently with specific names, specific reasons, sometimes specific communities. We actually saw this happen in real time with one of our other clients, Garman Homes, here in Raleigh. A buyer searched for a boutique home builder with four bedrooms, a home office, and a kid-friendly communities. Garman came up as the number one recommendation across multiple AI systems, not because they paid for it, because their digital presence was built in a way that AI could actually understand and trust. That is the opportunity, and that is what most builders are not optimizing for yet. Okay, so back to our Florida builder. When we started working with them on this, we needed to understand exactly where they stood, so we did something pretty simple, but really illuminating. We took 50 real buyer questions, the kinds of questions actual home buyers type into AI tools when they're searching their, options. Things like discovering which builders are worth considering, comparing communities, understanding what makes one builder different from another, figuring out production versus semi-custom. Real questions from real buyer journeys. And we ran those questions through ChatGPT and Gemini and tracked three things. First, mentions. How often was builder named at all in the AI response? Second, citations. How often was their website actually used as a source, meaning that AI actually trusted their content enough to reference it directly? And third, top recommendations. So how often did they actually appear in the top spots when AI was answering questions about who to consider? Citations are the most important of those three, by the way. A mention means AI knows you exist. A citation means AI trusts you enough to point a buyer directly to your content. That is a completely different level of credibility here. At the start, their Chat GPT citation rate was 16%. The number one ranking rate, meaning AI named them first when recommending builders, was 7%. Their top three rankings was 16%. Those numbers tell you that AI knew this builder existed, it just didn't trust them enough or understand them well enough to recommend them with confidence. Now, here's the part where I'm going to be intentionally a little general, because the specific approach we use with clients is something we've invested a lot in developing, and I'm not going to walk through the whole playbook on air. But I can tell you the philosophy, because the philosophy is what matters for your business here. The goal was never to trick AI or game the system. That is not how this works. And honestly, anyone telling you that they can game AI recommendation engines is not someone you should trust with your marketing budget. The goal was to make this builder genuinely easier for AI to understand, verify, and to confidently recommend. Think about it from AI's perspective. AI is like world's most well-read research assistant. It has absorbed an enormous amount of information about businesses and markets and brands. So when a buyer asks it a question, it goes to work synthesizing everything it knows and making a judgment call. If what it finds about your business is thin, vague, or inconsistent, like generic taglines, community descriptions that could apply to any other builder anywhere, or marketing language that sounds great to humans but gives AI nothing specific to work with, it hedges Or it skips you entirely. What we did, fell into two categories. The first was content. Not blog posts stuffed with keywords. Content that actually answers the real questions buyers ask when they're building their shortlists. Practical, specific, genuinely helpful content written the way a knowledgeable friend would explain the market to someone making one of the biggest purchases of their life. And the second was clarity. Making the builder's core facts crystal clear across the digital presence. Where they build, what they specialize in, their communities, what actually sets them apart, not in, adjectives, but in actual specifics. Making that information easy for both human visitors and AI systems to find, read, and trust. That's the plain language version. Essentially, content that serves buyers and clarity that AI can work with Here's what happened. On ChatGPT, the citation rate went from 16% to 72%. That is a 56 percentage point jump. Their number one ranking rate went from seven to 36%. Their top three ranking rate went from 16 to 59%. On Gemini, citation rate went from 25 to 63%. Number one ranking went from nine to 23%. ChatGPT showed the most dramatic improvement. Gemini improved too, though the gains were smaller because they already had decent baseline visibility there. And here's the thing that I think is actually the most important part in this story. The website traffic improved too. During the same period, engaged sessions went up and average engagement time nearly doubled from 51 seconds to just under two minutes. Why does this matter? Because it tells you that the content we created wasn't just AI friendly. It was genuinely useful to actual human visitors. Real buyers were spending more time on the site because the content was better, more specific, more helpful, more honest about what this builder is and who they're right for. This is not a coincidence. Good AEO and good content marketing are not in conflict. When you do this right, they reinforce each other. So let me bring this out of the case study and make it relevant for where you are right now. Most builder websites were built for two audiences, human visitors and Google, and they did a reasonable job of that. But AI reads your website differently than either of those audiences. AI is trying to build a mental model of your business, who you are, who you build for, where you build, what makes you different, and a lot of builder websites make that really hard Think about how many builder websites lead with something like, "Building your dream home since 1987", or, "Quality craftsmanship you can trust". I'm not picking on anyone here. These phrases have been on builder websites forever, but they mean almost nothing to AI trying to figure out whether you're the right answer for a specific buyer with specific needs. So vague taglines, generic community descriptions, the same three adjectives like quality, craftsmanship, trust, that literally every builder in the country uses. As well as inconsistent information between your website and other digital presence. All of that creates uncertainty for AI, and uncertain AI does not recommend you. It moves on. The bar has moved, and builder websites haven't moved with it yet, which honestly is good news for the builders in this audience because the window to get ahead of this is still open. Okay, so here's what I actually want you to do with this. Number one, go search yourself the way a buyer would. Right now, after this episode, open ChatGPT or Perplexity or Gemini and ask for a builder like you in your market, something like, "What's the best boutique home builder in your city for families with kids?" Or whatever fits your buyer profile. See what comes up. If it's not you, or if you're buried, or if AI mentions you but doesn't cite you as a source, that's your baseline. That's where you're starting from. Number two, read your own website like a stranger would, not like someone who already knows your story. Like an AI that only knows what your website tells it. Is your brand story specific and consistent? Are your communities described in enough detail that AI could accurately summarize them to a buyer? Do you actually explain what makes you different, or do you just assess it with adjectives and hope buyers fill in the blanks? Number three, get an AEO assessment from us. This is something we do at Anewgo. We look at how your site is currently performing with AI systems, where the gaps are, and what it could take to move the needle. If you want us to take a look, email your website URL to beth@anewgo.com, the subject line AEO assessment. That's beth@anewgo.com, no obligation, just a real look at where you stand. Beth will reach out to you with the report as well as a scheduled time to go over the, results with you. I will link that in the show notes to make this easier for you guys to find. So here's my take. A 16% citation rate jumping to 72% in a matter of months is not magical. It's not a loophole. It's the result of making builder's digital presence genuinely clearer, more specific, and more trustworthy for AI and for buyers. And here's what keeps me up at night a little bit honestly, the builders with the most resources, the large production builders with data teams and marketing budgets are paying attention to this. Once they move on AEO in a serious way, they're going to move fast. And because AI compounds, meaning the more credible and consistent your presence is over time, the more confidently AI recommends you, the gap between builders who move now and builders who wait is going to widen. The good news is that right now, in most markets, most builders have not touched this. A boutique regional builder can absolutely own their AI recommendations before a national builder figures out how to target that market. That is the window. It is open, but it will not stay open forever. We have also written a full blog post on this case study going out, on Thursday, so check out anewgo.com, our blog section,, for that. And in the next episode, we're going to go deeper on AEO itself, what it is, what good AEO looks like in practice, and a broader framework for thinking about this for your business. Thanks for listening, and I'll see you guys next week. Bye.