Lead follow-up & inside sales

Upload five years of past closings to Claude, ChatGPT, or Gemini, rank each past client by likelihood of moving in the next year with a reason to reach out, check each address for active, pending, or sold status, then send five short personal emails

In Agent Intelligence issue 21 (February 7, 2026), Matt Spangler shares a sphere workflow from his AI Marketing System course. A student named Lorraine used AI to look through her past emails, found 7 people in her sphere who had mentioned buying or selling in the next couple of years, then ran a property search on her past clients and found that one past client had a pending sale with another agent. Step 1 has two options: connect Claude, ChatGPT, or Gemini to your email and search for anyone who mentioned buying, selling, moving, job changes, or life changes that trigger moves; or, if your brokerage does not allow an email connection, export your closings from the last 5 years, remove private information, and ask the AI to rank past clients by likelihood of a near-term move using ownership duration (5 to 7 years is prime), property type (condos turn over faster), life-stage signals, and purchase price against likely equity gain, with a personalized reason to reach out. Step 2: check every address on Zillow, Redfin, or your MLS and sort into active or pending (call anyway), recently sold (they still need an agent where they land), and off-market. Final step: have the AI draft five warm emails under 75 words that reference something specific and end with a soft question. Spangler suggests running the workflow monthly.

At a glance

Tools used
ClaudeChatGPTGeminiZillowRedfin
Category
Lead follow-up & inside sales
Audience
agent, team lead
Difficulty
Copy-paste
Source
How Real Estate Agents Use AI to Find $48k in GCI
Matt Spangler (Agent Intelligence) · Newsletter · Feb 7, 2026
First step
Export your closings from the last 5 years, delete phone numbers, emails, and other private fields, upload the closings file to Claude, ChatGPT, or Gemini, and run the move-likelihood ranking prompt.

Compliance note: Strip private client data before uploading, follow your brokerage's rules on connecting email to AI tools, and do not contact a past client whose home is listed with another agent about that listing; keep the note to the relationship.

Suggested prompt

Written by this directory, inspired by the source. Replace the [bracketed] parts.

Rank these past clients by likelihood of a near-term move based on: ownership duration (5–7 years is prime), property type (condos turn over faster), life stage signals (first-time buyers who may have outgrown the home, empty nesters), and purchase price vs. likely equity gain. Rank them by likelihood to move in the next year and provide a personalized reason to reach out.

Then, for the top 10, give me a table with name, property address, closing date, years owned, and a status column I will fill in after checking each address on Zillow, Redfin, or my MLS: Active/Pending, Recently Sold, or Off-market.

After I fill in the status column, write a short email to [NAME] who [bought/mentioned CONTEXT]. Keep it warm, not salesy. Reference something specific: the neighborhood, the market, what they told me. End with a soft question. Under 75 words. No subject line that sounds like marketing.

Share on X Read the source

← All use cases