AI Listing Presentation: How I Use Claude to Build Mine in Minutes

I get some version of this question every week: “How long does it actually take you to put together a listing presentation?” Agents assume it’s a Saturday afternoon of dragging photos into a template, adjusting fonts, and trying to remember what comps you pulled last week. So this is the honest answer: I built an AI listing presentation workflow in Claude that returns a full 15 to 17 slide deck in 5 to 10 minutes, and I want to walk you through exactly how it works on a real listing.

I’m William Zhang, a real estate agent in Austin, Texas, and the founder of Real Estate AI Society. Before I get into the tutorial, I want to be straight with you about what this video is and isn’t. This isn’t a “here’s how I won the listing” story. The property I used, 1117 Terrace View, was already under contract by the time I filmed this. It’s a follow-along tutorial showing you the actual build, not a highlight reel.

I get why the temptation is there to dress it up. A “this deck won me the deal” hook performs better than “here’s the tool I use.” But I’d rather show you the real process on a real address than manufacture a result that didn’t happen on camera. If this workflow is going to earn a spot in your business, it should earn it on what it actually does, not on a story I bolted onto it after the fact.

By the end of this post you’ll know why a listing presentation matters more than most agents treat it, what my AI assistant Betty actually does with one prompt, and how the deck walks a seller from your value proposition through live market data to the close. If you want the exact prompt instead of rebuilding this from scratch, it’s in the free Claude starter guide.

Why a Listing Presentation Still Matters

A listing presentation isn’t just a price opinion. Its job is to communicate the value you bring to the table in the room with the seller. That means covering what the home is worth and what you think you can sell it for, but it also has to cover your skills: do you have a real social media following you can market the home to, a newsletter or email list full of potential buyers, or years of experience in that specific neighborhood.

At the same time, a good presentation works both directions. It should give you a way to understand what the seller actually wants: their pain points, how they prefer to communicate, what they’re worried about. That’s what turns the meeting from a pitch into a real conversation, whether you’re doing it in person or over a Google Meet call.

Setting Up the AI Listing Presentation Prompt

Claude is a strong option for this specific job because it’s good at taking a task and working on it independently instead of needing you to babysit every step. I have my assistant, Betty, set up with a working understanding of my entire business: who I am, my production, my past deals, and she’s connected to both the MLS and my CRM, Follow Up Boss. That context is what makes a single prompt actually usable instead of returning something generic.

The prompt itself walks through the basics first: your name, your brokerage, the years and areas you serve, your contact information, and the property address. Then it asks for your unique selling point. For me, that’s a real YouTube following, people relocating to Austin who watch my videos before they ever call an agent, so I make sure that’s in the pitch: I’ll do a home tour video for their house too, and they can see exactly what that looks like before they hire me.

If you don’t have a channel, that’s fine. Maybe you’ve sold a lot of homes in that specific neighborhood, or you’ve lived in the city for over a decade, or you’re a brand new agent who’s simply a strong communicator and a hustler. Every agent needs a real unique value proposition in this document, not a placeholder. Set up your assistant correctly once, with the right format and enough real information, and it should populate all of this without asking you again.

I ran this for real on 1117 Terrace View. I gave it the prompt, stepped away for about 5 to 10 minutes, and came back to a finished deck.

What the AI Listing Presentation Actually Builds

Here’s the slide order Claude produced, and it’s worth knowing because it’s the structure I’d tell any agent to use.

It opens with the address and my name, then moves straight into an outcome-not-resume slide: net the most, the least friction, a clean and predictable close, full transparency throughout. That’s a deliberate choice. Sellers don’t care about your resume. They care about what happens to their sale.

From there it builds a slide on the five things beyond your final numbers that a seller should understand: price, preparation, marketing, negotiation, and management of the whole process. This is where you separate yourself from every other agent in the interview, because if the seller is talking to three or four agents, this is the slide that makes yours stand out.

Then it goes into live MLS market data. Betty is connected to the MLS, so she pulls actual current numbers: months of inventory, average days on market, current conditions. If you don’t have MLS access wired in, you can ask your assistant to pull from a source like Zillow instead so the slide still carries real numbers instead of a placeholder.

Right after that comes the comps slide: listings from the neighborhood that closed in the past 90 to 180 days. Because Betty is connected to the MLS, this pulls straight from there rather than me hunting it down manually. I’ve written more about wiring that connection up in how I use Claude to run my Follow Up Boss CRM.

Next is the marketing plan, and because Betty already knows my YouTube channels and media presence, she listed them out along with the plan to run Facebook and Meta ads to get the listing more exposure, at no cost to the seller. That’s a real value-add you can put in front of a seller in the room.

Then it walks through a day-by-day launch strategy, treating the listing like a product launch: day minus 5 to zero is prep and the photo shoot, day zero is coming soon, day one is the Friday launch, days two and three are the open house weekend with a social media push, and day seven is reviewing the traffic and feedback and adjusting from there. This matters more than it sounds like, because it walks the seller through what to expect emotionally, not just logistically. Sellers appreciate knowing the next step instead of sitting in the dark.

The deck closes by asking for the listing, with the listing agreement ready so you can close on the spot.

The Follow-Up Doesn’t Stop at the Presentation

Because of the same prompt, Claude also builds a recommended follow-up email to send right after the appointment, one that recaps exactly what you talked about in the room. And because Betty is connected to Follow Up Boss, the next step is a real follow-up sequence: emails at 1, 3, 5, and 10 days after the appointment to help you understand whether the seller is ready to move forward and sign.

That’s the part most listing presentation templates skip entirely. You get a deck, and then you’re on your own for what happens after the meeting. This workflow treats the presentation and the follow-up as one connected system, not two separate tasks.

If you want to see how a similar single-prompt build works on the front end of a listing, I covered the same connected setup in how I built a real estate listing website with Claude, which is the piece that usually comes together with this presentation for the same address.

What This Workflow Doesn’t Do

I want to be direct about the limits here too. Claude isn’t going to walk into the room and read the seller for you. It isn’t going to negotiate live or decide your price on its own; the market data and comps are there to support your judgment, not replace it. It also isn’t going to know your specific seller’s pain points until you tell it, so the first pass is a strong starting draft, not a finished script you read word for word. What it does is remove the hours of manual assembly, so you walk into the appointment with a document that would have taken you an afternoon, built in the time it takes to make coffee, and you spend the time you saved actually thinking about how to run the conversation.

If you want to build this exact setup for your own business, along with the assistant, MLS connection, and CRM follow-up behind it, that’s most of what I teach in the Real Estate AI Bootcamp, a live cohort where we build your AI operating system together over four weeks.

Get the Prompt

You can get the exact prompt I used on 1117 Terrace View for free through the Claude for real estate agent starter guide. Paste it into your own assistant, fill in your name, brokerage, and unique selling point, and you’ll have a real listing presentation to review in minutes instead of a Saturday. For the rest of the tools I run my practice on, my AI tools page has the full list.