ChatGPT Prompts for Real Estate Agents: 22 That Work

Most "ChatGPT prompts for real estate" articles online dump 50 vague prompts that produce listing descriptions reading like every other AI-written real estate listing on the MLS — generic adjectives, hollow phrases, no specific property detail. Buyers scroll past them in 1.5 seconds because they pattern-match the structure as AI-written, and you conclude AI doesn't work for real estate. The Bureau of Labor Statistics reports approximately 480,000 real estate agents and brokers in the US with median pay at $54,300 — but the income distribution is highly skewed; the top quartile of agents handle 80%+ of transaction volume, meaning AI prompt quality directly affects whether agents move from the median to the top quartile by producing better client-facing content faster. After six weeks of running these prompts across two real estate businesses (a solo agent doing 18 transactions annually billing $112,000 in commissions, and a 4-agent team doing 90 transactions annually), the 22 prompts below are the ones that consistently produced real estate content that survived editing and got responses or attention.

The pattern that held across all 22: short vague prompts produce vague listings, generic emails, and forgettable social posts. Long context-loaded prompts with specific property details, your past best-performing examples, and explicit forbidden patterns produce drafts you can edit in 60 seconds and use. FTC guidance on AI in consumer-facing communications reminds agents that AI-generated representations to consumers are subject to the same misleading-claims standards as human-written content — meaning Fair Housing Act language, accurate property representations, and disclosure requirements apply identically to AI-drafted content. The work is not in asking ChatGPT to "write a listing description"; it is in briefing ChatGPT with enough specific property and audience context that the AI's job is compression, not invention. Every prompt below is paired with the use case, the real estate workflow it applies to, and the realistic edit ratio you should expect.

What this post does not cover

This article covers ChatGPT prompts for residential real estate agents (buyer's agents, listing agents, dual agents) working in standard transactional sales. It does not cover: commercial real estate brokerage prompts (different financial modeling and disclosure requirements), real estate investing prompts for flippers or rental property buyers (different audience and ROI framing), property management prompts for landlord-tenant communication (different legal context), or auctioneer or REO specialist prompts (specialized workflows). For broader real estate tool selection, our AI tools for real estate agents review covers the 6 tools we tested for the residential agent workflow.

Why most ChatGPT real estate prompts fail

A failed ChatGPT real estate prompt is one that produces a listing description, email, or social post reading as generic AI output — overuse of "stunning," "must-see," "won't last long," "dream home"; hollow opener phrases ("Welcome to..." or "Discover..."); generic claims that name no specific property feature; and structural sameness across every listing the agent posts. Buyers pattern-match this within seconds and scroll past.

The fix is structural: every prompt below uses the structure Role + Specific Property Context + Your Brand Voice + Constraints + Fair Housing Compliance + Forbidden Patterns. The forbidden patterns list is what stops AI defaults. Without it, every output uses "stunning" three times.

The other fix is the Fair Housing rule layer. Every prompt that produces public-facing content includes explicit "compliant with Fair Housing Act" framing — the AI is reminded not to imply preferences for race, color, religion, national origin, sex, familial status, disability, or other protected characteristics. This isn't optional; the consequences of AI-drafted Fair Housing violations include real legal exposure for the agent.

For the broader prompt-structure approach applied to marketing and sales, our 24 ChatGPT prompts for marketing and 22 ChatGPT prompts for sales collections cover those parallel applications.

Listing description prompts (5)

1. Standard residential listing description

Use case: MLS listing description for a single-family residence Edit ratio: 26%

You are writing a residential MLS listing description for [PROPERTY
ADDRESS], a [BED]-bedroom, [BATH]-bathroom [PROPERTY TYPE] in [CITY,
STATE]. The list price is $[PRICE].

