This report is available exclusively to subscribers of Inman Intel, a data and research arm of Inman offering deep insights and market intelligence on the business of residential real estate and proptech. Subscribe today.
Like a lot of real estate professionals, Kenny Klaus wasn’t entirely sold on artificial intelligence — at least at first.
But Klaus, who runs the Klaus Team Real Estate Solutions in the Phoenix area, started to come around after asking his on-staff tech guru to experiment with the famed chatbot ChatGPT, and the results were both surprising and promising. But that experience wasn’t what actually changed Klaus’ mind about the technology.
Instead, he truly became a convert when has asked the bot for a love note, as if a message from a Hallmark card.
Kenny Klaus. Image: The Klaus Team
“My favorite one so far is I asked it to write a love note to my wife,” Klaus recalled. “I knew she’d know it wasn’t me. But it wrote a note to my wife and I thought, ‘Holy cow, this is good.'”
Klaus said the note was longer, more detailed and generally better than he had anticipated — especially because he was just looking for a few lines. But over the months since, Klaus and his team have leaned into AI. Today it’s a “regular thing” that his team uses to write newsletters and community pages and for brainstorming ideas.
Klaus isn’t alone.
For this story, Intel set out to understand how brokers are deploying AI in their everyday businesses. It was a project that could have been quixotic. Tools like ChatGPT and its image-generating counterpart DALL-E have only been publicly available since late last year. It seems entirely possible that many brokers and team leaders still aren’t actually using generative AI at all.
But just the opposite proved to be true. The industry professionals who spoke with Intel for this story all reported experimenting with AI. Many had experiences that paralleled Klaus’s, saying they approached AI with some skepticism before ultimately being won over.
From these conversations, there are two big takeaways. First, AI has already deeply infiltrated the real estate industry. This technology is not just for early adopters, tech enthusiasts and self-identifying nerds. It’s already everywhere, and that sets it apart from other fad-like platforms — say, Amazon’s Alexa Skills or social networks, such as Clubhouse — that ultimately proved to be flashes in the pan.
But second, it also remains to be seen how exactly real estate professionals will end up using AI in the long run. Many of the brokers who spoke to Intel are still in the experimental phase of their journeys, and few have committed any meaningful amount of resources to the field. There are others, however, who believe that brokers need to commit more fully to AI and who are actually building their own technology from the ground up.
The variation highlights how new this space is, and suggests real estate is at an inflection point as it grapples with the potential of an entirely new tool set.
Coming around to AI
Among the brokers who spoke to Intel, the most common way to use AI is, as Tiffany McQuaid put it, as “a foundation that you can catapult off of.”
McQuaid is the broker-owner of McQuaid & Company in Naples, Florida, who told Intel that she has instructed her marketing and graphics teams to experiment with the technology. It hasn’t always been easy.
Tiffany McQuaid
“There’s been a huge amount of hesitation and even a little revolt,” she said, explaining that her marketing staffers were apparently afraid that the technology could ultimately put them out of a job. McQuaid said that she was actually headed over to talk to the team and assuage their fears about AI as soon as her call with Intel wrapped up.
McQuaid doesn’t see AI as a replacement for human work. She compared the technology to Zillow’s Zestimate, which has existed for years now, but which is “still not totally accurate.” Her point was that as she and her team have experimented with AI, they’ve found that it’s something they can build on but not rely on wholeheartedly.
Case in point: McQuaid said that her company recently sponsored a wine event at a local animal preserve. As her team brainstormed ideas for the event’s branding, someone in the room came up with the idea of using an image of a lion holding a glass of wine. In the past, they might have searched for or built an image fitting that description. This time, they typed the idea into an image generator.
“Some amazing things came out,” McQuaid said.
In the end, the team treated the experience as a brainstorming exercise and didn’t end up actually using the AI-generated images. But McQuaid said the exercise was still a useful one because it “opened our eyes to a few possibilities that we hadn’t thought of.”
The eye-opening possibilities of AI are something that many brokers mentioned. Courtney Poulos is the broker-owner of Acme Real Estate in Los Angeles who told Intel that — much like Klaus — she was initially skeptical of AI.
