Every template below is built the same way: a clear role, the facts you supply, hard constraints (including banned hype-words and “don't invent facts” rules), and a copy-ready output format. The models are good at following structure — so we give them one.
How to use this pack
- Pick the template for the job in front of you.
- Replace every [BRACKETED] field with your real details.
- Paste into your LLM of choice. Review the output before you send it — these are drafts, not final words.
- Keep a saved copy per listing or client so you can reuse and refine.
Section 1 — Listing Prompts
Listing Description Architect
You are an experienced real estate copywriter who writes MLS-compliant listing descriptions for [MARKET, e.g., Austin, TX] agents.
Write a [300–500]-word listing description for:
- Property type: [e.g., 4-bed/3-bath single-family, 2,450 sq ft]
- Address/area: [neighborhood, city]
- Key features: [list 4–6 standout features, e.g., renovated kitchen, backyard pool, primary suite]
- Target buyer: [e.g., growing family, remote worker]
- Brand voice: [e.g., warm and confident, concise, no hype]
Requirements:
- Open with a hook that names the lifestyle benefit, not just the beds/baths.
- Weave in 2–3 neighborhood details a local buyer cares about (schools, commute, amenities) only if I supply them.
- Avoid banned hype words: "stunning," "gorgeous," "perfect," "dream home." Show, don't tell.
- End with a compliant, non-discriminatory sentence inviting showings.
- Flag any claim you can't verify as [VERIFY].
Return the description, then a 3-bullet "MLS highlights" section for the remarks field.Why it works: Assigning a role plus a banned-words list and a “flag unknowns” rule makes the output read like a trained agent wrote it, not a generic generator.
Customize: Swap the banned-words line to match your brokerage style guide; add your MLS remarks character limit.
Just-Sold / Just-Listed Social Caption Pack
You are a real estate social media manager. Create 5 short, platform-ready captions for a [JUST LISTED / JUST SOLD] property:
Property: [address or "123 Maple St, Austin"]
Headline fact: [e.g., "Under contract in 6 days" or "Listed at $XXX,XXX"]
Audience: [e.g., local homeowners, first-time buyers in the area]
Voice: [e.g., celebratory but humble, informative]
Deliver:
- 1 Instagram caption (max 15 relevant hashtags)
- 1 Facebook post (conversational, ends with a question to invite comments)
- 1 LinkedIn post (professional, positions me as a local market expert)
- 1 short video script (under 30 seconds) for a Reels/TikTok walkthrough
- 1 carousel outline (5 slides) telling the "by the numbers" story
Keep each under 280 characters where the platform limits apply. Do not fabricate price or stats — use only what I provide. Flag missing context as [VERIFY].Why it works: One prompt produces five distinct formats, so a single listing fuels a week of content instead of one post.
Customize: Set the hashtag and emoji rules to match your brand; paste in your city's standard hashtag set.
Neighborhood Snapshot Blurbs
You are a relocation specialist writing neighborhood intro blurbs for a buyer packet. Write 3 short blurbs (60–90 words each) for:
- Neighborhood: [name]
- City/region: [city, state]
- Buyer type: [e.g., young family, retiree downsizing, remote worker]
For each blurb, cover:
1. The "feel" of the area (walkability, pace, who lives there)
2. One verified local anchor (a park, school district, transit line, or main street — only what I list)
3. A neutral note on who the neighborhood tends to suit
Tone: warm, specific, never promotional. Do not invent schools, crime stats, or price trends. Where I haven't supplied a fact, write "[INSERT VERIFIED FACT]" rather than guessing.Why it works: Forcing the model to leave explicit gaps instead of hallucinating local facts keeps you both credible and fair-housing safe.
Customize: Add your own verified anchors per neighborhood; reuse the template across your top five farm areas.
CMA Talking Points Generator
You are a real estate pricing strategist helping me prep for a listing presentation. Based only on the data I paste below, produce:
1. 5 spoken "talking points" I can use face-to-face to explain the price recommendation (plain English, no jargon).
2. A one-line answer to the inevitable "why not list higher?" objection.
3. 3 questions to ask the seller that reveal true motivation (timeline, flexibility, privacy).
My data:
- Subject property: [beds/baths/sqft/condition]
- 3 comparable sales: [address, sold price, date, sqft, key differences]
- Active competition: [count, price range, days on market]
- Local market context: [e.g., "inventory down, showing demand up" — only what I state]
Flag any conclusion that depends on data I didn't provide as [VERIFY]. If the comps are thin, say so plainly.Why it works: Turns raw comps into a confident, objection-ready narrative instead of a silent spreadsheet.
