Real estate agents can use ChatGPT to instantly create hyper-local listing descriptions and community pages that rank in AI searches.
Part 1 of 2 of the AI Masterclass Series. Based on a conversation featuring Ylopo's Head of Growth Aaron 'Kiwi' Franklin, and Realtors-in-Residence Matt Croteau, and Chris Phares from the "AI Masterclass: How Top Ylopo Coaches Use ChatGPT to Work Smarter, Not Harder" Webinar.
Agents are drowning in content creation. Every listing needs a description.
Every community page sits there blank, staring back at you. Somewhere between showing properties and closing deals, you're supposed to find time to write compelling copy that shows up in searches.
The AI noise problem
Everyone in the market today is pushing some AI product to help you "revolutionize" your business. There's a lot of "noise" in the industry today about AI texting and AI voice systems.
Before Kiwi got into ChatGPT tactics, he made a point about this that could save a lot of people money wasted on AI solutions that simply won't work for you.
"If someone is trying to sell you some sort of chat widget or voice widget that is not tied to your IDX, then it's pretty much worthless compared to what Ylopo's AI text and voice are doing for you," Kiwi stated.
The difference is important because many of the stand alone products available today are simply unconnected and therefore provide no value to your business.
These products may do their job but they don't know what your leads are actually doing, nor can they view which properties your leads were viewing or which properties they favorited last night at 11PM.
What makes Ylopo's AI² different
Ylopo's AI² technology (or AI Squared) combines both Ylopo AI Text and Ylopo AI Voice.
The key difference between Ylopo's technology and the rest of the market is that it's connected to your IDX activity and your CRM, allowing the AI to tie all your conversations and activities together.When leads view properties on your Ylopo powered website, the AI sees it. When a lead favorites a property, the AI reacts immediately.
When a lead shows strong intent, the AI escalates the conversation to a live transfer. All conversations flow smoothly and seamlessly into your Follow Up Boss CRM.
"Tying it all together is the hardest part," Kiwi said. "Creating a bot is really simple.
We've taken the hard part and created the heavy prospecting part, which is where you're going to make the most money and benefit the most."
This establishes the foundation for lead engagement but leaves open the question of how people locate your listings and pages in the first place. That process is changing fast, faster than most agents realize.
Strategy 1: Rewrite listing descriptions for large language models
The issue: Older descriptions don't rank any longer
Home search behavior has shifted, though it's easy to miss if you're still optimizing everything for Google.
Many people are now entering their home search into ChatGPT, Perplexity, or whatever large language model they've gotten comfortable using for other questions.
The way people enter their home search into ChatGPT versus Google is completely different.
"They'll type in Google 'Homes for Sale in [specific market].' However, with ChatGPT, they'll say, I'm looking for a four bedroom, five bedroom house that's within walking distance to such and such school and only 10 minutes away from such and such," Kiwi explained.
The queries are extremely specific.
They're based on location. Therefore, if your listing description includes general phrases such as "Close to City Center" or "Close to Shopping", you'll never appear in a search for someone using ChatGPT.
Listing descriptions require actual location data in order for the AI to determine if your listing meets the buyer's criteria. Those standard MLS descriptions most of us have been using?
They're optimized for a search behavior that's becoming less common. Rewriting descriptions with the correct location data takes forever unless, of course, you utilize a tool that knows exactly how to pull that data for you.
The solution: The RE AI listing description creator
Kiwi demonstrated a custom GPT titled "RE AI Listing Description Creator". This is a great example of a tool that allows you to generate listing descriptions that meet the requirements of a buyer who is utilizing large language models to conduct their home search.
Like many tools, the first step is to identify the GPT and access it. To access the RE AI Listing Description Creator in ChatGPT, follow these steps:
Step 1: Locate the custom GPT
Click the "Explore" button located in the lower left corner of the ChatGPT window. Once you are in the "Explore" window, locate the section labeled "GPTs."
Type "listing description creator" in the search bar. The "RE AI Listing Description Creator" GPT should appear in the search results.
Step 2: Provide the tool with your existing description
Once you have accessed the RE AI Listing Description Creator, simply type something similar to "Can you please help me re-write this listing description?" Paste your current MLS listing description into the box provided.
The MLS listing description could have come from Zillow, your broker or even your own MLS listing.
Step 3: Answer the GPT's clarifying questions
The RE AI Listing Description Creator GPT will ask several clarifying questions. Those questions include:
Property Address (critical for location data)
Neighborhood Name
MLS Character Limit
Desired Tone (Conversational, Professional, etc.)
Include Point of Interest Data
Step 4: Receive your optimized listing description
After you complete the clarifying questions, the GPT will create your listing description. Unlike the generic "charming home with plenty of natural light" descriptions found in virtually all listing descriptions, the GPT creates specific descriptions based upon the actual data for the property.
Instead of boring headlines, you get specific ones. Things like "Half-acre lot, no HOA, resort-style, RV parking, minutes to the strip and Allegiant Stadium."
Someone searching for "homes with RV parking near Allegiant Stadium" is going to find that.
