Lead Nurture Sequences with ChatGPT

Most agents have a database of two thousand contacts and a follow-up plan that consists of "I'll call them when I think of them." That is not a system. AI lets you build the system in an afternoon.

The principle: segmentation by intent, cadence matched to intent, content that earns trust before it asks for anything. Buyers in motion get weekly touches with new listings and education. Future sellers get monthly touches with market data and case studies. Renters thinking about owning get bi-weekly touches built around financial readiness. Three sequences. Twelve touches each. Thirty-six pieces of content. AI writes the first draft of all of them in under an hour.

The Master Nurture Prompt

Here is the structure. Run it three times — once per segment — varying the lead-type and cadence sections.

You are a senior real estate agent in [Ames, IA].
You write nurture emails that sound human, give
real value, and never beg for the appointment.

LEAD SEGMENT
Active buyer, 90-180 days from purchase,
pre-approved at $400-475k, prefers single-family
3+ bed in [school district zone].

CADENCE
Twelve weekly touches. Mix:
- 4 listing-feature touches (new active comps)
- 3 education touches (financing, inspection,
  closing-cost expectations)
- 3 market-data touches (median price, DOM,
  inventory months — local zip-level)
- 2 soft-ask touches (offer to tour, offer
  to refine search criteria)

CONSTRAINTS
- Each email: subject line (under 55 chars),
  preview text (under 90 chars), body
  (140-220 words), single CTA.
- No "Hi [First Name]!" exclamation-point energy.
- No real-estate cliches.
- No phrases that imply preferences about who
  should live somewhere ("perfect for families,"
  "great for kids," "active adult community,"
  "exclusive neighborhood"). Property facts only.
- Comply with CAN-SPAM: include a sentence at
  the bottom about how the recipient can opt out.
- Comply with Fair Housing Act 42 U.S.C. §3604.

OUTPUT
Numbered email 1 through 12. Each email labeled
with its purpose (listing / education / market
data / ask). End with a one-line note on what
behavior triggers move to the next segment
(reply, click, tour request).

Now run that for each segment. For the future-seller segment, change the cadence to monthly, swap listing features for "just-sold in your zip" touches, and lean heavier on equity and timing content. For renter-to-owner, run bi-weekly with first-time-buyer financing content and rent-vs-own math.

Mixing Voice Back In

The first draft will be ninety percent of the way there. The last ten percent is voice. Pick three of the twelve emails — usually the ones that introduce you, the one with the strongest market take, and the one that asks for a tour — and rewrite them in your voice. Then paste your rewrites back into the model and ask:

Below are three emails I rewrote in my voice.
Re-write the other nine to match this tone,
keeping content unchanged.

That single move turns a generic sequence into one that sounds like you wrote every line. It is the single biggest difference between AI-assisted nurture that works and AI-assisted nurture that gets unsubscribed.

Fair Housing Act & NAR Article 10 — Nurture-Side Pitfalls

Specific pitfalls AI will create unless you constrain it:

• Familial-status targeting language: "Since the kids are getting older, you might be thinking about a bigger place." If the lead never mentioned kids, you cannot infer them. Strip phrases like "for your family," "as a parent," "perfect for kids."
• Religious holiday hooks: Easter, Christmas, Ramadan, Diwali openers segment your audience by religion in real time. Use seasonal hooks tied to weather, school calendars, or market cycles instead.
• Steering by neighborhood description: "I think you'll feel right at home in [neighborhood] — it has a great community feel." That sentence is steering bait. Describe the property and the data; let the lead choose the area.
• CAN-SPAM violations: 15 U.S.C. §7704 requires a working unsubscribe link, a physical postal address, and accurate header info. AI will not put any of these in unless you tell it to. Add the requirement to every nurture prompt.
• TCPA on the SMS variant: If you adapt the sequence to texts, the Telephone Consumer Protection Act (47 U.S.C. §227) requires prior express written consent. Document it before the first text leaves your phone.

Add this guardrail clause to every nurture prompt:

Do not infer protected-class membership of the
lead. Do not use family, religious, racial,
national-origin, or disability framing in any
email. Comply with 42 U.S.C. §3604(c), NAR Code
of Ethics Article 10 SOP 10-3, and CAN-SPAM
15 U.S.C. §7704. Include a working unsubscribe
sentence and my brokerage's physical address
in every email's signature block.