Best AI Prompts for Realtors

AI summary

Seven AI prompts for real estate agents that respect Fair Housing rules and avoid the “don’t miss this stunning opportunity” defaults: listing drafts, offer cover letters, buyer consultations, CMA narratives, open house follow-ups, inspection triage, and niche newsletters. Built to amplify the agent’s local knowledge, not paper over it.

Prompt 1

Listing Description Drafter

Most MLS descriptions sound the same. This prompt produces copy that pulls a buyer through the listing without resorting to “don’t miss this rare opportunity.”

I am writing the listing description for:

ADDRESS: [ADDRESS]
LIST PRICE: [PRICE]
BEDS / BATHS / SQ FT: [SPECS]
KEY FEATURES (be specific, not generic):

[LIST: school district, recent renovations, lot size, special architecture, view, etc.]

NEIGHBORHOOD VIBE in one sentence: [VIBE]
TARGET BUYER (be direct about who would actually buy this): [BUYER PROFILE]

Draft a listing description with:

1. OPENING: one sentence that gives the buyer a reason to keep reading. Not a tagline.
2. ROOM-BY-ROOM HIGHLIGHTS: 3-5 sentences walking through the home in the order a buyer would experience it.
3. THE STANDOUT: the one feature that makes this listing different from the others on the block.
4. LOCATION CONTEXT: one paragraph on what living here actually looks like (walkable to X, commute to Y).
5. CLOSING NUDGE: one sentence that invites a private showing without sounding pushy.

Do not use the following banned phrases: "don't miss", "stunning", "breathtaking", "must-see", "opportunity of a lifetime", "hidden gem", "a true entertainer's delight." Show the reader why; do not tell them how to feel.

Keep it under 300 words. Fair Housing compliant: no demographic preferences, no language steering to specific groups.

When to use: After you have walked the house twice and have the specifics. · Best model: Claude (most disciplined about banned-phrase rules) or ChatGPT.

Prompt 2

Offer Cover Letter Drafter

Some sellers still read the cover letter. This prompt drafts one that is specific and Fair Housing compliant.

I am submitting an offer on behalf of my buyer:

BUYER NAME: [NAME]
BUYER STAGE OF LIFE (factual, not demographic): [e.g., first-time buyer, relocating for new job, downsizing after kids moved out]
WHY THEY WANT THIS SPECIFIC HOUSE: [THE NON-PRICE REASONS]
OFFER TERMS (price, contingencies, closing timeline): [TERMS]

Draft a cover letter from me as the listing buyer's agent that:

1. Opens with one specific thing about the property the buyer noticed and loved.
2. Gives the seller a sense of who the buyer is as a HOUSEHOLD (not a demographic).
3. States the offer terms clearly in one paragraph.
4. Closes professionally.

FAIR HOUSING COMPLIANCE IS NON-NEGOTIABLE. Do not include:

- Race, ethnicity, national origin, religion of the buyer.
- Family status (number of kids, plans for kids, etc.).
- Disability status.
- Marital status.
- Sexual orientation or gender identity.
- Anything that could be construed as discriminatory.

If I have provided any prohibited information in my prompt, omit it and flag it in your output. Keep the letter under 200 words. Some MLS rules forbid offer letters; if my prompt suggests I should check that first, say so.

When to use: Same day you submit the offer. · Best model: Claude. The Fair Housing discipline is non-negotiable and Claude is most reliable about flagging issues.

Prompt 3

Buyer Consultation Prep

First-time-buyer consultations either go great or burn an hour. This prompt structures the conversation so you cover what matters.

I have a buyer consultation tomorrow with:

NAME: [NAMES]
WHERE THEY ARE IN THE PROCESS: [first conversation / been looking 3 months / specific timeline]
THEIR STATED PRICE RANGE: [RANGE]
WHAT THEY HAVE TOLD ME SO FAR ABOUT WHAT THEY WANT:

[BULLET POINTS FROM PRIOR CONVERSATIONS OR INTAKE FORMS]

Draft a consultation outline with:

1. THE FIRST 5 MINUTES: warm-up questions that get them talking about their day-to-day life, not their dream home.
2. THE BUDGET CONVERSATION: 3 specific questions that surface whether their stated range is realistic (without making them feel judged).
3. THE NEEDS vs WANTS SORT: the 4-5 questions I should ask to separate must-haves from nice-to-haves.
4. THE MARKET REALITY: 2-3 facts about the current local market they should know up-front (frame as one-liners I can deliver naturally).
5. THE NEXT STEPS: what I should leave them with so we know what is happening this week, next week, and the week after.
6. ONE QUESTION I AM AFRAID TO ASK: based on their profile, the question I might be avoiding that I should actually raise.

Do not assume a household structure. Do not stereotype based on the demographic clues in my prompt.

When to use: Night before the consultation. · Best model: Claude is well-suited because of the discipline about not stereotyping.

Prompt 4

Comparative Market Analysis Narrative

Most CMA conversations are about the spreadsheet, not the story. This prompt produces the narrative that helps a seller actually internalize the comps.

