Zillow can’t smell if 20 cats live there. I’ve said that for years, but here we are in a new era where so many consumers are using AI to understand value. So, it’s time to amend the statement. AI can’t smell if cats live there. I think we’re at an interesting place with AI and consumers, so I wanted to share some brief thoughts. And then I want to talk about a growing portion of the market – lower prices.
UPCOMING SPEAKING GIGS:
4/22/26 EDCAR
4/28/26 PCAR Rocklin
5/7/26 Empire State of Mind
5/13/26 Investor Webinar TBA
5/14/26 Event TBA
5/15/26 Nevada County TBA
5/21/26 Event TBA
5/27/26 Solano County TBA
6/3/26 Wisdom Wednesday in Elk Grove
8/6/26 PCAR Auburn
10/2/26 PCAR Rocklin
10/21/26 Coldwell Banker Sierra Oaks / EDH
It seemed fitting to have an AI image today (made with Grok).
BUT AI SAYS MY HOUSE IS WORTH…
I’m starting to get more conversations with clients and consumers based on what AI says. It seems like, “But AI says…” is the new “But Zillow says…” What I find to be unique about AI though is the conversations are going beyond just the home value. With AI, for me, it’s been more about what one particular item might contribute to the value or the return-on-investment for various improvements. I’ve had a client share the value of an ADU based on what ChatGPT says, and I had a recent client provide an itemized list of what they spent on improvements over the past four years and what AI says each improvement contributes to value.
TWO EXAMPLES OF QUESTIONS I ASKED AI:
I’m not saying these examples are accurate (be super careful with stuff like this). I am saying consumers are at the other end of the phone or desktop and asking questions. So, this is where real estate professionals have to be prepared for conversations.
I asked Grok about an HVAC:
I asked ChatGPT about the value of an ADU:
SOME THOUGHTS ABOUT AI
- Be ready for more real estate conversations. AI feels like it’s going to help real estate conversations get deeper, and I’m here for it.
- Don’t forget to fact check what AI says. Dig into some of the cited data sources. If we only read headlines and AI summaries, that can be dangerous.
- AI is going to be valuable for business and valuing properties. I’ve only scraped the surface, but I’m blown away at what some colleagues are doing.
- AI can make life efficient, but it can’t live for us, so we have to remember to build relationships with other humans. Look, I love technology, and we’re clearly advancing in some ways, but it seems like society is getting lonelier despite technology, and that’s a huge concern.
- As a side note, I’ve noticed people use AI on me in two ways. Sometimes my profile pic is being sent through AI for speaking engagement flyers. It’s a little weird to be honest. I’d rather be shown “as is.” Last year somebody even put my profile pic in a suit instead of the blue shirt. Haha. The second thing I’ve noticed is a few people send my charts through AI and then strip out credit to me. This is a rarity, but it’s not cool. The wild part though is AI sometimes botches the columns and stats. Yikes.
Maybe AI me is hotter though. Onlookers might not see any difference, but my family thinks I look weird (AI version) because they know me best.
Now let’s look at a growing part of the market…
MORE STRENGTH AT LOWER PRICES IN 2026
Do you know where we’ve seen the most growth in closed sales volume this year? It’s at the lower end of the price market – particularly between $400,000 to $500,000. This isn’t true in every single local county, but it’s true in the region as a whole and many counties. My thesis is prices have dipped a little from last year, and that’s caused more properties to sell within lower price points.
Scroll down for twelve counties of stats like this.
ARE THERE MORE FIRST-TIME BUYERS?
It’s possible we have more first-time buyers, but I can’t really say for sure. There isn’t a perfect way to uncover this with MLS stats. I will say it’s a positive trend to see more sales at lower prices though since that’s a range where first-time buyers are playing the game. By the way, here’s financing and cash since 2009 (FHA alone is not a good metric to gauge first-time buyers).
A VIRAL NARRATIVE ABOUT FIRST-TIME BUYERS
The National Association of Realtors came out with some new stats this week to show first-time buyers are at record low levels, but there are some questions about the study. Look, it’s possible NAR is correct. I’m just not confident about their stats. I am not anti-NAR here either, so save your hate mail. These stats are from a national survey that got 6,103 responses across the country. This is the same 120-question survey mailer that was used to say first-time buyers are now forty years old (other sources don’t show that same age). I’m not aware of another data source that can confirm or contest this survey, but I’m open to suggestions if anyone has a preferred source.
TAKE A LOOK AT THE METHODOLOGY
Here is the methodology for this study. You can click here to see directly that this was the methodology for the screenshot above. I like other NAR stats, but I’m not willing to share their narrative from the survey. It’ll go viral everywhere, but I’m going to hold back on this one.
WE DO HAVE AN AGING POPULATION
All that said, we have an aging population nationally with a birth rate that is slowing, so it wouldn’t be surprising to me in decades to come to see more transactions at older ages. There is so much talk about a Silver Tsunami as Boomers grow in age. Just remember that not every Boomer is going to die and list their home in one moment. Here’s a visual from my ResiClub Pro Membership to show the share of population that is 55 and older. Interesting, right?
TWELVE COUNTIES OF DATA (FLEXING AT LOWER PRICES)
Check out twelve counties of stats to see what I mean about more strength between $400K to $500K. Take smaller counties with a grain of salt since three months of data isn’t much to work with, and recognize that some counties have had a struggle with volume this year (San Joaquin in particular). If you work in real estate, keep asking yourself which price points are growing and contracting. And if $400K to $500K keeps growing, connect with investors who may want to sell also.
Anyway, I hope that was helpful. Thanks for being here.
Questions: How are you seeing AI show up in real estate conversations? What are you noticing happening at lower prices right now?
If you liked this post, subscribe by email (or RSS). Thanks for being here.