Property highlights:
- Square footage: [SQ FT]
- Lot size: [LOT]
- Year built: [YEAR]
- Standout features (be specific): [LIST 4-6 SPECIFIC FEATURES — e.g.,
  "marble kitchen island with waterfall edge", "primary suite walk-in
  with custom closet system", "screened porch overlooking pond", not
  "spacious" or "gorgeous"]
- Recent updates: [LIST WITH DATES]
- Neighborhood / school district: [SPECIFIC]

Brand voice samples (paste 3 of my best past listings):
1. [PASTE]
2. [PASTE]
3. [PASTE]

Write a 280-word listing description that:
- Opens with the specific feature most likely to drive buyer interest
  (not "Welcome to..." or "Step into...")
- Names 4-6 specific property features by their concrete attributes
- Includes the neighborhood/school context briefly
- Ends with a specific call to action ("call to schedule a private
  showing this weekend", not "don't miss this opportunity")

Fair Housing compliance: do not reference, imply, or suggest preferences
for race, color, religion, national origin, sex, familial status,
disability, or any protected characteristic. Describe property and
neighborhood features only.

Forbidden phrases: "stunning", "gorgeous", "must-see", "dream home",
"won't last long", "welcome to", "step into", "boasting", "discover",
"prime location" without specifics.

2. Luxury property listing description

Use case: $1M+ residential listing requiring elevated tone Edit ratio: 28%

Write a luxury MLS listing description for [PROPERTY ADDRESS], a
$[PRICE] [PROPERTY TYPE] in [LUXURY MARKET CONTEXT].

Property highlights:
- [SPECIFIC LUXURY FEATURES — architect name if relevant, design
  pedigree, specific finishes, view specifics, lot features]
- [PRIVACY / GATED / SECURITY FEATURES IF APPLICABLE]
- [LIFESTYLE FEATURES — wine cellar capacity, gym square footage,
  guest house bedroom count]

Target buyer profile: [SPECIFIC — "primary residence buyer relocating
from [region]", "second-home buyer", "international investor", etc.]

Write a 320-word description that:
- Opens with the property's defining architectural or design statement
- Names specific design pedigree (architect, designer, builder if
  notable)
- Describes 3-5 spaces with specific dimensions and features
- Includes the privacy/lifestyle differentiator if relevant
- Closes with a private showing call to action through agent only

Tone: confident, restrained, not gushing. Luxury buyers find effusive
descriptions cheap.

Fair Housing compliance: same as standard listing.

Forbidden phrases: "stunning", "breathtaking", "unparalleled",
"once-in-a-lifetime", "world-class" (unless objectively true),
"opulent", "majestic".

3. First-time buyer listing description

Use case: Entry-level listing under $400K targeting first-time buyers Edit ratio: 24%

Write an MLS listing description for [PROPERTY ADDRESS], a $[PRICE]
[PROPERTY TYPE] positioned for first-time buyers in [MARKET].

Property highlights:
- [PRACTICAL FEATURES first-time buyers value — move-in ready, low
  maintenance, fenced yard, primary suite, parking, storage]
- [LOCATION ADVANTAGES — commute time to [employer], walk to [amenity],
  school district]
- [STARTER HOME ECONOMICS — utility costs if known, low HOA if any,
  recent updates that reduce immediate spend]

Write a 240-word description that:
- Opens with the practical first-time buyer advantage (not generic
  "perfect starter home")
- Names 5-6 specific practical features that reduce buyer anxiety about
  hidden costs
- Includes the commute / lifestyle context relevant to typical
  first-time buyer demographics
- Closes with a CTA that names a specific easy-to-attend showing time

Tone: warm, practical, not aspirational. First-time buyers value
"this works for my life" over "this is a dream".

Fair Housing compliance: do not target by familial status, age, or any
protected characteristic. Use lifestyle and practical descriptors only.

Forbidden phrases: "starter home" (perceived as condescending),
"perfect for", "won't last long", "investment opportunity" (different
buyer profile), "fixer-upper" if it's not.

4. Open house social media announcement

Use case: Social post announcing an upcoming open house Edit ratio: 22%

Write 4 social media post variants announcing an open house for
[PROPERTY ADDRESS], $[PRICE], on [DATE] from [TIME] to [TIME].

