ChatGPT Prompts for Real Estate Agents: Fair-Housing-Safe Listings, Follow-Ups, and Client Touches That Actually Get Sent

Every agent knows the writing backlog: the listing description that's due before photos go live, the follow-up to Saturday's open house sign-in sheet that never happens, the price-reduction conversation you rehearse in the car, and the past-client check-in that would have generated a referral if it had actually gone out. It's high-frequency, formulaic writing — exactly what ChatGPT handles well. It's also writing where one lazy phrase ("perfect for young families," "safe neighborhood," "walking distance to the church") walks you straight into Fair Housing territory.

That's what most "AI for realtors" content skips: the prompts aren't the hard part. The hard part is building the guardrails into the prompt itself, so every draft comes out describing the property instead of the buyer, and so no client's personal details ever land in a public AI tool. This article gives you both: the baseline rules, and 12 ready-to-paste prompts covering listings, follow-ups, open houses, price conversations, and the referral touches that compound over years.

TL;DR

ChatGPT can draft a listing description, a post-showing follow-up, an open house invite, or a price-reduction email in seconds — but only safely if two rules hold every time. First: describe the property, never the people. Fair Housing law prohibits steering and any language that signals a preference about who should live somewhere — so the prompts below are built to focus on features, condition, and location facts, with placeholders like [NEIGHBORHOOD FEATURE] that stay factual and property-focused. Second: client-identifying details never enter the prompt. Names, addresses tied to a person, financial situations, and family circumstances stay out; draft with [BUYER], [SELLER], and [PROPERTY ADDRESS] placeholders, then personalize inside your CRM — never inside the AI chat window. Below are 12 prompts organized by where an agent's week actually leaks time, plus the one-line rule worth memorizing: AI drafts the words; the agent owns every fact, every number, and every Fair Housing call.

The Fair Housing baseline, before a single prompt

Three rules govern everything below. They're not optional extras — they're the difference between a productivity tool and a compliance incident.

Market the property, not the neighbor. The Fair Housing Act prohibits discrimination based on race, color, religion, sex, disability, familial status, and national origin — and many states add more protected classes. That means no drafts describing an "ideal buyer," no demographic language ("great for young professionals," "family-friendly block"), and no proximity references that signal religion or ethnicity. Every prompt below is written to describe features, condition, and verifiable location facts. If a draft comes back with people-language, cut it before it ships.
Never paste client-identifying information into a public AI tool. Your clients' names, finances, motivations, relocation reasons, and pre-approval details are confidential. Draft generic with [BUYER]-style placeholders; fill in real specifics only after the draft leaves the AI tool, inside your CRM or email client.
You own every fact. AI will happily invent square footage, school ratings, HOA terms, and market statistics. Give it the facts from your listing sheet and MLS data, instruct it to use only what you paste, and verify every number before publication. A wrong fact in a listing is your misrepresentation problem, not ChatGPT's.

With those pinned above your desk, here are 12 prompts for the writing that eats an agent's week.

Listing descriptions (3 prompts)

1. The core listing description

Write a [WORD COUNT, e.g., 200]-word listing description for [PROPERTY ADDRESS], a [BEDS]-bed, [BATHS]-bath [PROPERTY TYPE] listed at [PRICE]. Use only these facts — do not add or embellish any detail: [PASTE FEATURE LIST FROM LISTING SHEET]. Lead with the strongest feature, describe the property and its condition only — no language about what kind of buyer or household it suits, no demographic or lifestyle assumptions. Close with a neutral invitation to schedule a showing.

2. The feature-to-benefit rewrite

Rewrite these raw property features as benefit-focused sentences for a listing: [PASTE FEATURES, e.g., "new HVAC 2024, south-facing windows, 0.4-acre lot"]. Keep every claim tied to the physical property — light, storage, maintenance, layout, energy costs. No assumptions about who lives there or how they live. Return as short sentences I can mix into a description.

3. The listing headline options

Write 8 headline options (under 65 characters each) for a listing at [PROPERTY ADDRESS]: [ONE-LINE PROPERTY SUMMARY]. Focus each on a concrete feature or on [NEIGHBORHOOD FEATURE, e.g., "two blocks from the riverfront trail"] — factual location references only, nothing about the type of person who should buy. No all-caps, no exclamation marks.

Buyer and seller follow-ups (3 prompts)

4. The post-showing buyer follow-up

Write a follow-up email to [BUYER] after their showing of [PROPERTY ADDRESS] on [DATE]. One line thanking them, one line inviting their honest reaction, and an offer to answer questions or line up similar properties based on the criteria they've given me: [CRITERIA, e.g., "3+ beds, under $450K, large yard"]. No pressure language. Under 100 words.

5. The seller weekly update

Write a weekly update email to [SELLER] about their listing at [PROPERTY ADDRESS]. Report only these facts: [SHOWINGS COUNT] showings this week, [ONLINE VIEWS] online views, and this feedback summary: [PASTE FEEDBACK NOTES]. One line on next week's plan: [PLAN, e.g., "refreshed photos, one open house"]. Honest, calm tone — no spin, no invented market commentary. Under 150 words.