Courtney Poulos
“When ChatGPT, first came out, I was like, ‘This is terrible. It’s going to take away the skill of writing beautiful language for your listings,'” she told Intel. “We’re really good at that and not everyone is. So that’s a competitive advantage for us.”
Poulos’ first exposure to the new generation of chatbots came via a news article about a journalist who “had a terrifying encounter with ChatGPT.” It didn’t sound good. But after reading the article, Poulos started “fiddling around” with the tools, and eventually realized, “It’s not the worst thing ever.”
The key for her was realizing that while AI can produce useful text, it can’t replicate the moments of human connection that lie at the core of what real estate agents do.
“Even if I use ChatGPT to give me a basis, I have to add beauty and those elements of actually being at the property,” she said. “This process can’t go on without me, and that’s what made me realize it’s a tool, not a replacement.”
Over the ensuing months, Poulos and her team have come to use ChatGPT as a “collaborative tool” for generating text. She said they still have to make a fair number of edits, including for issues such as fair housing, but the initial concerns have melted away.
The company is now leaning so hard into AI that Poulos recently brought in an expert to do an AI training for her agents. The tables have clearly turned.
Overall, the brokers who spoke with Intel for this story mentioned using AI mostly for tasks that involve writing. But as the technology infiltrates further and further into the industry, some are also coming up with other novel uses as well. Angie Stumbo, broker-owner of EXIT Realty Stumbo & Co. in Dayton, Tennessee, is among them.
Angie Stumbo
“I had ChatGPT write my voicemail message, which my husband hates,” she told Intel with a laugh. “I said, ‘Write a voicemail that encourages people to text me instead of leave a voicemail, and make it funny.'”
Stumbo — who said she “jumped on the ChatGPT trend pretty much when it came out” — explained that she has also used AI in concert with other tools, such as the image editing website Canva. And she’s begun experimenting with Microsoft’s AI features in such programs as PowerPoint.
These experiences have led her to agree with other broker-owners, noting that “it’s not going to be a substitute for making phone calls and showing properties.”
However, the thing Stumbo most likes about AI is that it potentially levels the playing field. She pointed out that as a broker of a small company — she has 12 agents — in a relatively rural area, she doesn’t have access to the same tools and platforms that are available to brokers at bigger firms. But as AI evolves, she sees it as a way to accomplish more with less.
“If I can do 10 times the amount of things with some assistance and free up time for my family or free up time for other things, I don’t see the negative of it,” she said.
Such comments highlight the fact that while AI is new, many brokers are already moving beyond the shiny-new-thing phase. They’re talking about it the way they might discuss a CRM or transaction management platform. The exact uses are up in the air, but everyone seems to be at least grasping for them.
Explosive growth
Hard data on how exactly real estate professionals are using artificial intelligence right now doesn’t seem to exist. But generally speaking, AI platforms have seen explosive growth.
Analytics firm Similarweb tracts AI usage and found that over the past seven months, ChatGPT in particular experienced “dizzying growth,” according to a report the company shared with Intel. That growth involved rising from effectively zero monthly visits in the latter part of 2022 to nearly 2 billion by this May.
Credit: Similarweb
Interestingly though, Similarweb has also found that the days of ChatGPT’s explosive growth appear to be ending: From May to June, the chatbot’s site saw a 9.7 percent drop off in traffic. In the U.S. alone, the traffic dip amounted to 10.3 percent during that period.
Unique visitors also fell during that time period by 5.7 percent, while the time spent on the bot’s website went down by 8.5 percent. Despite that dip, however, time spent on ChatGPT’s site is still up compared to where it was last fall when OpenAI, the company that makes the bot, first opened access to the public.
Credit: Similarweb
Even with a leveling out or falling in engagement, Similarweb found that “ChatGPT still attracts more worldwide visitors than bing.com, Microsoft’s search engine, or Character.AI, the second-most popular stand-alone AI chatbot site.”
What these numbers mean remains to be seen. ChatGPT’s website is something of a loss leader for OpenAI, costing as much as $700,000 per day to keep running, Similarweb noted. The analytics firm also points out that ChatGPT was never really “set up to run a mass-market, ad-supported website.”
