AI Appointment Setter for Real Estate Agents: Setup and Scripts

For real estate, an AI appointment setter answers every buyer and seller lead within seconds, by text, DM or email, asks the questions an agent would (area, budget, timeline, pre-approval, selling first?), and books showings, buyer consultations or listing appointments on the agent's calendar. It's built for the reality of real estate leads: lots of them, most not ready yet, and the fastest agent usually winning the ones who are.

I've looked at plenty of real estate lead flows while building ReplySetter. The pattern is always the same: great lead generation, slow follow-up. Here's how I'd fix it.

The real estate lead problem

Real estate agents often pay a lot per lead from portals and ads. Then:

• leads arrive while the agent is at a showing or a closing,
• the agent replies hours later, after another agent already did,
• most leads are 3–12 months from buying, and nobody nurtures them,
• the CRM fills with untouched contacts.

Speed and follow-up are the whole game. See what is speed to lead.

What an AI setter does for agents

• Instant replies to portal, Facebook ad, website and text leads.
• Buyer qualification: areas, budget, property type, timeline, pre-approval status, first-time buyer or not.
• Seller qualification: address, reason for selling, timeline, whether they're also buying.
• Booking showings, buyer consultations or listing appointments.
• Long-term nurture for leads who aren't ready, with follow-ups timed to what they said.
• CRM updates: fields, tags and pipeline stage for every lead.
• Handoff to the agent for hot leads and anything legal or contractual.

Sample buyer flow

Step 1, Greet and confirm interest. "Hi Jordan, thanks for asking about 14 Oak Street! Are you looking to buy in the next few months, or just starting to look around?"

Step 2, Qualify. Objectives: target areas, budget range, timeline, pre-approved yes/no, must-haves. One question at a time.

Step 3, Route. Switch on timeline and pre-approval:

• Ready within 3 months and pre-approved → book a showing or buyer consultation.
• Ready but not pre-approved → offer a lender introduction, then book.
• 6+ months out → nurture sequence, tag accordingly.

Step 4, Book. Offer two times for a showing or call, book, confirm with the address.

Sample seller flow

Step 1. "Thanks for requesting a home value estimate for 22 Pine Road! Are you thinking of selling soon, or mainly curious about the value?"

Step 2, Qualify. Objectives: timeline, reason for moving, whether they need to buy next, any repairs or updates.

Step 3, Book a listing appointment for sellers within 6 months, or a valuation call for others.

Scripts that work in real estate

Portal lead:

Hi Jordan, it's Sam's assistant from [brokerage]. Saw you asked about 14 Oak Street. It's still available! Would you like to see it this week, or do you have a few questions first?

Not ready yet:

Totally makes sense to start early. Want me to send you new listings in [area] under [budget] as they come up? And I'll check in around [month] when you're closer to moving.

Pre-approval:

Have you had a chance to talk to a lender yet? If not, happy to connect you with one we trust, it makes things much easier when you find the right place.

Compliance notes for real estate

Not legal advice, but worth keeping in mind:

• Fair housing. The AI must never steer buyers based on protected characteristics (race, religion, family status, disability and so on), or describe neighbourhoods in those terms. Put this explicitly in the instructions and test it.
• Texting consent. In the US, the TCPA and A2P 10DLC rules apply to texting leads. Only text people who gave consent.
• No advice it shouldn't give. Legal, tax and mortgage questions go to the professional.
• Disclosure. Be honest that it's the agent's assistant. See should you tell leads they're talking to AI.

How I'd set it up for a real estate team

With ReplySetter:

1. Connect your channels. Most teams run SMS, Facebook and website chat through GoHighLevel; email leads through Gmail or Outlook.
2. Knowledge base: service areas, your process, buyer and seller FAQs, current listings or a link to them, and your team's bios.
3. Two flows, buyer and seller, or one flow with a switch at the start.
4. Calendars: each agent's calendar, with the AI booking the right one based on area or lead type.
5. Scenarios for fair housing, legal questions and "I want to speak to the agent".
6. Test with real past leads in the sandbox.
7. Go live in suggest mode, then auto.

Teams that already use a custom property search API can connect it as a custom tool so the AI can look up listings matching what the buyer said.