How can real estate agents use AI to answer hold vs. sell vs. 1031 exchange questions for clients?
Use a 3-step AI workflow — Paste, Push Back, Layer In. Paste the client’s question into Claude along with the property’s MLS history and a current comp set. Ask Claude to push back on the framing of the question. Then layer in the local context Claude can’t see. The whole thing takes about 12 minutes, and it almost always points to one clear answer.
One of my clients almost cashed out $400,000 in equity last week.
He texted me three options. Sell the rental, do a 1031 exchange into a new-build, or hold it another five years. He wanted my take by the end of the day.
Twelve minutes later I sent him a real answer. Not a gut take. Not a “let’s set up a call to chat through it.” A defensible recommendation backed by his actual numbers, his actual market, and the math he hadn’t run yet.
He kept the rental. It was the right call. And the whole thing came out of a workflow you can run on any client question Monday morning.
Here’s exactly what I did and the 3-step process you can steal.
Step 1: Paste — and don’t skip the MLS exports
The first move is to paste the client’s actual question into Claude. Don’t paraphrase it. Don’t clean it up. Drop in the email or the voicemail transcript exactly the way it came to you.
But that’s not the part most agents skip.
The part most agents skip is uploading the two MLS exports that turn this from a hot take into a defensible answer:
- The property’s own history — original purchase price, any price changes, current loan info, and the rent the client is collecting.
- A current comp set — what’s actively listed, what’s pending, and what’s sold in the same neighborhood in the last 90 days, broken out by similar bed/bath count.
Without those two files, Claude is just a smart friend with opinions. With them, it’s a decision-grade analyst that can actually do the math on your client’s specific situation.
This is the same data-first habit I teach in 3 Claude AI Skills Every Busy Realtor Should Run This Week. Real numbers in, real answers out.
Watch Kristi walk through the exact paste at 2:35.
Step 2: Push Back — Claude’s job is to challenge the question
Here’s where most agents get stuck. They paste the client’s three options and ask Claude to pick one. So Claude picks one. But it’s picking from the wrong menu.
The better move is to ask Claude to push back on the framing of the question first.
In this case, Claude came back with the answer I didn’t expect — “you’re asking the wrong question.” The client wasn’t trying to choose between three properties. He was trying to decide what to do with $400,000 of equity. That’s a completely different question, and it changes the whole analysis.
This is what good coaching feels like — and it’s what makes the AI usable for client work instead of just for content. You’re not outsourcing your judgment. You’re sharpening it.
The Push Back step is also where the “cash flow is the wrong metric” reframe comes out. Most investors lead with cash flow because it’s the number they can see on a bank statement. Claude will steer you toward the metric that actually matters: the blended return on the equity sitting inside the property.
When you add up appreciation, principal paydown, the tax shelter from depreciation, and the rent — and you divide that total return by the equity at stake — most low-rate rentals quietly print 20%+ a year. You cannot beat that by selling the property and parking the money somewhere else.
That’s the “rate as an asset” reframe in plain English. A 3% loan on a rental isn’t a liability. It’s an asset that’s quietly generating return on every dollar of equity it’s protecting. And it’s the single biggest reason most “should I sell my rental” conversations end with “hold it.”
Here’s another reframe Claude surfaced on the push-back step: the 1031-into-new-build trap. A 1031 exchange into a brand-new rental sounds like an upgrade. New construction, fresh tenants, less maintenance. But the new property comes with today’s rate, today’s basis, and today’s tax picture. You can wipe out the entire equity advantage of the old property in a single transaction without realizing it.
Watch Kristi explain the push back at 4:05.
If you want to put AI to work in your marketing without the tech overwhelm, check out the AI Marketing Academy (AIM) with Jason Pantana. It’s built for real estate agents who want practical results — not a PhD in machine learning. Learn more at kristijencks.com.
Step 3: Layer In — the part Claude can’t see
Even with the MLS exports, Claude is missing the layer that makes you valuable as the agent. Claude doesn’t know that the high school zoning is shifting next year. Claude doesn’t know that three new builds just hit the market down the street. Claude doesn’t know that the HOA is about to vote on a rental cap.
That’s the layer you bring in.
After Claude runs the math, take 60 seconds and add the local context out loud. “Here’s what I’m seeing in the neighborhood that the numbers don’t show.” Then ask Claude to weight that into the recommendation.
This is where the answer gets sharp. Tom Ferry-style — three options on the table, one clear winner, and a one-line rationale you can text back to the client.
In my client’s case, layering in two specific local data points — a backlog of competing new-builds and a softening rent trend at the higher price point — confirmed the hold. The blended return on the existing property was already crushing what the 1031 target could deliver. Adding the local picture just made it more obvious.
He kept the rental, kept the rate, kept the equity, and stopped second-guessing himself by lunch.
Why this matters more than another marketing post
Most of the AI conversations in real estate right now are about content. Caption generators. Headline writers. Reel hooks. Those are all useful — but they’re table stakes.
The leverage is on the advisory side. Your clients have real questions about real money, and most of them aren’t getting a real answer from anyone. Their CPA punts to their advisor. Their advisor punts to their agent. Their agent shrugs and offers to run comps.
An hour with this workflow puts you in a different tier. You’re the only person in their life who can take the email, run the math, push back on the framing, and hand them a decision they can act on the same day.
That’s the same shift I teach in How to Use AI to Find Hidden Leads in Your CRM — and it’s the same reason your database isn’t a list, it’s an ecosystem. Every past client has a question like this sitting in their inbox right now. Most of them just need someone with the math and the nerve to answer it.
The 3-step process, one more time
If you remember nothing else, remember this:
- Paste. The client’s question and the two MLS exports — the property’s history and the live comp set.
- Push Back. Ask Claude to challenge the framing. Most client questions are the wrong question. The reframe is where the answer hides.
- Layer In. The neighborhood-level context only you have. Then ask Claude to factor it in and recommend.
That’s the whole thing. Twelve minutes start to finish. Done is better than perfect, and a defensible answer the same day beats a polished one next week.
Pick a client question sitting in your inbox right now and run it. Then come back and tell me what Claude pushed back on — that’s almost always the part that changes the deal.
For more workflows like this one, more episodes of Take It and Run, and the systems I’m using with the agents I coach, head over to kristijencks.com and join the email list. The next playbook drops there first.
About Kristi Jencks
Kristi Jencks is a real estate coach, speaker, and trainer who helps agents, team leaders, and brokers stop overthinking and start taking action — with practical strategies that are easy to implement and built for the real world. As a Senior Coach in the Tom Ferry ecosystem and host of the Take It and Run podcast, she specializes in turning everyday agents into consistent, confident closers through systems that actually work. Kristi is based in Gilbert, Arizona, where she’s also an active agent in the East Valley market.
Leave a Reply