Blog

From her home in Mexico, JoAnne Kim started searching for a Philadelphia-area apartment by opening ChatGPT.

Kim, a regular user of the generative artificial-intelligence tool, told OpenAI’s chatbot what she and her husband wanted: a quiet place in Pennsylvania that was walkable and dog friendly with good transit access, a strong sense of community, cafes and restaurants, and a lot of things to do.

ChatGPT suggested Ardmore, Bryn Mawr, and Chestnut Hill. The couple chose Ardmore, and Kim asked the chatbot for a list of apartment communities that fit their criteria and included outdoor space.

ChatGPT “helped me become a better apartment hunter, because I was able to use all the data to help me make a decision,” said Kim, a 64-year-old career mentor for university students.

This month, the Kims moved into Montgomery Plaza near Suburban Square.

Generative AI tools are becoming increasingly popular as more people use them to write emails, summarize documents, and recreate pictures of pets. And, like Kim, renters are also using AI chatbots to find apartments.

In Philadelphia and beyond, the growth of AI is changing the way people search for homes — and it’s prompting some in the industry to change how they do business.

In April, Philadelphia became one of 15 cities where RentCafe, a national apartment search website, rolled out its AI assistant, Ren. The platform is preparing to make the tool available nationwide. Apartments.com launched its latest AI tool — Apartments.com Ai — in June and started an advertising campaign in August to get more renters to use it.

Property managers, too, are trying to find ways to reach people who use AI.

Charlie Marshall, a leasing associate at the Ardmore-based property management company Harrison Richards, which manages Montgomery Plaza, has been tweaking the company’s website to try to get its seven properties more prominent placement in generative AI search results.

Marshall said the handful of people who said they found a building using AI tools seemed more informed and committed than other prospective tenants.

“It was immediately clear they had a good idea of the places they wanted to lease at,” he said. “It was not like some people you see that are touring seven or eight properties. They had narrowed it down to like three.”

Stuart Richens, vice president of product at Apartments.com, said because the company’s AI tool “is immersive and kind of keeps guiding you and helping you, renters are getting much deeper into the search more quickly.”

“And that’s helping them get to that short list more quickly,” he said.

Kim and her husband, a senior data analyst who uses AI, worked with ChatGPT to focus their search. But to find a place that felt like home, Kim also watched local real estate agents’ YouTube videos, read reviews, watched local news coverage, and talked to property managers.

Kim said that because she started with virtual conversations on ChatGPT, “I spent more time thinking about the kind of life we wanted to have. It wasn’t just about finding the apartment. It was the neighborhood, the vibe.”

‘A more advanced Google search’

AI tools can only do so much.

Vee Gordon, a rental agent with Philly Home Girls, said she worked this year with a New York City couple in their 20s who wanted a two-bedroom home in Philadelphia. They insisted on seeing places that were hundreds of dollars over their $6,000 monthly budget. During a meeting with Gordon, they pulled up ChatGPT.

“They were asking it how to negotiate a rental price down,” she said. “I feel I have more insight on that than what ChatGPT was saying.”

Gordon has worked exclusively with renters — many in their 20s and 30s — for four years in the city and its suburbs. She estimates she has helped sign almost 500 leases.

“This is my expertise,” she said.

The couple didn’t end up finding a place through her.

Jack Olmanson, 27, doesn’t regularly use generative AI tools. But this year, when he and his girlfriend needed to move from Washington, D.C., to the Philadelphia area for her work in the equestrian industry, he turned to Anthropic’s Claude for help.

The couple used it “as a more advanced Google search,” said Olmanson, who works in business management.

“You can be very customized in your asks,” he said. Claude “was good at giving us everything that’s out there that meets our initial criteria.”

The couple wanted a two-bedroom apartment in a walkable area and a building that welcomed Mya, their yellow Labrador retriever.

Olmanson and his girlfriend are originally from Minnesota and didn’t know the Philadelphia area, so they asked Claude to compare suburban towns here to those they knew in the Minneapolis area. They also asked it to make a map of locations with reasonable commute times.

Once they got suggestions from Claude, they did their own research, exploring websites, looking at photos, reading reviews, and taking tours. And they still reviewed Zillow listings they got through email alerts.

They moved into an Ardmore apartment building managed by Harrison Richards in May.

Olmanson recommends renters verify everything an AI chatbot tells them to make sure it’s not making things up. And remember, he said, it’s just a tool.

“You’re still in charge of making sure [a home] is the right fit for you,” he said.

How AI has changed apartment listing websites

When Philadelphia renters begin searching RentCafe’s website of home listings, a button prompts them to “Ask Ren.” Apartments.com encourages them to “Meet your Ai rental adviser.”

Renters using AI tools tend to start broadly with a location and budget and refine searches through a virtual back-and-forth to get to a smaller number of the most relevant listings. Renters can type or talk. Apartments.com’s chatbot supports about 50 languages.

RentCafe found that people who use its tool are twice as likely to click into a listing and 50% more likely to contact property managers than people who use traditional search tools. Users of Apartments.com’s chatbot request 144% more tours.

“Conversational search, we’ve seen, has been very helpful for renters,” said Samantha Skrobot, renter advocate at RentCafe.

Renters can use it to figure out realistic budgets and determine locations accessible by public transit. They can search for requirements that aren’t included in traditional filters on apartment search websites, such as proximity to a laundromat or library. Users can ask AI tools to compare properties’ amenities and tell them about nearby schools.

Through the chatbots, renters can prioritize certain property characteristics (say, gets a lot of natural light) and indicate where they’re flexible, which they can’t do with traditional search filters. AI chatbots can suggest locations a renter may not have considered that meet their requirements.

On newer 3D tours on Apartments.com, renters can tell the AI chatbot to remove furniture from a room and give dimensions of a space.

Adjusting for AI

Because of the growth of generative AI tools, Richens at Apartments.com said that for property managers, “having a very complete listing with every piece of data about the property … is very important because that’s what renters are asking for.”

“And if you don’t have it, how is the AI going to answer?” he said.

Apartments.com is working with property managers to beef up listings.

At Harrison Richards, Marshall has worked to improve how the company’s properties show up in generative AI searches. He asked questions that a renter would ask and noted when the company’s properties did and did not appear.

A Redditor posted a couple years ago looking for Ardmore-area apartments, and someone said they’d lived at the company’s Montgomery Plaza property and liked it.

“That was apparently gold for ChatGPT,” Marshall said.

The property management company has changed heading sizes and where information shows up on its website, so AI tools can more easily pull out answers for renters. Marshall has noticed a difference.

After the changes, he said, the AI tools “would draw out the things we think are important and the things people might have questions about.”