Property in one line: [ONE SENTENCE].
Specific feature most likely to draw open-house visitors: [ONE].
Local neighborhood angle: [ONE — specific to the area].

Each variant should:
- Open with a specific hook (not "Open house this weekend!" or
  "Mark your calendars!")
- Name the specific draw feature
- Include date, time, address concisely
- End with directions to RSVP, get listing details, or just show up

Variant breakdown:
1. Instagram/Facebook (under 280 chars, conversational, with 4
   relevant hashtags)
2. LinkedIn (under 400 chars, more professional tone)
3. NextDoor / local Facebook group (under 200 chars, hyper-local
   voice)
4. Twitter/X (under 240 chars, with 2 hashtags)

Forbidden phrases: "Open house alert", "save the date", "drop by",
"come check it out", "you don't want to miss this".

5. Sold listing announcement

Use case: Post-closing announcement that the property sold Edit ratio: 25%

Write a social media post announcing [PROPERTY ADDRESS] sold for
$[PRICE] on [DATE].

Specifics:
- Days on market: [#]
- Number of offers (if multiple): [#]
- Buyer story (if shareable): [BRIEF — "first-time buyers", "relocating
  family", etc. — must be Fair Housing compliant]
- What I did specifically to make the sale happen: [ONE SPECIFIC THING
  — staged with specific photographer, ran specific marketing campaign,
  negotiated specific concession]

Write a 180-word post that:
- Opens with the sale outcome (not "another one closed!" or "sold!")
- Names the specific value I delivered as the agent
- Includes the closing context (multiple offers, days on market) as
  social proof
- Closes with what I'm taking on next (positions me for next listing
  conversation)

Tone: professional, specific, not gloating.

Fair Housing compliance: do not reference protected characteristics
of the buyer in any way.

Forbidden phrases: "another one in the books", "crushing it", "best
week ever", "thrilled to announce", emoji-only celebration.

Lead nurture and client communication prompts (5)

6. Cold outreach to a FSBO

Use case: Email to a For Sale By Owner about representation Edit ratio: 28%

Write a cold email to a FSBO seller at [PROPERTY ADDRESS] who has had
the property listed [#] days at $[PRICE].

My specific credibility for selling in this area: [SPECIFIC NUMBERS —
"sold 14 homes in this neighborhood last year averaging 97% of list",
not generic claims].
What I observed about their current FSBO listing: [ONE SPECIFIC
OBSERVATION — staging, photography, pricing, MLS visibility — must be
constructive].
The specific value I'd add: [ONE SENTENCE].

Email must:
- Under 130 words
- Open with the specific observation (not "I noticed your FSBO listing")
- Acknowledge their decision to FSBO without dismissing it
- Name the specific credibility number relevant to their situation
- End with a low-friction next step (15-min phone call, not a listing
  presentation)

Tone: respectful peer, not pushy salesperson. FSBOs hate being pitched.

Forbidden phrases: "I noticed your listing", "I'd love to discuss",
"have you considered using an agent", "the value of representation",
"market expertise".

7. Buyer's agent representation pitch email

Use case: Email to a potential buyer about working with you as their agent Edit ratio: 26%

Write an email to [PROSPECT NAME] who reached out about [PROPERTY
ADDRESS or PROPERTY TYPE] in [AREA].

What I know about their search: [SPECIFIC — price range, must-have
features, timeline].
My specific credibility for this buyer's situation: [SPECIFIC NUMBERS].
The 30-minute discovery call I want to book: [SPECIFIC PURPOSE — "to
understand what's must-have vs. nice-to-have so I show you only homes
that fit"].

Email must:
- Under 150 words
- Open with a specific reference to what they told me about their
  search
- Briefly position my buyer's agent value with one specific number
- Suggest a 30-min discovery call with a clear agenda
- Make it easy to schedule (specific calendar link)
- Avoid pitching representation in the email — that's the call

Fair Housing compliance: do not assume household composition or
preferences based on protected characteristics.

