In NAR's 2025 Technology Survey, 20% of Realtors reported using AI tools daily and 22% weekly, while 32% had not yet used AI in their business.Source ↗
82% of agents use AI to write listing descriptions, up from 58% in 2024, and 74% use it for marketing content like email and social posts, per Delta Media's 2026 AI survey.Source ↗
ChatGPT is the most-used AI tool among Realtors at 58%, followed by Gemini at 20% and Copilot at 15%.Source ↗
Only 17% of Realtors say AI has had a significant positive impact on their business, while 46% see no noticeable difference — results depend heavily on how it's used.Source ↗
writingChatGPTClaude

First-draft listing descriptions in minutes

Writing a fresh MLS description for every listing eats an hour you don't have, and recycled phrasing ("charming, must see!") does nothing for the property. This is now the single most common AI task in the industry — 82% of agents draft listing copy with AI — but the draft still has to be fact-checked and Fair Housing-clean before it hits the MLS.

Prompt
You are an experienced real estate copywriter who writes MLS listing descriptions that sell homes without hype or filler. Write a listing description for the property below.

Property facts (use ONLY these — do not invent or embellish anything):
{{property_facts}}

The single most compelling feature to lead with: {{standout_feature}}

Requirements:
- One headline option under 10 words, then a full description of no more than {{character_limit}} characters, then a 150-character teaser for portal previews.
- Describe the property and its features, never the buyer. No language implying who should live there — nothing like "perfect for families," "ideal for young professionals," or "walking distance to church," and no references to any protected class under the Fair Housing Act.
- Lead with the standout feature, not "Welcome home to..."
- Concrete and specific over generic: skip "charming," "must see," and "won't last."
- Plain sentences, active voice, no exclamation points.
- If a fact seems ambiguous or is missing (square footage, year built, school district), insert [VERIFY] rather than guessing.

Return the headline, the description with its character count, and the teaser, then a one-line list of any [VERIFY] items.

Fill in your details and the prompt updates live — then copy.

automationChatGPTClaude

Open house follow-up sequences that actually go out

Open house leads go cold within days, and NAR data cited across the industry says most buyers end up working with the first agent who responds. Drafting a personal-sounding message for every name on the sign-in sheet is exactly the work that doesn't happen at 9 p.m. on Sunday — unless AI writes the first pass.

Prompt
You are a real estate agent's assistant drafting follow-up messages that sound like a real person, not a drip campaign. I just held an open house at {{property_address}}. Key selling points: {{property_highlights}}.

Here are my notes on each visitor from the sign-in sheet: {{visitor_notes}}

For EACH visitor, write a 4-touch follow-up sequence:
1. Same-day text (under 300 characters) — thank them and reference something specific they said or asked about.
2. Day-2 email — answer their open question if I noted one, add one worthwhile detail about the home, and ask ONE question about what they're looking for.
3. Day-5 email — offer to send similar listings matching what they described.
4. Day-10 email — a soft value touch (neighborhood price snapshot or buyer consult offer), no pressure.

Rules: conversational tone, short paragraphs, one question max per message, no "just checking in," no urgency tactics like "this won't last." Subject lines under 6 words. Where I lack a detail, use [FILL] placeholders instead of inventing one.

Fill in your details and the prompt updates live — then copy.

analysisChatGPTGemini

Monthly market updates clients actually read

"How's the market?" is the question every past client asks, and a monthly plain-English update is one of the best ways to stay top of mind — but translating MLS stat sheets into readable copy takes hours. Agents and teams now routinely paste MLS data into ChatGPT for quick market overviews to send clients or take into appointments.

Prompt
You are a local real estate analyst who explains market data to regular homeowners without jargon or spin. I'm writing my monthly market update for {{audience}}.

Here are this month's MLS statistics for {{market_area}}, with prior-period comparisons:
{{mls_stats}}

Write:
1. A 200-word plain-English market update for an email newsletter. Open with the one number that changed most and what it means in practice. Keep it at an 8th-grade reading level.
2. Three one-sentence takeaways labeled "If you're buying," "If you're selling," and "If you're staying put."
3. A 40-word version for a social caption.

