I Asked ChatGPT to Find Me an Apartment – Here’s What Happened

What happens when a renter skips Google altogether and asks AI to find them an apartment?

In this episode of The Multifamily Marketers Podcast, hosts Josh Grillo and Caroline Mills run a real-world experiment using ChatGPT to search for an apartment in Atlanta—without mentioning a single property by name.

The results reveal how AI is changing apartment discovery, why some communities are recommended while others are ignored, and what multifamily marketers need to do today to improve their visibility in the next generation of search.

You’ll learn:

• What happens when renters use ChatGPT instead of Google to find an apartment
• Why AI recommends some communities—and completely overlooks others
• How reviews, neighborhood content, and third-party listings influence AI recommendations
• Why having a great website alone is no longer enough
• How conversational search is changing traditional SEO strategies
• Practical steps to improve your property’s visibility across AI-powered search platforms
• A simple exercise every multifamily marketer should try this week

To suggest a topic, get in touch!

For more multifamily marketing help, connect with us at Resident360.

Drop Us a Review If You Enjoyed the Episode!

TRANSCRIPT

Episode 53: I Asked ChatGPT to Find Me an Apartment – Here’s What Happened

Podcast Co-Hosts: Josh Grillo and Caroline Mills, Resident360

Josh: Welcome back to the Multifamily Marketers Podcast. I’m your host, Josh Grillo, co-founder here at Resident360, and I’m joined by Miss Caroline Mills, our director of digital marketing. Caroline, how’s it going today?

Caroline: Hello! We’re hanging out well here in Florida today. How about you guys? Your kids just started school.

Josh: Oh, man. Yeah, they are not loving life right now. They just finished their first week of school. It was fine. One’s in high school and one’s in seventh grade, but it’s the same conversation every day. “Hey, how was school today?” “Not good.” “What do you mean?” “Oh, it’s boring. School’s always boring.” It’s like pulling teeth trying to get information from these kids. But that’s okay, because we have an exciting topic. We’re talking more about AI. More specifically, I asked AI to find me an apartment. What happens next?

Let me set the stage for everybody listening. What if I told you the next person looking for an apartment might never do a Google search? They might not go to Apartments.com or even Zillow. Instead, guess what? They might go to ChatGPT and say, “Help me find an apartment.” That’s what we’re covering today. We’re going to run a little test, right, Caroline? I know you spent some time setting this up, and I know our listeners can’t see the screen, but you’re going to walk us through exactly how this works. Set the stage for us.

Caroline: We thought it would be really interesting to do this exercise live together. Josh actually hasn’t seen the results yet, so he doesn’t know what I’m going to pull up. I gave ChatGPT a renter profile as if I were someone starting my search, and we’re going to look at it together from a marketing perspective. What information is AI using? Which properties are showing up? Who isn’t showing up? And what does that mean for marketers?

Before we look at the results, let’s talk about what we’re actually testing. I wanted to act as a real renter, not a marketer trying to force AI to recommend a specific property. I know nothing about the properties that came up. I don’t know the management companies or anything about them. I created a renter profile, and what I said was: “I’m a 28-year-old woman looking for an apartment in Atlanta, Georgia. I want a short commute; I work hybrid, in the office downtown. I’d like a one-bedroom with a den or a two-bedroom. I also need it to be pet-friendly and have an elevator.” So I gave it the kind of long-form query we always talk about. I intentionally didn’t include any apartment community names or tell it where I wanted it to send me. I just gave it information a renter could realistically provide.

Josh: That’s important, because we’re really asking: If AI had no reason to know our property existed, would it still find us?

Caroline: Yes, exactly. I know you can’t see this at home, but I’m sharing my screen with Josh right now so he can see my ChatGPT results.

I chose those details because I thought Atlanta would be very broad. It’s a huge market with a lot of saturation. I said I wanted to be near downtown, which also has a lot of saturation, but I wanted to give it something to work with on the commute, because I always hear about Atlanta traffic. Then I gave it very specific things: an elevator and pets. And I asked for a one-bedroom with a den or a two-bedroom, because in this scenario, the person works hybrid and wants space to work from home.

I think that’s important because these aren’t SEO-driven keywords. I’m just saying I want pets and an elevator. That’s nothing crazy. I didn’t even give it a budget, because I didn’t want to narrow it down to luxury apartments or a certain area. A renter isn’t sitting there thinking, “I need a luxury apartment with a 4.7-star Google rating and strong local authority.” They’re thinking, “I have a dog. I work downtown. I want to be near restaurants, and I don’t want to spend more than X.”

Josh: Which is a completely different way of searching.