The distances are exact. Not "close to schools" but "0.4 miles from Pacific Ridge Elementary School."
It specifically lists top-rated schools with actual names and ratings, coffee shops and restaurants that people actually care about, parks and recreational facilities with distances, major employers and attractions nearby, how far to the airport and transportation hubs.
The RE AI Listing Description Creator pulls hyper-local data that homebuyers are actually searching for. The degree of specificity in the prompts will cause the large language models to believe that the description matches the buyer's inquiry.
Results from the webinar
During the webinar, Kiwi generated a listing description for a Las Vegas property using the RE AI Listing Description Creator. The resulting description included the following information:
That level of detail is what these large language models are looking for when someone asks a specific question. Having the actual information that a buyer might inquire about is vastly superior to keyword stuffing.
"This is a very powerful tool, especially if you guys want to start appearing in large language models," Kiwi stated. "It's free. Take advantage of that."
Most agents will likely ignore this. They'll continue to utilize the same listing description template they've always used.
But the agents who do use this? They'll begin to appear in searches that everyone else is missing.Listing descriptions are not the only area where this occurs.
Strategy 2: Create SEO-optimized community pages in minutes
The issue: Unused community pages represent lost opportunities
Another area that represents lost opportunities for visibility is the unused community pages on virtually all Ylopo websites. There's another set of pages sitting empty on most Ylopo sites, representing the same kind of missed visibility.
If you are a Ylopo user, you have a branded website that is built on Squarespace and features community pages for different neighborhoods.
According to Kiwi, the majority of users have left these community pages completely blank or with just a very small basic piece of text.
Understandably, who has the time to research and write comprehensive community guides for every single neighborhood they serve?
You're supposed to be an expert on Lake Las Vegas and Henderson and Summerlin and whatever other communities are in your market, somehow also writing 2,000-word blog posts about each one?
But having an empty community page represents a missed opportunity. Properly written community pages can attract organic traffic from Google searches, rank in large language model results, and position you as the local market expert.
Additionally, properly written community pages can capture leads who are searching for specific neighborhoods.
The question is not if community pages offer value, but how you can fill them without dedicating 40 hours to researching schools and restaurants and commute times. Fortunately, there is a way to avoid most of the research required to create proper community content.
The solution: The "Five things every homebuyer wants to know" prompt
Kiwi spent considerable time developing what he refers to as a "one-shot prompt." Essentially, a one-shot prompt is a comprehensive ChatGPT prompt that produces professional quality community content in less than a minute.
Location of the one-shot prompt
To obtain the "Five Things Every Homebuyer Wants to Know" prompt, go to the Ylopo Success Community Facebook group.
Click on the "Files" tab and download the "Five Things Every Homebuyer Wants to Know."
How the prompt works
According to Kiwi, there are two essential elements that should be included in every ChatGPT prompt you write.
Although not complicated, the inclusion of these elements can greatly affect the quality of the response produced by the prompt.
Define the Role
"You are a real estate expert and community blogger who has lived in the market for decades. You've seen the area develop, the restaurants come and go, the schools open and close. You write with authority, personality, and deep local knowledge that homebuyers trust."
By defining the role you wish to portray to the AI, the tone and perspective of the response will adjust accordingly.
Without defining the role, the AI will respond with generic content that lacks the authority and local knowledge you would expect from someone who has truly lived in the area.
Identify the Target Audience
"People thinking about purchasing property in the area; whether relocating from other states, or local residents seeking to relocate within a new part of town."
Defining the target audience impacts what type of information ChatGPT identifies and provides. For example, someone relocating from outside of Nevada needs to be provided with different types of information than a local resident moving to a new part of town.
The prompt establishes a basic format for the article based on the types of information that potential buyers seek about a neighborhood:
Lifestyle and Local Vibe (How is it to live there?)
Schools and Education (School rating, etc.)
Market and Home Types (Cost, Architecture)
Community and Accessibility (Transportation, etc.),
Activities and Local Attractions (Food, Entertainment)
These are the questions buyers actually ask when determining if a particular area is appropriate for them.
Results
During the Webinar, Kiwi ran the prompt for Lake Las Vegas.
In approximately 30 seconds, ChatGPT created a title that was not generic, created an article that was between 1800 and 2500 words in length, identified the schools in the area and their respective ratings, identified current market data and pricing for homes in the area, identified businesses and restaurants in the area that buyers may recognize, identified transportation options and realistic commutes to and from employment, and offered insights that seemed to originate from someone familiar with the area.
As Kiwi stated,
"It is almost as if you hired someone to do the research and then write the article. I see a lot of protective mechanisms to prevent this from appearing as a generic AI article or being flagged as such by Google."
That last part matters. Google's gotten pretty good at detecting thin AI content and burying it.
Providing ChatGPT with enough structure and context via a well-structured prompt produces an article that appears to have been written by someone knowledgeable about the subject matter rather than generic AI drivel. As a result, you can utilize the content without hours of editing.