Happy Friday. Have you shifted your publishing date?
Interesting post this morning. I have yet to have someone ask about AI results so appreciate the heads up.
I like the change in volume by price point but it leads me to wanting more. Is there a change in the composition of sales, i.e. smaller homes selling, or is it another way to point to declining prices, or both? My next steps would be to trend prices over time and to trend house size over time to learn more. But cool idea that I’ve not considered.
Happy Friday. I haven’t. I have just been so busy that I’ve written last-minute. Last week it was about wanting more time to pull March stats. Ideally, I’d like to push out a post on Wednesday or so, but it’s just been an overwhelming season really.
Hmm, I’ll think on that. I do track size every month for all the counties, but I really only compile it for the entire region. It’s been trending down a little lately year-over-year (not a shocker if we have more sales at lower prices). This is on my radar. We did have slightly more sales at lower prices for all of last year too, but I’m still watching this as an emerging trend.
I’ve been seeing ChatGPT give clients really bad information that has caused issues in deals. At one point it told these buyers to submit under asking for a home that had already received multiple offers. These same clients were later told by ChatGPT that the sewer line material (ABS) was problematic which led to a mini panic attack before I explained to them that information is misleading and ABS is the current standard for sewer lines. I’m all for clients doing their own research but concerned that especially with Chat it tends to conform to whatever the user wants to hear.
Yikes. And that’s not the pretty party. I like the idea of AI spurring on real estate conversations, but unfortunately it will do stuff like this too. I heard Ryan Serhant mention something about a deal almost blowing up because of this. I guess it goes along with what I was saying. We have to fact check AI. On that note, I notice when I do fact check AI, it’s very quick to say I’m right about something. If it was that quick to admit an error, couldn’t have just gotten it right in the first place?
Thanks, Gaea. We are in new territory here, and we need to keep our antennas up for more conversations like this. I was just talking with a client who is nearly 80 years old, and she uses ChatGPT. Wild.
There are 21 cats in the AI image
Oh my gosh, that is so funny. I didn’t even think to count. And you’re so right. I think I told AI to add 20 cats, but I guess they added 21. Haha.
Great conversation. AI can’t smell the cats, but humans can tell AI about the cat smell and ask for analysis. I have too many questions that I don’t expect you to try and answer. The response from AI about the ADU is completely useless. $140k to $300k plus. Why even give a range of value at that point? Also, should AI be allowed to give a number or range? Isn’t that unlicensed appraisal practice? I guess it’s not specific to a property so maybe not but maybe something to regulate AI in the near future.
I hear you. I’m very concerned about some of the answers. Here we are though. New conversations will emerge. The danger is so many people don’t fact check. I like the HVAC answer. Not digging the ADU one though. And that’s a viable question. I’m not sure what the answer is. I do know people will likely tend to gravitate toward answers they like to hear. Unlike Zillow, consumers can now ask about so many things (including what the market is doing). There are interesting times.
Great post, prompting so many thoughts! One thing I do like about AI in general is that it knows what I just asked, so unlike a google search, I can dig into a topic with it for a while. It’s actually been helpful on non-real-estate topics for me. But I’ve tried to use it to get some quick ideas on comps for an offer on a home, and it’s terrible. I would say something about searching within one mile, similar sized homes with pools, and at first it seemed good with interesting narrative about what the strengths of the subject property were, but when I fact checked the comps, quite a few were not within the mile I requested. On followup, I told it to exclude those in gated or senior communities, and it assured me that the new list had excluded them, but it was lying. 🙁 It was a place to start, MAYBE, but it was inaccurate and needed a lot of manual followup.
Regarding comparisons to last year, do you think the coming months might show a kind of illusory jump in sales over 2025, since the market slowed way down last year starting in April with the tariff drama?
So interesting. Thanks, Steve. This is where AI is only going to be as good as available data, and I can’t imagine the stats would be there for a detailed search like that.
I’m glad you brought this up. And I’m also thankful last year we didn’t have 2025 uncertainty in the year-over-year stats I shared today. Some of the comparisons are going to get weird. I’ll have to write about that. Sellers really started to back off the market after April, and if they continue to come back this year, the year-over-year comparisons are going to look stronger this year. This is where stats for both active listings and new listings could get a little awkward. And while we’re at it, stats for days on market are going to start looking faster than one year ago maybe even by next month. The % of sales selling above and below the original list price are going to start to look more aggressive during parts of the next two quarters too. Closed sales get a little interesting because we actually had more volume last year compared to 2024 overall, but we had a couple of strong months in the fall, and that really helped. When I think about 2025, both May and August were weaker than 2024 for closed sales, but there wasn’t anything too crazy. I think the bigger story in the stats was sellers backing off the market, and buyer demand remained pretty lackluster. I’m not terribly concerned about the comparison for closed sales numbers because it was so low each year, but there will be other stats I mentioned that are going to be a different story because it will look more competitive today. As always, we need to be aware of what we’re comparing to last year… I’ll dig in and say otherwise if needed, but this is my general take for now. Do you anticipate any change in the stats?
I enjoy Zillow back in the day, Sellers would tell me Zillow says. My response was did you check their error rate in the fine print. They would say NO and I would follow go take a look and get back with me. Usually, they would call back and ask when I can start or say, okay guess your value was okay. Now I have to work on the AI Hallucinations and bring that to the forefront. But I will finish with none of this really matters. What really matters is what your family thinks of your dashing good looks. Hope all is well my friend. Take good care of yourself and remember to recalibrate from time to time. You and I both have to keep track of us.
Haha. Yes, it’s all about good looks. And unfortunately, that is what fades as time goes on…. Thanks for the reminder on health. Have a wonderful day, Brad. And a tremendous weekend.