The Neighbourhoods AI Doesn't Know About (And Why That's Your Edge)
By Matt Basedow
Every buyer sitting across from you has already asked ChatGPT what they should pay.
That's not a threat. It's just where things are now. Two out of three agents report using AI in their own business according to NAR's 2025 Technology Survey, so it would be strange if their clients weren't doing the same thing on the other side of the table.
The question isn't whether your buyers and sellers are consulting AI. They are. The question is what it's telling them, and what it's leaving out.
The Gap Nobody's Naming Out Loud
Kamini Lane, CEO of Coldwell Banker Realty, put her finger on it in a CNBC interview in June. She said the one thing AI consistently misses is knowing what's "up and coming," whether that's a neighbourhood, a street, or a design feature.
That's not a small gap. It's the difference between a client who prices a home off last year's comps and one who understands why the block behind the new cafe strip is about to move. AI tools are built on historical data. They're excellent at telling you what already happened. They have no mechanism for picking up the things that haven't shown up in the numbers yet: the new school zoning rumour, the developer quietly buying up three lots on the same street, the fact that everyone under 35 wants to live near that one particular strip now.
You know these things. You picked them up from twenty conversations at open homes, three coffees with a local town planner, and a decade of watching which streets turn over fastest.
The problem is that knowledge almost never leaves your head.
Why This Matters More Than It Did Last Year
Trust in AI for home-related decisions is actually falling, not rising. In the US, trust in AI to help find a home dropped from 30% to just 16% in a single year, according to Cotality's 2026 survey reported by HousingWire. Buyers are using these tools more, and trusting the output less.
That's the opening. Clients aren't looking to be told AI is bad. They're looking for someone who can catch what AI missed, and say it with confidence.
Buyers are using AI more than ever and trusting it less than ever. That gap is where agents win.
Most agents already have the knowledge to fill that gap. Almost none of them are showing it to anyone.
Show What You Know, Don't Just Say It
Telling a buyer "I know this area well" is a claim. It sits in the same category as every other agent bio on every other listing. What actually lands is showing the specific, current, slightly insider detail that no AI tool would ever surface: why this particular pocket is about to move, what's changing two streets over, why the design choices in this renovation aren't random.
This is where your listing marketing does more work than most agents ask of it. A property video isn't just a walkthrough of rooms. It's a chance to narrate the context an algorithm can't see: the up-and-coming strip nearby, the reason this street is different from the one next to it, the detail a buyer would only get from someone who actually works this patch. Platforms like PropertyVideos.ai make it fast to turn your listing photos into a branded video, which means there's no excuse for that local knowledge to stay locked in your head when you could be putting it in front of every buyer who views the listing.
Turning What You Know Into Something Clients Can See
Write down the "up and coming" detail for every current listing. One sentence. What's changing nearby that a data model wouldn't catch yet.
Say it in your video, not just your head. Add it to the voiceover script or your own commentary on the listing video, not buried in the listing description nobody reads fully.
Repeat it across channels. The same detail works in your Instagram caption, your buyer follow-up email, and your open home conversation. Consistency builds the reputation.
Track which details land. If mentioning the new zoning change gets more enquiries than mentioning the renovated kitchen, you've found your edge for that suburb. Keep using it.
What Agents Ask About This
Isn't this just marketing spin?
No. Spin is saying something that isn't true to sound good. This is the opposite: saying something true that happens to be invisible to any tool your client might have already consulted. The advantage only works if the detail is real and specific.
What if I don't actually have insider knowledge on a suburb?
You have more than you think. Every conversation at an open home, every chat with a local tradesperson, every "why did that one sell so fast" moment is data an AI model doesn't have access to. The habit isn't gathering more of it. It's writing it down before it evaporates.
Doesn't AI already do local market reports?
It aggregates historical sales data. That's useful, but it's backward-looking by definition. What Lane pointed to on CNBC is forward-looking knowledge, the stuff that hasn't sold yet, which by nature can't show up in a dataset.
The Gap Won't Stay Open Forever
AI will keep getting better at pattern-matching historical data. It will not get better at knowing something before it happens. That's not a technical limitation that improves with the next model release. It's a structural one.
The agents who win the next few years won't be the ones who ignore AI or the ones who over-rely on it. They'll be the ones who take the one thing AI genuinely can't do, and make sure their clients actually see it. Right now, most of that knowledge is still sitting in your head instead of your marketing.
That's an easy fix. It's also the whole edge.