Implementation: Simply copy & paste onto your Ylopo website
After ChatGPT generates the community content, simply copy the article content, log into your Ylopo-branded website, find the community pages, and paste the article content directly into the page.
However, here is a very important aspect of this process:
Make certain your IDX Feed is displayed among the sections of the article. The community content draws visitors to your website.
However, your IDX Feed is what allows you to capture leads from those visitors.
"There is no purpose to have an article without the IDX feed included in it because the IDX feed is how you monetize the traffic and views you receive," Kiwi explained.
If you create the most appealing community guide possible, yet fail to provide a mechanism for individuals to view available properties on the page, you are essentially sending them back to Zillow to continue their search. Once they are on Zillow, they are likely communicating with whoever has listed their property on Zillow.
An effective community page can assist you. Thirty effective community pages will begin to appear as a significant competitive advantage.
Repeat process for each community
A key benefit of the system described above is its scalability. You are not required to spend hours researching and writing for each neighborhood.
You're not spending hours researching and writing for each neighborhood. You run the same prompt for every area you cover: Same prompt structure for each neighborhood.
Completely different content for each area due to differences in the data. Build your entire website in a weekend.
Establish yourself as the hyperlocal expert in multiple areas.
"You can simply repeat this for every single community," Kiwi said. "ChatGPT will produce entirely different content for each neighborhood as the markets are clearly different."
And it will. ChatGPT isn't going to give you the same article about Lake Las Vegas and Summerlin.
The schools are different, the prices are different, and the vibe is different in each of the two areas.
Therefore, when you provide it with the name of the neighborhood, it performs the research related to that neighborhood and creates content that accurately represents the distinct features of that neighborhood.
This is likely the least exciting feature of this strategy, however, it is possibly the most important. You can actually finish it.
You will not create custom content for thirty different neighborhoods in your free time. However, you can run this prompt 30 times during an afternoon.
At that point, the content will begin generating revenue for you in many ways beyond just the content itself.
The Ylopo advantage: AI generates revenue for you
Why these ChatGPT-based strategies work particularly well for Ylopo users is how the content produced integrates with the existing automation capabilities of the Ylopo platform. Having better content matters, but how the content ties together with everything else matters more.
The content flywheel:
ChatGPT creates optimized listing descriptions and community pages.
Those pages rank in Google and large language models.
Organic traffic flows to your Ylopo website.
Viewers browse the IDX feed of properties on your website.
Ylopo AI Text immediately engages them with chat.
Ylopo AI Voice follows up with calls.
Leads with high intent trigger priority notifications.
Conversations from all sources synchronize to Follow Up Boss.
You intervene when leads are actually ready to engage.
This is what Kiwi referred to when he mentioned that you should not purchase random AI widgets. A standalone chatbot cannot perform this function.
It does not understand what properties someone is viewing, therefore cannot escalate based upon intent signals, and just sits there chatting.
When your content, IDX, AI engagement, and CRM are all integrated, each component enhances the effectiveness of the others. The listing description attracts viewers to your website, the community page maintains their interest, the IDX feed captures their interest in specific properties, the AI follows up automatically, and you only engage when the lead demonstrates high levels of intent.
So what does it take to build that system?
Summary and key takeaways from Part 1
For listing descriptions:
Utilize the RE AI Listing Description Generator Custom GPT.
Provide exact distances to amenities (not just claim proximity).
Include the names of specific schools along with their ratings.
Add hyper-local details that reflect how buyers are searching.
Focus on specific keywords that large language models prefer.
For community pages:
Obtain the "5 Things" prompt from the Ylopo Success Community.
Define the role and target audience in your prompts (the output quality will be impacted by the selection).
Allow ChatGPT to perform the research while you spend time doing something else.
Include your IDX feed throughout the content to convert traffic to leads.
Scale the process by running the same prompts for each of the communities you serve.
Kiwi emphasized that you should not let every AI tool vendor distract you.
"Do not be distracted by every person selling a bot. If it is not connecting to all your IDX and driving into your Follow Up Boss, it is probably not going to do you much good, and it is probably going to cause you additional work."
This is likely true for most of the AI tools vendors pitching to agents today. They seem great until you discover that they are creating another system that you have to review, and another group of conversations that are not connected to your actual lead funnel.
Listing descriptions and community pages are merely the beginning though.
Upcoming topics in Part 2
In Part 2 of this series, Matt and Chris will discuss more advanced topics. They will explain how to create automatic "Deal of the Week" email campaigns utilizing Zillow integration, set up scheduled tasks that operate independently of you, create customized GPTs for addressing lead objections, use Follow Up Boss's built-in AI for follow-up communications tailored to individual leads, analyze completed sales to determine recurring trends to emulate, and use the RELIC framework to enhance your prompts for writing better content.
Matt and Chris use these methods on a regular basis to close more deals with less manual effort.
All of the methods are based on the Ylopo platform since that is where the integration actually adds value.
Ylopo Success Community on Facebook contains all the prompts, templates, and resources from the previous webinar.
Matt and Chris conduct daily office hours at 9:00 AM Pacific if you require assistance with implementing any of the techniques discussed in this webinar.