In real estate, time is money and if you’ve been investing for a while, I’m sure you’ve felt the sting of a missed opportunity. One facet of the due diligence phase that can be time consuming is finding comparable sales to determine your subject property’s after repair value or even estimated market value if it’s not a fixer upper. This raises the question: Can a ChatGPT real estate plugin replace manual property record due diligence?
In our ever evolving technology landscape, AI tools, most notably, OpenAi’s ChatGPT, have become extremely popular and the real estate industry is not exempt from these advancements in technology.
Today, we’ll put the TopHap ChatGPT real estate plugin to the test to see if it can help streamline our due diligence process and give us a ballpark estimate of a property’s market value to help us make offers more efficiently and quickly.
Note: at this time, in order to use ChatGPT plugins you must be a ChatGPT Plus user. Additionally, we wanted to compare this tool to the Zillow Plugin but it was unavailable for testing.
Enabling TopHap ChatGPT Real Estate Plugin
Before we dive into how to use TopHap, let’s run through how to enable ChatGPT plugins within your dashboard. If you already know how to install plugins, you can skip this part by clicking here.
ChatGPT Settings Menu
First, you’ll navigate to to the pop-up menu on the lower left hand corner of your ChatGPT interface and select the “Settings Option”:
Enable Plugins
Next, you’ll click on “Beta Features”. Then, you’ll click on “enable plugins”:
Start a New Chat
Once the plugins have been enabled you will click on “New Chat” in the upper right corner. Then select the “ChatGPT-4” model. Next, you’ll see a dropdown screen, click on “Plugins (beta)”.
ChatGPT Plugin Store
Under the model selection buttons, you’ll see a small toggle arrow below, click on it and then click on “Plugin Store”.
You’ll then see a dashboard like this:
Installing Plugins
Now, within the plugin store, you can search for “TopHap” or, you can do what I did and search for “real estate”. Once you see the plugin, click on “Install”
Once the plugin is installed, you’ll see the logo beneath the model selector, make sure that the checkmark is filled, this will let you know that the plugin is installed and active.
TopHap ChatGPT Real Estate Plugin
When it comes to tech innovations, TopHap is a real estate intelligence platform that is known for its geospatial data visualization and analysis that has a focus on residential and commercial real estate. I was curious to see how refined the data extraction capabilities and how the quality of this plugin stacks up against manual property record searching. Let’s see if this tool can save us time.
Property Characteristics
First, I wanted to see what kind of data it would give me with a property in a non-disclosure state. To test the database, I went on Zillow and randomly selected a property to search. Next, I simply put in the address and this is the result it gave me:
The data produced is plentiful, easy to digest and includes relevant information such as the:
- address,
- estimated Value,
- last sale record,
- property classification,
- lot size,
- occupancy status (owner occupied vs non-owner occupied),
- neighborhood,
- school District and,
- miscellaneous other features.
I then did a search on four properties that I own, three were returned with the correct characteristics while the last one was returned with characteristics that were completely incorrect. Although I strongly believe that the inaccuracies were a byproduct of incorrect information being on file with the City of Philadelphia. A prime example of inputs determining outputs.
Comparable Sales
Next, I wanted to see how TopHap would fare with the task of producing comparable sales in the Philadelphia market, so I started a new chat and entered the following prompt:
“Give me 3 comparable sales that closed within the last 6 months and that are within 1/2 a mile from 533 N Paxon St., Philadelphia, PA 19131. The sales should have 2-3 bedrooms, and a square footage ranging from 700-1200.”
After two attempts and about 15 seconds of processing, here’s what the TopHap plugin returned:
Which shockingly was not bad, I recently had the property appraised and the first comparable in the recommendations the plugin returned was a sale that the appraiser used to support their opinion of value.
Zoning Data
Next, lets see if TopHap can deliver accurate zoning information. To test this out, I entered the prompt:
“what is the zoning of 3107 North Stillman Street, Philadelphia, PA 19132”
TopHap then returned the following information:
“The property at 3107 North Stillman Street, Philadelphia, PA 19132 is zoned as “RSA5” according to the data retrieved.