Forbidden phrases: "I'd love to help you find your dream home", "let's
hop on a call", "no pressure", "trust", "my mission".

8. Follow-up after an unanswered first email

Use case: Follow-up to a lead who hasn't responded to first outreach Edit ratio: 23%

Write a follow-up email to [PROSPECT] who hasn't responded to my first
email [DAYS] days ago. Original email subject: "[SUBJECT]".

New angle I want to lead with: [SPECIFIC — a piece of useful market
data, a new listing matching their criteria, an answer to a likely
question they have about the market].

Follow-up must:
- Under 70 words
- Open with the new angle (not "I wanted to follow up" or "just bumping
  this")
- Provide actual value (insight, listing, market data) without expecting
  a response
- End with a single soft option that's easy to ignore

Forbidden phrases: "I wanted to follow up", "circling back", "bumping
this up", "just checking in", "I'm here if you need me".

9. Post-showing follow-up email

Use case: Email after showing a property to a prospective buyer Edit ratio: 22%

Write a post-showing follow-up email to [BUYER NAME] who I showed
[PROPERTY ADDRESS] on [DATE].

What they said during the showing about what worked / didn't work:
[SPECIFIC OBSERVATIONS].
Their specific concern they raised: [SPECIFIC].
My honest assessment of whether this property fits them: [PICK ONE —
"strong fit, recommend offering", "good fit but concerns to address",
"poor fit, here's what we should look for instead"].

Email must:
- Under 200 words
- Reference what they specifically said during the showing
- Address their specific concern with honest analysis
- Give my honest recommendation (don't push toward offer if it's not
  right)
- Include next step appropriate to my recommendation

Tone: trusted advisor, not pushy salesperson. Honest "this isn't right
for you" emails build referrals.

Forbidden phrases: "I think this could be the one", "won't last long",
"competitive market", "we should move quickly" unless objectively true.

10. Market update email to past clients

Use case: Monthly or quarterly market update to past client list Edit ratio: 24%

Write a market update email to my past client list about the [AREA]
market this [PERIOD].

Specific market data:
- Median sale price this period vs. [comparison period]: [#]
- Average days on market vs. comparison: [#]
- Inventory levels: [#]
- Specific notable transaction or trend: [SPECIFIC]

Email must:
- Under 280 words
- Open with the most specific data point relevant to past clients
- Provide 3-4 specific market data points with context
- Include one actionable insight (e.g., "if you've been considering
  selling, current inventory levels mean...")
- Close with a soft positioning that doesn't push for a transaction

Tone: trusted market expert sharing useful information, not salesperson
fishing for listings.

Forbidden phrases: "hot market", "perfect time to sell", "values are
skyrocketing", "limited inventory" without specific numbers.

CMA and pricing prompts (3)

11. Comparative Market Analysis narrative

Use case: Written CMA narrative for a seller meeting Edit ratio: 23%

Write the narrative section of a CMA for [PROPERTY ADDRESS], a [BED]/
[BATH] [PROPERTY TYPE] at [SIZE] sq ft built in [YEAR].

Recent comparable sales in [DISTANCE] miles:
1. [ADDRESS, SALE PRICE, DOM, KEY DIFFERENCES FROM SUBJECT]
2. [ADDRESS, SALE PRICE, DOM, KEY DIFFERENCES]
3. [ADDRESS, SALE PRICE, DOM, KEY DIFFERENCES]

Active competitive listings:
1. [ADDRESS, LIST PRICE, KEY DIFFERENCES]
2. [ADDRESS, LIST PRICE, KEY DIFFERENCES]

My recommended list price range: $[LOW] - $[HIGH], with target list at
$[TARGET].

Write a 400-word CMA narrative that:
- Explains the market context for this property and price range
- Walks through each comparable sale with specific dollar adjustments
- Discusses active competition and how this property positions
- Justifies the recommended list price range with specific numbers
- Closes with the recommended strategy (list, time on market expectation,
  price reduction strategy if applicable)

Tone: data-driven, confident, not promotional.