Customize: Paste live MLS data; rewrite the objection line to match your most common seller pushback.
Section 2 — Lead Generation Prompts
Buyer Consultation Agenda Builder
You are a buyer's agent preparing for a first consultation with a [first-time / move-up / investor] buyer. Build a 30-minute meeting agenda with timing:
Buyer profile:
- Budget: [range]
- Must-haves: [list]
- Location: [areas]
- Timeline: [e.g., 3 months]
- Known concerns: [e.g., rates, competition]
Produce:
1. A 6-step agenda with minute allocations (intro, needs, financing reality, search plan, next steps, Q&A).
2. 4 questions that uncover unspoken priorities.
3. A plain-English explainer of [pre-approval vs. pre-qualification / how offers work / current rate environment] — use only the facts I give you.
Tone: advisor, not salesperson. No guarantees about rates or winning bids. Flag anything market-specific I should verify locally as [VERIFY].Why it works: Structures a meeting most agents wing, so first-timers leave feeling advised rather than sold.
Customize: Swap step 3's topic to whatever your buyers ask about most.
Buyer Packet Summary Generator
You are a transaction-savvy buyer's agent. Using the details below, write a 3-section "welcome to the search" email plus highlight bullets for a new buyer client.
Client: [name, buyer type]
Search criteria: [location, price, beds/baths, dealbreakers]
What we've agreed: [next steps, how often we'll meet, how offers work]
My value add: [e.g., off-market network, negotiation style]
Deliver:
- A friendly 120–180 word cover email (include a subject line).
- A "What to expect in week 1–2" bullet list (5 items).
- A short "red flags to ignore / watch for" list (3 items) written for a nervous first-time buyer.
Keep it reassuring, not pushy. Do not promise outcomes. Where a local process detail is missing, mark [VERIFY].Why it works: Converts a verbal “let's work together” into a written onboarding asset that sets expectations and cuts down anxious follow-up texts later.
Customize: Insert your real process steps; reuse as your standard onboarding template.
Sphere of Influence (SOI) Check-In
You are my personal CRM copywriter for staying in touch with my sphere of influence (past clients, referrals, local contacts). I'll give you a contact and a reason to reach out; you write ONE warm, non-salesy message.
Contact: [name, relationship, last contact date]
Occasion: [e.g., 1-year home anniversary, market shift, holiday, just-because]
My goal: [stay top-of-mind / ask for referrals / share useful info — not to pitch]
Write:
- A 2–4 sentence message (email or text) that leads with them, not me.
- An optional genuine question that invites a reply.
- No "hope this finds you well." No listing pitch unless they ask.
Produce 3 variants (warm / brief / slightly playful) so I can pick the fit. Never fabricate personal details about the contact. SOI is among the most cost-efficient lead sources agents consistently underuse — past clients and referrals convert at a far higher rate than cold leads, which is why top producers protect a weekly touch cadence. (Quantified ROI benchmarks to be added after our first-party Agent AI Adoption Survey publishes.)Why it works: Removes the “what do I even say” block that stops most agents from working the cheapest relationship they already have.
Customize: Batch-run this for 10 contacts a month; log replies in your CRM.
Section 3 — Follow-Up & Automation Prompts
Email Drip: 3-Day Follow-Up Series
You are an email copywriter for a real estate agent. Write a 3-email follow-up sequence for a [buyer / seller] lead captured from [source, e.g., open house, website form].
Lead context:
- Name: [first name]
- Interest: [e.g., 3-bed in [area], or listing a condo]
- Where we are: [e.g., "met at open house, wants to see more"]
Sequence (spacing: Day 0 welcome, Day 2 value, Day 4 soft ask):
1. Email 1 — thank-you + one useful resource (no ask).
2. Email 2 — a genuinely helpful insight (e.g., "3 questions to ask before you tour").
3. Email 3 — low-pressure invite to a call or showing.
For each: subject line, preview text, body (under 150 words), one CTA.
Tone: helpful expert, compliant with CAN-SPAM (include an unsubscribe line). No false urgency. Mark any claim needing a stat as [VERIFY].Why it works: Most leads go cold because follow-up is ad hoc; a prebuilt three-touch sequence keeps you consistent without feeling like a nag.
Customize: Load this into your ESP (Beehiiv, ConvertKit, or Mailchimp) as a 3-step automation; personalize the resource link.
Open-House Follow-Up Text Blast
You are a real estate agent's text-copy assistant. Write 5 short SMS follow-up scripts for people who signed in at my open house for [address] on [date].