RSA5 is a type of residential zoning in Philadelphia. It’s typically used for areas with row houses or townhouses. However, for the most accurate and up-to-date zoning information, I recommend checking with the Philadelphia Department of Planning and Development or the city’s official government website.”
Which is actually correct!
However, it appears that the TopHap’s outputs will only be as good as the data it can fetch. I tested this by asking for the zoning of a property in another city that I’m familiar with that doesn’t have much of its public records online. The results as you can imagine came up empty:
Property Taxes
Another area where this chatGPT real estate plugin shines is providing property tax data. To test this out, I entered the following prompt:
“What are the current property taxes of 3107 North Stillman Street, Philadelphia, PA 19132?”
The answer again was spot on:
“The current property taxes for the property at 3107 North Stillman Street, Philadelphia, PA 19132 are $1,411.00 for the year 2022. Please note that property taxes can vary year by year due to changes in tax rates or property assessments. For the most accurate and up-to-date information, I recommend checking with the Philadelphia Department of Revenue or the city’s official government website.”
Here’s a cross reference to what is on the City of Philadelphia’s website:
Limitations of the TopHap ChatGPT Real Estate Plugin
Next, let’s explore the limitations of the TopHap ChatGPT real estate plugin.
Estimated Values
I put the plugin to the test in the City of Philadelphia which is a fairly easy market to comp properties in as it’s mostly composed of attached brick townhouses. All of which are nearly identical in square footage, bedroom count and bathroom count.
To prompt the TopHap plugin to return an estimated value, you would just simply type in the address of the home you are researching:
The values provided from their AVM (automated valuation model) are not that far off if the home is in Average condition. However, the drawback is that the tool does not account for the property condition. As an investor, the AVM is useless for homes that are fixer uppers or for determining the ARV of a property you are analyzing. The Zillow Group’s zestimate AVM slightly outshines TopHap’s offering. For the time being, a better suited technique for these kinds of homes would be doing a good old fashioned comparative market analysis.
Additionally, a prompt requesting sales price ranges within a specific zip code also turned up empty, the best this real estate plugin could do was provide a median sales price point.
Comparable Sales
As mentioned before, asking for comparable sales with a specific condition rating is currently not possible. When I asked for comparable sales that were in good condition based on previous real estate listing descriptions, the generative AI plugin indicated that it did not have access to specific property listings.
Additionally, although this plugin can return valuable insights and filter through prompts like the number of bedrooms, and square footage, an apparent limitation that the TopHap plugin struggles with is returning results based on the type of home. When I asked the AI to search for 6 comparable sales for a two-family home within a 1/2 mile radius, the list of homes generated were a mixture of single family and two-family homes.
Overall, until this technology reaches the alpha phase, a limitation of credible real estate information is to be expected.
Rental Data
Another limitation of this plugin is searching for market rent within a specific market area. After many unsuccessful attempts to generate rental data ranging from specific bedroom count to general fair market rents, it appears as though the data is only limited to sales records.
Building Permit Data
TopHap’s website states that it compiles data on building permit counts, which could be excellent for assessing builder/rehabber confidence. However, when I attempted to search for issued permit counts in the Cities of Philadelphia, New York, and San Francisco, the AI apologized for its inability to source the data. Despite each city maintaining robust online databases.
However, the lack of this information may be due to
The Wrap Up: How Does the TopHap ChatGPT Real Estate Plugin Compare to Conventional Methods?
Overall, for quick property characteristics, this plugin does a good job of delivering accurate information. The user interaction is very similar to using ChatGPT without any plugins. The total time between a prompt and delivery of information was on average less than 30 seconds. However, for value estimations, rental data, building permit counts and more detailed records requests, the performance of this plugin was limited (it will be interesting to see how the zillow ChatGPT plugin will compare).
All in all, It appears that the results will only be as good as the source of public record data and scraping methods this ChatGPT plugin is using. While there is a basic framework and immense potential for a more efficient and quick due diligence workflow. Doing anything more than just fetching general property characteristics leaves a lot to be desired. There is still a lot that needs to be fine tuned before one can realistically expect AI to change the way people perform due diligence when purchasing a rental property. It will be interesting to see how future real estate plugins will shape tomorrow.