Forbidden phrases: "hot market", "obvious pricing", "no-brainer",
"comparable" without specific differences.

12. Pricing conversation script for seller meeting

Use case: Spoken script for the pricing portion of a listing presentation Edit ratio: 26%

Write the spoken script for explaining the recommended list price to
[SELLER NAME] for [PROPERTY ADDRESS] during a listing presentation.

The recommended list price: $[PRICE].
The seller's hoped-for price (if known): $[SELLER PRICE].
The gap to explain: [DOLLAR DIFFERENCE].
Specific market data supporting the recommendation: [3-4 DATA POINTS].

Script must:
- Sound natural when spoken (not read)
- Acknowledge the seller's price expectation respectfully
- Walk through the supporting market data simply
- Explain the consequences of overpricing (DOM, price reductions,
  staleness)
- Stand firm on the recommendation without being pushy

Tone: trusted advisor explaining hard math, not pushy negotiator.

Forbidden phrases: "the market doesn't lie", "buyers won't pay that",
"you need to be realistic", anything that talks down to the seller.

13. Price reduction conversation script

Use case: Conversation with a seller about reducing list price after time on market Edit ratio: 28%

Write the spoken script for recommending a price reduction to [SELLER]
on [PROPERTY ADDRESS] after [DAYS] days on market.

Current list price: $[CURRENT].
Recommended new list price: $[NEW].
Specific evidence supporting the reduction: [SHOWING COUNT, OFFER COUNT,
COMPETITIVE LISTING ACTIVITY, COMPARABLE SALES DURING TIME ON MARKET].

Script must:
- Open by referencing the specific evidence (not "we need to talk about
  the price")
- Acknowledge the difficulty of the conversation
- Walk through what's been tried (marketing, showings, feedback)
- Recommend the specific reduction with the supporting math
- Discuss what happens if we don't reduce (additional weeks, further
  price drop, staleness)

Tone: trusted advisor delivering hard news, not pushy salesperson.

Forbidden phrases: "the market has spoken", "buyers don't want it",
anything blaming the seller for the lack of activity.

Buyer process and offer prompts (3)

14. Buyer offer letter (personal note to seller)

Use case: Personal offer letter from buyer to seller (where allowed by state) Edit ratio: 30%

Write a personal offer letter from [BUYER NAME] to the seller of
[PROPERTY ADDRESS], offering $[PRICE] with [SPECIFIC TERMS].

Buyer specifics that are appropriate to share:
- General timing (move-in by [DATE])
- Why this property specifically (the feature that matters most to them)
- Their plans for the property (continuing existing landscaping,
  preserving the [feature], etc.)

Letter must:
- Under 220 words
- Open with the specific reason this property stood out
- Mention one concrete thing they appreciate about the home
- Be warm and personal without being effusive
- Avoid pressure or competitive offer comparisons

Fair Housing compliance (CRITICAL): do not reference race, color,
religion, national origin, sex, familial status, disability, source of
income, or any other protected characteristic. Some states prohibit
buyer love letters entirely — confirm legality in your state before
sending.

Forbidden phrases: "we fell in love", "your beautiful home", "perfect
family home", "growing family", "first home", "dream home", or any
reference to specific personal characteristics.

15. Counter-offer explanation email

Use case: Email to buyer explaining a seller counter-offer Edit ratio: 25%

Write an email to my buyer client [NAME] explaining the seller's
counter-offer on [PROPERTY ADDRESS].

Original buyer offer: $[PRICE] with [TERMS].
Seller counter: $[COUNTER PRICE] with [COUNTER TERMS].
Key differences from original offer: [SPECIFIC DOLLAR / TERM CHANGES].
My recommended response: [PICK — accept, counter-back with $[X],
walk away].

Email must:
- Under 240 words
- Explain the counter in plain language (no jargon)
- Walk through what changed from their original offer with specific
  numbers
- Give my honest recommendation with reasoning
- Make clear the decision is theirs

Tone: trusted advisor, not pushy. Buyers feel pressured easily during
negotiation.