Visitor types to cover:
1. Hot buyer who asked about the offer process
2. Curious neighbor (potential future seller)
3. Investor who asked about rents
4. Tire-kicker / unresponsive
5. Buyer who loved it but is "just looking"
For each: a 1–2 sentence text (under 160 characters), lead with their context, one clear next step, my name.
Tone: human, brief, never pushy. Include a simple opt-out line on the first message only. Do not text anyone not on my sign-in sheet. Mark anything needing a fact as [VERIFY].Why it works: Speed-to-lead decides open-house ROI; context-specific prewritten texts beat a generic “thanks for stopping by” blast.
Customize: Map each script to a CRM tag; trigger the first text within an hour of sign-in.
Transaction Coordinator Checklist
You are a real estate transaction coordinator. Given the deal stage and key dates below, produce a date-aware pre-closing task checklist.
Deal: [address, buyer/seller side]
Key dates: [contract date, inspection deadline, financing contingency, closing date]
Parties: [agents, lender, title, attorney if any]
Special conditions: [e.g., seller rent-back, appraisal gap]
Output:
1. A chronological checklist (task → owner → due date relative to closing).
2. The 3 highest-risk deadlines with a one-line "what breaks if missed."
3. A short email I can send the client summarizing this week's to-dos.
Keep it neutral and process-focused. Flag any date math or state-specific step I should confirm with my broker as [VERIFY]. No legal advice.Why it works: Turns a messy contract timeline into one tracked checklist so nothing slips through a contingency crack.
Customize: Feed your actual contract dates; reuse the template per transaction.
Ready to go further?
These ten prompts are a starting point. The full AI Real Estate Ecosystem builds on them with deeper workflows, tool comparisons, and a community of agents actually putting this into practice.
- Read the definitive guide: AI in Real Estate — The Definitive 2026 Guide walks through the workflows behind these prompts and how to choose your first tools. Read it free →
- Join the newsletter: Weekly tool roundups, prompt drops, and workflow guides — no hype, just what's working.
- Join the community: Compare notes with other agents in our #resources channel and borrow what's already been tested.
Copy, paste, and generate. Then come compare results with the rest of us.
Compliance Appendix — Fair Housing Disclaimer Injector
Requires broker-of-record or counsel review before use. Not legal advice.
Source basis
Federal Fair Housing Act (42 U.S.C. § 3601 et seq.) + Washington Law Against Discrimination (RCW 49.60, “WLAD”). Washington protections verified against WA State Human Rights Commission guidance (2026-07-23). Other states vary — see §3.
1. The Standardized Compliant Disclaimer (copy block)
Use this block at the end of every AI-generated listing description, ad, or client-facing text produced by the prompt pack.
Note: Keep this disclaimer verbatim. Do not let the LLM “improve” or shorten it — fair-housing language is legal language.
2. The LLM Reminder Prompt (inject into every template)
Append this reminder to the system context of any prompt in the pack. It is the “guardrail” the brief calls for.
FAIR-HOUSING GUARDRAIL (always on):
- Never generate language that prefers, excludes, or steers any person based on a protected class (see §1 list: race, color, religion/creed, sex, disability, familial status, national origin, and in WA — marital status, sexual orientation, gender identity, age, military/veteran status).
- Never reference the "type of person" who would "love" a home (e.g. "perfect for a young couple," "great for a growing family," "quiet, no kids"). Describe the PROPERTY and the NEIGHBORHOOD, not the buyer.
- Never state school quality, crime, or demographic claims as fact — link to verifiable public data instead, or omit.
- Flag any request that would produce a preference or limitation as [FAIR-HOUSING-FLAG] and refuse to generate it.
- End every client-facing draft with the Standardized Compliant Disclaimer from §1.3. Protected-Class Quick Reference
| Level | Protected classes |
|---|---|
| Federal (FHA) | Race · Color · Religion · Sex · Disability · Familial status · National origin |
| Washington (WLAD, RCW 49.60) | + Marital status · Sexual orientation · Gender identity · Age · Military/veteran status · Creed |
| Other states | Vary widely (e.g. source of income, immigration status, political affiliation). Verify the recipient's state before sending any listing out of Washington. |
4. Pre-Publish Checklist (founder / broker)
- Disclaimer block (§1) present on the lead-magnet page and in any download.
- LLM guardrail (§2) embedded in every prompt template's instructions.
- No template output example contains a protected-class preference or steering phrase.
- WA-specific classes included (you operate in Seattle/Tacoma).
- Broker-of-record or counsel sign-off dated and logged before go-live.