Caroline: Right. It’s a lot more conversational and contextual, and there’s a lot more intent behind it. So Josh, here are our results.

The first thing I noticed is that it brings up a map. I thought it was interesting that it pulled in a lot of Yelp ratings for reviews. I didn’t ask for the best apartments or anything like that. It just brought up places with a huge range of ratings: one with a 4.7 from Google, and another with a 3.6 from Yelp. It didn’t say anything about them.

Next, ChatGPT breaks my priorities down into a chart. Something interesting it does that Google doesn’t: I told it my age, and it said, “Age 28: useful for lifestyle and community considerations.” I thought that was interesting, because you can’t really put that into Google and match it to keywords on someone’s website.

Then it says, “Here’s where I would start looking.” Again, we know nothing about these communities. It says: “Old Fourth Ward: probably my first search. This would be my starting point because it checks a lot of the boxes simultaneously. It’s immediately east of downtown and has walkability, BeltLine access, restaurants, bars, and a large supply of newer apartments. Current sources put a typical drive downtown at five to 10 minutes, although Atlanta traffic obviously makes the actual commute variable.” The information it’s citing is from Apartments.com; actually, both of those sources are from Apartments.com. So it tells me why I’d choose that area, and then it gives me some apartments there to investigate. That’s the first area. Then it goes into Midtown and Inman Park. I didn’t tell it where I wanted to live, so it’s giving me a lot of information about these different areas.

Josh: I’ll tell you what, Caroline, scroll back up. I know everybody can’t see the screen, but I think this is super interesting, because I haven’t done this personally in a long time, and it’s changed dramatically since the first time I tried it. On the screen, there’s a map of Atlanta. It basically looks like Google Maps, with little star ratings all around it: a 4.3, a 3.1, a 3.5, a 3.4, a 3.0. At the bottom, it highlights two apartment communities, with another one coming off the screen, so it looks like you can scroll over to see more.

Wait, go all the way back to the left, Caroline. Okay. What I think is really interesting, guys, is the order of the properties showing on the screen. And again, this is desktop; it might look different on mobile, too. Think about that. A 4.0-rated property is showing first. The second property is a 3.7. Scroll over a little, Caroline. The next one is a 3.6, and the fourth one, which isn’t even showing on the map until you scroll, is rated 4.7. So ChatGPT isn’t putting the highest-rated property first on the map. I just thought that was pretty interesting. Okay, but what does it recommend first, based on what you put in?

Caroline: It recommends the Ponce City Market and North Avenue area along the BeltLine. It tells me about that neighborhood because I mentioned walkability and commuting, and it really focused on that first. It puts a lot of emphasis there. Then it goes into Midtown next: “Best if minimizing the commute is the absolute priority.”

Josh: Okay, let’s look at the very first property. It says Anthem on Ashley Apartments. Scroll back up to your map. That’s the 4.0-rated one, the first one on the screen, right?

Caroline: It was.

Josh: Interesting. So visually, it shows the 4.0 first, the 3.7 second, and so on until it gets to the 4.7. I think that’s really interesting. It’s taking your profile, what you put in, and it’s not just pushing you toward the highest-rated property. It’s playing matchmaker and trying to match you with what it thinks is the best fit for you, regardless of star rating. I think that’s fascinating.

Caroline: Yes, and I like that you mentioned that. Another interesting thing, which you can’t see at home, is that you can now navigate without leaving ChatGPT. I clicked the link it included, and it opened on the side of the screen, so I can still see my results. I clicked on Anthem on Ashley Apartments, and I can see a few photos and what looks like a basic overview. Then there’s a “What people say” section, and the very first line, in bold, is “Residents frequently highlight the location.” So even though it doesn’t have the 4.7 rating, residents mention the location most, and I talked about walkability, so that’s what it put first for me. I thought that was interesting.

Josh: All right, let’s jump into the meat of it: the why. Caroline, if you’re a marketer listening to the show right now, what are you supposed to take away from this screen share we’re describing?

Caroline: I think there are a few things. One, which we talk about often: AI needs information. It can’t recommend what it doesn’t understand. If your property website says, “Luxury living in the heart of the city,” that’s great branding, but it doesn’t necessarily answer the questions a renter is asking. Where are you actually located? How far are you from downtown? Are you pet-friendly? What’s nearby? What’s the price range? What floor plans do you have? And what really makes you different?

Second, you want a great website, but your website isn’t the entire internet. It’s only one piece of the puzzle. AI can encounter information about your property through search results, reviews, directories, ILS listings, your Google Business Profile, social content, and other third-party sources. So you can have a really beautiful website and still have a really weak digital presence.

Josh: And I think that’s something multifamily marketers have traditionally struggled with anyway, right?