Forbidden phrases: "they're playing hardball", "we should accept",
"won't accept less", "take it or leave it".

16. Multiple offer situation explanation

Use case: Email/text explaining a multiple offer situation to a buyer Edit ratio: 27%

Write an email to my buyer client [NAME] explaining we're in a multiple
offer situation on [PROPERTY ADDRESS].

Specifics:
- Number of competing offers: [#]
- Seller's deadline for highest and best: [DATE / TIME]
- My read on how to be competitive: [SPECIFIC RECOMMENDATIONS]
- What the buyer's stretched price would be: [$ ROOM TO MOVE]

Message must:
- Under 200 words
- Lay out the situation clearly without alarm
- Explain "highest and best" honestly (no false urgency tactics)
- Give specific recommendations on price, contingencies, escalation
- Make clear the decision is theirs and there will be other properties

Tone: trusted advisor in a high-pressure moment, not panicked
salesperson.

Forbidden phrases: "we have to act now", "this might be your only
chance", "you'll regret it", "stretch as much as you can".

Social media and content prompts (3)

17. Weekly market insight social post

Use case: Educational social post about local market conditions Edit ratio: 24%

Write a social post sharing one specific market insight about [AREA]
this week.

Specific data:
- The specific stat: [#]
- The comparison context: [vs last week, vs last quarter, vs last year]
- What this means for buyers or sellers: [ONE SPECIFIC IMPLICATION]

Post must:
- Under 240 chars for Twitter/Instagram caption opening
- Lead with the specific stat (not "Did you know..." or "Just in...")
- Provide context that makes the stat meaningful
- Close with one specific actionable insight

Variants needed:
1. Twitter/X (240 chars, 2 hashtags)
2. Instagram/Facebook (under 400 chars, 4 hashtags)
3. LinkedIn (under 600 chars, professional tone)

Forbidden phrases: "Did you know", "Just in", "Hot off the press",
"You won't believe", emoji bullets.

18. Neighborhood guide intro

Use case: Blog or social long-form intro to a neighborhood Edit ratio: 26%

Write the opening 200 words of a neighborhood guide for [NEIGHBORHOOD],
[CITY, STATE].

Specifics:
- The defining feature of this neighborhood: [ONE SENTENCE]
- Median home price range: $[LOW] - $[HIGH]
- Typical buyer demographic (housing market terms only, not protected
  characteristics): [first-time buyers, downsize buyers, etc.]
- 3 specific amenities (specific names, not generic): [LIST]

Opening must:
- Lead with what makes this neighborhood distinct (not "Welcome to..."
  or "Located in...")
- Name 2-3 specific amenities by name
- Give the price context briefly
- Set up what the rest of the guide will cover

Fair Housing compliance: do not characterize the neighborhood by
protected characteristics. Describe amenities, geography, architecture,
commute access.

Forbidden phrases: "Welcome to", "Located in", "Nestled in", "Hidden
gem", "Up-and-coming" (loaded with implications), "great place to raise
a family" (familial status), "diverse community" (protected
characteristic).

19. Just-listed announcement post

Use case: Social post announcing a new listing Edit ratio: 22%

Write 3 social post variants announcing a new listing at [PROPERTY
ADDRESS], $[PRICE].

Property in one line: [ONE SENTENCE].
The specific feature most likely to drive showings: [ONE].
Open house timing if scheduled: [DATE/TIME or "private showings only"].

Each variant should:
- Open with the specific draw feature (not "Just listed!" or "New on
  the market!")
- Include price, beds/baths, neighborhood concisely
- End with a specific viewing CTA

Variant breakdown:
1. Instagram/Facebook (under 400 chars, 4 hashtags, visual-first)
2. LinkedIn (under 500 chars, market-positioning angle)
3. NextDoor / hyper-local (under 200 chars, neighbor-to-neighbor tone)

Forbidden phrases: "Just listed", "new to market", "you've gotta see
this", "won't last", "first to see".

