I think we're measuring AI completely wrong.
Everyone talks about it the same way. How many people can it replace. How much work can it eliminate. Both of those are subtraction, and both are easy to put in a slide.
In real estate, the better question is how much humanity can it give back.
What does an agent's day actually look like?
Chasing signatures. Finding documents. Remembering dates. Filling out forms. Working out what comes next.
Nobody got into this business for any of that. It's the work that piles up around the work, and it's the same pile that stands between a finished deal and getting paid.
What should they be doing instead?
Negotiating. Reassuring. Building trust.
That's the stuff humans are actually really good at, and it's the part a client remembers. Nobody ever chose an agent because the paperwork was tidy. They chose the person who was present and paying attention when the deal got hard.
That's the whole product. It's also the first thing to go when the day fills up with everything else, because presence is what gets sacrificed when you're behind.
So why does the replacement framing stick?
Because it's easy to count.
Headcount is a number. Hours saved is a number. "The agent was fully present for the call that decided the deal" isn't, so it doesn't make the slide, and the thing that doesn't make the slide stops getting measured.
That's how an industry ends up optimizing for the wrong half of the job.
What's the better measure?
How much of the day went back to the parts only a person can do.
Not how much got removed. What got returned. Those sound like the same measurement and they are not, because one of them counts what an agent stopped doing and the other counts what they got to do instead. If an agent ends the week having spent more of it with clients and less of it hunting for a document, that's the number that matters, and it happens to be the one nobody reports.
The point of AI here isn't to remove a human from the transaction. It's to remove everything that keeps the human from being human.
Quick takes
How should real estate measure the value of AI?
By how much of the day it returns to the work only a person can do, not by how many people it replaces. Judd Hoffman argues headcount and hours saved are easy to count, which is why they crowd out the measure that matters.
What work should a real estate agent be doing instead of paperwork?
Negotiating, reassuring and building trust. Chasing signatures, finding documents, remembering dates and filling out forms are the work that piles up around the job rather than the job itself.
Is the point of AI in real estate to replace the agent?
No. Judd Hoffman's framing is the opposite: the point is not to remove a human from the transaction, it's to remove everything that keeps the human from being human.
Sources
- Nobody Gets Paid Until the File Is Complete: The same pile of admin work, seen from the end of the deal where it delays the check.
- Organization Is the Service: On presence being the thing a client actually notices, and the first thing lost when the day fills up.
Transcript
118 words
I think we're measuring AI completely wrong. Everyone talks about AI in terms of how many people it can replace or how much work it can eliminate. But in real estate, in my opinion, I think the better question is how much humanity can it give back? I see what agents do. They chase signatures, they find docs, they have to remember dates, they fill out forms, and they have to figure out what comes next. They should be negotiating, reassuring, building trust. The stuff that humans are actually really good at. The point of AI isn't to remove a human from the transaction at all. Maybe it's to remove everything that keeps the human from being human.