It’s certainly possible that some of AI’s luster has worn off. But it’s also possible that the changing traffic numbers reflect a new idea reaching a saturation point in the public consciousness; six months ago, few had heard of ChatGPT, while today it makes headlines daily. In other words, companies typically grow fastest when they’re new.
That’s true of social networks like Facebook, but it’s also true of firms in other sectors, including real estate. EXp Realty, for instance, saw stratospheric agent count growth for many years but lately has seen that growth slow significantly.
And that brings us back to real estate. If the numbers above do reflect a sector that has broken through to become a mainstay in the public imagination, it makes sense that the sector is also getting widespread use among housing professionals. Artificial intelligence isn’t just some niche thing for early adopters. It’s mainstream.
If you build it
Most of the brokers who spoke with Intel for this story are in the early stages of experimenting with AI, and few have devoted significant resources to the topic.
Stumbo, for example, said that she loves new technology — “I’m like a squirrel with a shiny object” — but so far hasn’t actually paid for any AI tools, aside from accidentally forgetting to cancel a free trial. Ditto for Klaus, who said he hasn’t yet found the need to plunk down cash for AI. Poulos personally pays for the pro version of ChatGPT but hasn’t rolled out any company-wide platform at her brokerage.
By and large that is the consensus: Brokers are playing around with free versions but don’t even yet have a line item in their budgets devoted to AI.
Sean Frank
But that approach is not universal. Sean Frank, founder and CEO of brokerage Mainframe Real Estate in Orlando, is going further — much further.
“We’re working on a chatbot,” he told Intel, adding that his company is specifically focusing on training the bot on topics, such as the Realtor Code of Ethics and fair housing regulations. “And later on we’re going to be rebuilding our back-end systems from scratch and are looking to integrate AI into that.”
Mainframe employs two U.S.-based software programmers and depending on market conditions also contracts for additional work with a tech team that is based in Ukraine. The company has long had a technology focus and thus quickly embraced chatbots such as ChatGPT when they burst onto the scene.
Today, Mainframe’s own chatbot will be built on ChatGPT’s API, or application programming interface, which OpenAI has made available for others to use in their own projects.
Like other broker-owners, Frank said his team has used AI to produce written content. But they’ve also leaned on it for more complicated tasks, such as writing computer code. And he thinks there’s value in the customization approach he’s taking.
“Most brokerages are going to look for off-the-shelf solutions, but I think there needs to be a lot of caution,” Frank said. “I think that a lot of people understand what it is but I don’t think they understand how to customize it for their needs.”
Frank’s approach to AI highlights an alternative path compared to what many other brokers are doing right now and highlights that inflection point in which AI penetrates ever deeper into the housing industry. Some experts see the trend as the future.
Bobby Bryant
“I do believe that the brokerage of the future is going to have to bring their development inside,” Bobby Bryant told Intel.
Bryant is the CEO of Ask Doss, a home search portal and virtual personal assistant that uses ChatGPT’s latest iteration — specifically GPT4, which is not part of the bot’s free offering — and which can recognize natural language.
Bryant has worked as a broker in the past, but today as a technologist he’s unsurprisingly bullish on the idea of artificial intelligence. The concept is integral to his company’s goal of building a better way to find houses.
But he said that merely using free, publicly available tools comes with risks. Brokers could lose control of their data, for instance, or can end up perpetually paying more of their commissions to more and more parties.
“It’s getting scarier and scarier to be third-party-reliant,” he said.
Bryant consequently advocated for a vision of AI that’s more along the lines of what Frank is doing, and he suggested that companies in that vein are likely to have a competitive advantage over the longer term. Agents, in other words, may begin choosing companies because of the way they’re handling AI.
Already something like that is playing out with Compass, among other firms, touting their “boomerang agents” who have returned for the technology. That trend isn’t specifically about AI, but it does show how technology can be a deciding factor in the competition between brokerages.
Whether the same thing now plays out with artificial intelligence remains to be seen. But Bryant at least believes that companies wanting to compete today need to dive into AI.
“What the brokerages have to understand is,” Bryant argued, “you either have to get into this AI fight or you’re gonna lose.”