Client experience and retention prompts (3)

20. Post-closing thank-you note

Use case: Personalized thank-you to a client after closing Edit ratio: 25%

Write a personalized thank-you note to [CLIENT NAME] after closing on
[PROPERTY ADDRESS].

Specifics about their journey:
- How long we worked together: [TIMEFRAME]
- One specific memory from the process: [SPECIFIC MOMENT]
- One specific thing they did that made the transaction work: [SPECIFIC]
- What I want to acknowledge about working with them: [ONE SENTENCE]

Note must:
- Under 140 words
- Reference the specific shared moment / journey detail
- Acknowledge what they brought to the transaction
- Avoid generic "it was a pleasure" language
- Close with a specific positioning for ongoing relationship (annual
  market update, holiday note, etc.)

Tone: warm, personal, specific. Generic thank-yous feel automated and
get filed forgettably.

Forbidden phrases: "It was a pleasure", "Thank you for trusting me",
"I'm so grateful", "Working with you was amazing".

21. Annual past-client check-in email

Use case: Annual check-in to past clients (1-3 years after closing) Edit ratio: 23%

Write an annual check-in email to [PAST CLIENT NAME] who closed on
[PROPERTY ADDRESS] [YEARS] years ago.

Specific local market data they'd find useful:
- Current estimated value of their property (Zestimate or my BPO):
  $[X]
- How that compares to what they paid: [% / $ change]
- Specific neighborhood activity worth noting: [SPECIFIC]

Email must:
- Under 200 words
- Reference the specific anniversary timing
- Provide the specific data about their property
- Note one specific neighborhood update they might not know
- Close with a no-pressure positioning ("just wanted to share since I
  thought you'd find it interesting")

Tone: genuinely useful, not sales fishing. Generic "thinking of you"
emails feel like setups; specific data feels like a friend who's a
real estate pro keeping them informed.

Forbidden phrases: "thinking of you", "just wanted to check in", "let
me know if you're considering selling", "I have buyers looking in your
area" (unless objectively true).

22. Referral request after a closing

Use case: Asking a happy client for a referral after closing Edit ratio: 24%

Write a spoken script for asking [CLIENT NAME] for a referral during
our post-closing celebration call.

What we accomplished together: [SPECIFIC OUTCOME].
The specific way I delivered value during the process: [SPECIFIC].
The kind of referral I'd value most: [SPECIFIC — "anyone else in your
network considering selling in this neighborhood", "co-workers who
might be relocating", "friends asking about market timing"].

Script must:
- Sound natural when spoken (not read)
- Acknowledge what we accomplished together specifically
- Make the referral ask specific (so they actually think of someone)
- Make it easy to say "yes, but I need to think about who"
- Offer something to the referred person that makes the intro easy

Tone: confident peer asking a favor, not desperate salesperson.

Forbidden phrases: "I'd love a referral", "do you know anyone", "all
my business comes from referrals", anything that puts them on the spot.

How to use these prompts effectively

The 22 prompts above are templates. The work is in the specificity you bring to each [PASTE] field. ChatGPT cannot invent the actual features of the property, the specific market data, the specific things your client said during the showing. You provide that material. ChatGPT compresses it into a usable draft.

The 60-second per-prompt workflow:

1. Save the prompt template that matches your use case (bookmark a ChatGPT conversation per use case)
2. Before opening ChatGPT, gather the specific context the prompt needs (MLS sheet data, market stats, client notes)
3. Fill in the [FIELDS] with that specific context
4. Run the prompt
5. Edit one sentence to add a specific detail only you would say
6. Use the output

The first 20 outputs will need heavier editing (40-50% edit ratio) as you learn what your AI defaults need correcting. By output 50, the edit ratio drops to 22-28% across most prompt types. The setup is the lever.

For the broader prompt-structure approach across other use cases, our 22 ChatGPT prompts for sales covers the sales-specific application, and our 24 ChatGPT prompts for marketing covers marketing email and ad workflows that apply to real estate marketing.