6. The cold-lead revival

Write a re-engagement email to [BUYER], who I last spoke with on [DATE] about buying in [AREA]. One line acknowledging time has passed with zero guilt, one genuinely useful observation from my own notes: [PASTE YOUR OBSERVATION, e.g., "three new listings matching their criteria this month"], and one low-commitment next step (reply, or a 10-minute call). Do not invent market data. Under 90 words.

Open houses and price conversations (3 prompts)

7. The open house invitation

Write an open house invitation for [PROPERTY ADDRESS], [DATE] from [START] to [END]. Include the property basics — [BEDS] beds, [BATHS] baths, [PRICE] — one standout feature: [FEATURE], and one factual location note: [NEIGHBORHOOD FEATURE, e.g., "around the corner from the farmers market"]. Invite everyone; no language about who the home is "perfect for." Give me two versions: a 60-word email and a 25-word social caption.

8. The open house attendee follow-up

Write a next-day follow-up email template for open house attendees who signed in at [PROPERTY ADDRESS]. Thank them for coming, offer the listing sheet and disclosures on request, ask one open question about what they're looking for, and mention I'm happy to help whether or not this home was the fit. Placeholder [ATTENDEE] for the name. Under 90 words.

9. The price-reduction conversation opener

Write an email to [SELLER] preparing them for a price conversation about [PROPERTY ADDRESS], on the market [DAYS] days with [SHOWINGS COUNT] showings and no offers. Use only the comparable data I paste here: [PASTE YOUR COMPS AND FEEDBACK]. Frame it as reading the market's response together, propose a call on [DATE/TIME OPTIONS], and take no position on a specific new price — that's for the call. Empathetic, direct, no blame. Under 160 words.

Past clients, referrals and reviews (3 prompts)

10. The home-anniversary touch

Write a one-year home anniversary note to [PAST CLIENT], who closed on [MONTH/YEAR]. Warm, short, zero sales pitch — one congratulations line, one line that I'm always available for anything home-related (contractor referrals, questions, a no-obligation value check if they're ever curious). Under 70 words.

11. The referral request

Write a referral note to [PAST CLIENT] a few months after their closing. One line referencing that I enjoyed working with them (no transaction details), one line explaining that my business grows through introductions, and a specific, low-friction ask: if someone they know mentions buying or selling, a simple intro by text or email is the best compliment. No incentives, no pressure. Under 80 words.

12. The review request

Write a short post-closing review request for [PAST CLIENT]: one line thanking them for their trust, one line noting that reviews help others choose an agent with confidence, and the placeholder [REVIEW LINK]. No incentives, no scripting of what the review should say, no transaction details. Under 60 words.

Save yourself the setup

The 12 prompts above cover the highest-frequency writing in an agent's week — a fraction of the full workload. If you'd rather skip straight to a tested, copy-paste library, that's what the 100 ChatGPT Prompts for Real Estate Agents pack is built for: 100 placeholder-first prompts across listing descriptions, buyer and seller follow-ups, open houses, price and negotiation conversations, farming and sphere touches, social content, and past-client referral systems — all written property-first with Fair Housing guardrails baked into the prompt language. Get it for $19.99 at thedailystackstudio.gumroad.com/l/gtfnu.

Frequently Asked Questions

1. Can ChatGPT write Fair Housing-compliant listing descriptions? It can produce compliant drafts if you constrain it — instruct it to describe only the property and verifiable location facts, and to avoid any language about who should live there. But compliance responsibility is yours: review every draft for steering language, demographic signals, and lifestyle assumptions before publishing. AI is a drafting tool, not a compliance officer.

2. Is it safe to put client information into ChatGPT? No. Treat public AI chatbots as non-confidential. Client names, finances, motivations, and personal circumstances stay out. The safe workflow is placeholder-first: draft the generic message with [BUYER]-style placeholders, then personalize inside your CRM or email client.

3. Will AI-generated listing descriptions hurt my listings? The risk isn't AI itself — it's unverified facts and generic sameness. Feed the prompt your actual listing sheet, instruct it to use only those facts, verify every number, and edit in your own voice. A description built on your real data reads better than most human-written boilerplate.

4. Can unlicensed assistants send these AI-drafted messages? Templates for scheduling and logistics are generally fine for staff to send after a broker-approved template review. Anything touching pricing opinions, negotiation, contract terms, or licensed activity stays with the agent — check your state's rules on what unlicensed assistants may communicate.

5. What about advertising rules beyond Fair Housing? Most states and MLSs add their own advertising requirements — brokerage identification, team name rules, disclosure language. These prompts handle the message body; your required disclosures and brokerage branding get added per your broker's policy before anything publishes.

Conclusion

None of this replaces your market knowledge, your negotiation, or your broker's compliance review — it replaces the blank cursor between you and the follow-up that should have gone out Tuesday. Draft with placeholders, keep client details in your CRM, describe properties instead of people, and verify every fact before it ships. Start with the three follow-up prompts above this week — consistent follow-up is where most deals are actually won. If they earn their place, the full 100 ChatGPT Prompts for Real Estate Agents pack ($19.99, thedailystackstudio.gumroad.com/l/gtfnu) turns it into a system for your whole pipeline, Fair Housing-aware from the first draft to the last.