I have a CMA presentation to make:

SUBJECT PROPERTY: [ADDRESS, SPECS]
LIST PRICE THE SELLER IS HOPING FOR: [PRICE]
COMPS (3-5 recent sales):

[FOR EACH: address, sale price, days on market, key differences from subject]

WHAT THE COMPS SUGGEST FAIR VALUE WOULD BE: [RANGE]

Draft a 5-minute narrative I can deliver to the seller that:

1. Opens by acknowledging this is their home, not a commodity.
2. Walks through one comp at a time, naming the specific feature that made it more or less than the subject.
3. Lands on a price range with the reasoning earned by the comps, not just stated.
4. Addresses the most likely objection they will raise ("but my house is special because X").
5. Reframes the conversation around days-on-market, not just price, if their target price would slow the sale.
6. Closes with my recommended list price and ONE concrete action they can take that affects value (specific staging, repair, or pricing strategy).

Do not condescend. The seller knows their house better than I do. Make the comps the authority, not me.

When to use: Day before the listing presentation. · Best model: Claude. Tone discipline matters when delivering hard pricing news.

Prompt 5

Open House Follow-Up Sweep

You had 12 people through Sunday’s open house. Three are warm. The pattern is to write the same email to all 12. This prompt drafts personalized follow-ups in 10 minutes.

Here are the sign-in sheets and notes from yesterday's open house at [ADDRESS]:

[FOR EACH VISITOR: name, what they said they liked, what they said concerned them, current housing situation if mentioned]

Draft a personalized follow-up for each visitor:

1. THE OPENING: references the specific thing they said about the house.
2. ONE PIECE OF VALUE: a useful piece of info about the property, the neighborhood, or the market relevant to their stated concern.
3. THE NEXT STEP: invitation to a private showing, a CMA on their current home, or a market-update email subscription, calibrated to their warmth level.
4. SIGNATURE: my standard sign-off (you can use "[YOUR SIGN-OFF]" as the placeholder).

WARMTH RATING: rate each visitor as HOT / WARM / COOL based on what they said, and tailor the ask accordingly. Do not over-ask the cool ones; do not under-ask the hot ones.

Keep each email under 120 words.

When to use: Monday morning after the Sunday open house. · Best model: Claude or ChatGPT. Both handle this competently.

Prompt 6

Disclosure Issue Triage

An inspection just came back. There are 14 items. Some are deal-killers, most are not. This prompt sorts them so you can have the right conversation with both sides.

Here is the inspection report summary for [ADDRESS]:

[PASTE THE ITEMS LIST OR SUMMARY]

The transaction context:
LIST PRICE: [PRICE]
OFFER PRICE: [PRICE]
MY CLIENT IS THE: [BUYER / SELLER]
CLOSING DEADLINE: [DATE]
MARKET CONDITION (buyer / seller / balanced): [CURRENT MARKET]

Triage the items into:

1. DEAL-RELEVANT: items significant enough they should affect price, terms, or the deal itself.
2. NEGOTIATION LEVERS: items that are real but minor and become bargaining chips.
3. ROUTINE: items that are normal for a home of this age and condition.
4. REQUIRES SPECIALIST: items where a structural engineer, plumber, or other specialist evaluation is needed before we know what we are dealing with.

For each DEAL-RELEVANT item, draft one sentence I would say to my client about what it actually means, in plain English.

Do not provide legal advice. Do not provide a definitive cost estimate for repairs; flag any item that needs a real contractor bid.

When to use: Within 24 hours of receiving the inspection report. · Best model: Claude. The discipline about not over-claiming on cost estimates matters.

Prompt 7

Niche Market Newsletter Drafter

Most agent newsletters get unsubscribed in three sends. This prompt drafts a newsletter that the niche actually wants to read.

I send a monthly newsletter to my sphere. My niche / focus area:

[NICHE: e.g., first-time buyers in Austin, downsizers in Westchester, investor-buyers in Phoenix]

What is happening in the local market this month:

[KEY DATA POINTS: median price change, days on market, inventory level, mortgage rate movement, recent comparable sales worth mentioning, any new development or zoning change]

Draft a newsletter with:

1. SUBJECT LINE: 6-9 words that name the specific thing inside (no "Market Update" generics).
2. ONE LEAD INSIGHT: the single most important thing my niche should know this month, stated in 2-3 sentences with the specific number that proves it.
3. ONE PRACTICAL TIP: a concrete action my niche could take based on this month's market.
4. ONE LISTING OR STORY: a brief mention of a recent sale or new listing that illustrates the insight (no hard sell).
5. THE QUIET CTA: a soft invitation to reply with questions, not a "book a call now" push.

Keep it under 300 words. Tone: a knowledgeable friend who happens to be an agent, not a sales rep.

Do not write generic real-estate advice ("spring is a great time to sell"). Every claim must be tied to local data I gave you.

When to use: First Monday of every month. · Best model: Claude (most disciplined about the no-generic-advice rule).

Connectors are now standard

Claude, ChatGPT, and Grok all support connectors that let your AI read live data from your work tools (Gmail, Notion, GitHub, Asana, HubSpot, Stripe, and many more) instead of relying on you to paste context. For agents this means the AI can read your Gmail thread with a client or co-op agent, your Google Drive folder of inspection reports and disclosures, your Calendly bookings for showings, or your Notion deal tracker.

For real estate agents, the connectors worth pairing with these prompts:

Get Smarter About AI Every Morning

Free daily newsletter. Built for people who want to use AI well, not chase every model.

Free forever. Unsubscribe anytime.

Two ways to go further