The Bottom Line

The 22 prompts in this article are templates that work for residential real estate agents — when you bring specific property details, market data, and client context to each [FIELD] and use the Fair Housing compliance reminders to keep your content legally safe. Realistic time savings: 15-20 minutes per listing description, 5-7 minutes per client email, 10-15 minutes per market update. For a solo agent producing 4-6 listings monthly plus 30-40 weekly client emails, that's 8-12 hours recovered every week.

The watch-out: AI cannot enforce Fair Housing compliance for you. The reminders in each prompt help, but the agent is legally responsible for every public-facing piece of content the AI drafts. Review every output against Fair Housing Act language before publishing or sending. The 30 seconds of compliance review per output is what keeps the time savings from becoming liability exposure.

Frequently Asked Questions

Are ChatGPT-written real estate listings legally compliant with Fair Housing? ChatGPT-written listings are compliant with the Fair Housing Act only when the agent reviews and edits the output to remove any language that references or implies preferences for protected characteristics (race, color, religion, national origin, sex, familial status, disability, source of income in some jurisdictions). The prompts in this article include explicit Fair Housing compliance instructions, which reduces but does not eliminate the risk. The agent remains legally responsible for every public-facing piece of content regardless of who or what drafted it. A 30-second compliance review of every AI output before publishing is non-negotiable. Agents who skip this review have already been sanctioned in multiple jurisdictions for AI-drafted listings that violated Fair Housing guidelines — the AI's mistakes become the agent's liability.

Can ChatGPT write a complete CMA, or just the narrative section? ChatGPT can write the narrative section of a CMA — the written explanation of comparable sales, adjustments, market context, and recommended pricing strategy — but cannot generate the underlying data. The agent provides the comparable sales, the active listings, the dollar adjustments, and the recommended price range from MLS data; ChatGPT compresses that into a 400-word narrative seller-meeting document. The full CMA package (data tables, adjustment grids, photos of comps, mapping) still requires CMA software like CloudCMA, RPR, or similar. ChatGPT replaces the 45-60 minutes the agent would spend writing the narrative, not the 30-90 minutes of data gathering. For solo agents producing 4-6 CMAs monthly, this represents 3-6 hours of recovered time monthly without affecting CMA quality.

How do I prevent ChatGPT from producing generic AI-toned listing descriptions? The three structural fixes that prevent generic AI listings: load 3 of your past best listings as voice samples in the prompt (the AI matches what you've written before); provide 4-6 specific property features by their concrete attributes ("marble kitchen island with waterfall edge") not generic descriptors ("spacious gourmet kitchen"); and include an explicit forbidden-phrase list ("stunning", "must-see", "won't last", "welcome to", "step into"). The forbidden list is what stops the AI from defaulting to real estate cliché openers. Without these three elements, the AI produces the same generic listing every other agent's AI tool produces. With them, the output reads as a specific agent describing a specific property — which is what buyers respond to.

Should real estate agents use AI for everything, or are there workflows that should stay human? AI is the right tool for listing descriptions, CMA narratives, market updates, social posts, and email drafting — anywhere the agent provides the specific data and AI handles the structure and prose. AI is the wrong tool for in-person buyer or seller meetings, negotiation phone calls with the other agent, difficult conversations (price reductions, walking away from a deal, complaint handling), and the relationship-building work that drives referrals. Use AI for the writing layer around the agent's work; never use AI to participate in conversations where the client is paying for human judgment and presence. The agents who get this wrong substitute AI for relationship work and watch their referral rates drop; the agents who get it right use AI to recover hours for more relationship work and see referral rates rise.

For the full Real Estate AI tool stack including the CRMs and listing platforms that pair with these prompts, see our AI tools for real estate agents review.

Sources

• U.S. Bureau of Labor Statistics — Real Estate Brokers and Sales Agents
• Federal Trade Commission — AI in Consumer-Facing Communications
• U.S. Department of Housing and Urban Development — Fair Housing Act
• National Association of Realtors — Code of Ethics

For the editorial standards behind every recommendation on this site — including how AI assists with our writing and how we verify sources — see our Editorial Process page.