Caroline: Exactly. We’ve spent years saying you need to own your search results, and now we’re entering a world where we’re really asking, “Does the internet actually understand my property?”

This is where I thought we could make the episode a little more interesting. I wanted to talk about who didn’t show up, because I think that’s more interesting than who did, especially if this is happening to you. Location was my main priority, so I searched Google, with no long-form query, for “apartments near downtown Atlanta, Georgia.” I got completely different answers than ChatGPT gave me. Sure, there were some sponsored results at the top, but I got Broadstone 230, with a 4.6 rating, right downtown. I got The Lofts at Centennial Yards. None of those came up in my ChatGPT results at all.

So I pulled up the Broadstone website to see why. As I scroll through, it’s very visually appealing, with lots of photos, but there isn’t a lot of content. It emphasizes simple pricing and real living, with lots of gorgeous photos, but nothing there answers the questions I just asked ChatGPT. This property is beautiful. It’s a few miles from downtown. It has amenities and everything I could be looking for, but AI didn’t recommend it.

Josh: Can we actually know why?

Caroline: Not necessarily. That’s a good question. We see these things and make our own assumptions right now, but AI doesn’t give us a transparent ranking algorithm. We can’t say a property is number one because it has exactly 127 reviews. What we can do is look at the information available to AI and start identifying patterns.

Josh: So does this mean SEO is seriously declining, or dead, or something else?

Caroline: That’s a good question, and a lot of people are asking it right now. I would say no. If anything, I think SEO is becoming broader than what we’re used to. We’re moving from “How do I rank this page?” to “How do I make my property discoverable across the entire search ecosystem?” And that’s a much bigger question.

Josh: Okay, so let’s say I’m a property marketer listening to this and thinking, “Great, AI is changing everything.” What am I actually supposed to do?

Caroline: That’s very valid. I’d start with a few things. First, do what we did: look at your digital footprint. Search your property name and see what comes up. Then search generic questions a renter might ask, like “best apartments in Atlanta,” “pet-friendly apartments near Midtown,” “apartments near [this employer],” or “best apartments under $2,000 in Atlanta,” and see whether your property is even part of the conversation.

Make your property information extremely clear. You don’t want renters or AI to have to figure out what you offer. Include your location, amenities, pet policies, floor plans, pricing, neighborhood, transportation, nearby businesses, fees, and everything else you can think of.

Then, just like you and I talked about, pay attention to your reviews. What are residents actually saying about you? You can’t control what they say, but is it helping you show up or not? In this case, “Residents frequently highlight the location” is what made that property rise to the top. So what are residents saying about the things renters care about, whether it’s location, staff, maintenance, noise, or parking?

Also, don’t ignore third-party listings: your ILSs, your Google Business Profile, and directories. They all need to tell the same story. If your website says one thing and Zillow says another, you could be creating a lot of confusion.

And then test it. This is the easiest part. Open ChatGPT, just like I did on the show, and ask it to find apartments in your market. Then change your renter profile and see whether someone different could find you. Ask about different neighborhoods, budgets, and amenities, and see what comes back.

Josh: So Caroline, do you think AI will actually replace Google for apartment searches? Obviously not right now, but in the near future?

Caroline: I think we’ll continue to see more fragmented search behavior. We’ve seen that this year. Reports have said Google searches are down a significant percentage year over year. Some renters will still use Google. Some will stick with Apartments.com and Zillow. Some will start on TikTok, some will search Reddit, and some will go to ChatGPT. The mistake would be assuming you only need to optimize for one search engine. The real goal is to be the property that gets recommended wherever the renter starts their search.

Josh: Yeah, and I think that’s probably the biggest takeaway from today’s experiment.

Caroline: Completely agree. AI search isn’t just an SEO conversation. It’s a digital presence conversation, like we talked about today. If AI or a renter tries to understand your property today, what will they find? Is it accurate? Are people talking about you? Are your reviews strong? And most importantly, would AI recommend you?

Josh: I guess we’ll have to find out, right?

Caroline: I’d say that’s your homework for this week. Go ask AI to find an apartment in your own market. Don’t give it a property name, and don’t tell it what you want it to recommend. Just be the renter and see what happens.

Josh: I think that’s a perfect wrap-up for today’s episode. Hopefully everybody got a lot out of it. Sorry you couldn’t see the screen, but hopefully we talked through it clearly enough for everyone to follow. If you have any questions, feel free to reach out to Caroline or me. We’re always available and love feedback from our listeners. As always, we’ll be back next week with another episode of the Multifamily Marketers Podcast. We appreciate everybody listening today. Thank you, guys. Bye-bye.

Caroline: Bye!