Rules: Use ONLY the numbers I provided, and always name the comparison period ("down 4% from May"). No predictions stated as certainty — "if this continues" is fine, "prices will rise" is not. No "great time to buy or sell" cheerleading; say only what the data supports. If a stat is missing for a claim, leave the claim out rather than estimating.

Fill in your details and the prompt updates live — then copy.

planningClaudeChatGPT

Negotiation prep briefs before every offer call

Before presenting or countering an offer you're juggling comps, inspection findings, the other side's motivations, and your client's priorities — and the synthesis usually happens in the car. Trade-press prompt guides now put negotiation-strategy prep among the highest-value AI uses because it turns scattered deal facts into a structured brief in minutes.

Prompt
You are a veteran real estate negotiation coach preparing an agent for an offer conversation. Build a one-page negotiation brief from the deal facts below. Do not add facts I didn't give you; if something important is missing, list it under "Find out first."

Deal facts (offer terms, comps summary, inspection findings, days on market, competing offers if any):
{{deal_facts}}

My client's priorities, ranked: {{client_priorities}}

Current market context: {{market_conditions}}

Format the brief exactly like this:
1. Position summary — 3 bullets on our leverage, 3 on theirs.
2. Three likely scenarios (counter on price, counter on terms, reject) — for each, a probability rating, the response that best protects my client's top priority, and what we'd concede first.
3. Client talking points — 5 short lines in plain language I can say on a call, including one that sets expectations if we lose the deal.
4. Questions to ask the other agent that reveal motivation without giving away ours.
5. Find out first — missing information that would change this analysis.

No legal advice; flag anything that belongs with an attorney or lender.

Fill in your details and the prompt updates live — then copy.

creativeChatGPTGemini

A month of social content in one sitting

75% of Realtors use social media in their business, but consistency — not creativity — is what kills most agents' accounts. Batching a month of post ideas grounded in your actual listings and neighborhood knowledge beats staring at a blank caption box every Tuesday.

Prompt
You are a social media strategist for a solo real estate agent — a person, not a brand account. Build a 4-week content calendar, 3 posts per week, for Instagram and Facebook.

My market and niche: {{market_area}}, focused on {{niche_audience}}.

What I actually have to work with this month (listings, open houses, closings, local events): {{upcoming_content}}

Mix these post types across the month: listing or open-house spotlights (only for properties listed above), local business or neighborhood spotlights, one market-stat post per week marked [INSERT STAT] for me to fill with real MLS numbers, myth-busting or how-to education for my niche, and one personal or behind-the-scenes post per week.

For each post give: day, post type, hook (first line, under 12 words), a 60-90 word caption draft in my voice — direct, warm, zero hype — a format suggestion (reel, carousel, or single photo) with a simple shot idea, and a CTA. Vary the CTAs; "DM me" may appear at most 3 times in the month. Never invent market numbers, events, or property details.

Fill in your details and the prompt updates live — then copy.

communicationChatGPTClaudeGemini

Translating listings and updates for multilingual clients

When a buyer's first language isn't English, a translated listing description or transaction update builds trust that a Google-Translate paste job can't. AI handles the first pass in seconds — but real estate idioms like "move-in ready" don't translate literally, so the prompt has to force flagging and back-translation for verification.

Prompt
You are a professional translator who specializes in US residential real estate and writes for {{target_language}}-speaking clients living in the United States.

Translate the following {{document_type}} into {{target_language}}:

{{source_text}}

Requirements:
- Use the terminology actually used by {{target_language}}-speaking homebuyers in the US — for example, how "escrow," "earnest money," or "HOA" are commonly rendered or kept in English — not textbook translations.
- Keep all names, numbers, addresses, dates, and dollar amounts exactly as written.
- Real estate idioms ("move-in ready," "fixer-upper," "turnkey") must be adapted for meaning, not translated word-for-word. Flag each with [IDIOM] and a one-line note on how you rendered it.
- Neutral, professional register — warm but not casual.
- Do not add, remove, or soften any information, especially about property condition or costs.
- After the translation, give an English back-translation of the 3 most consequential sentences so I can verify nothing shifted in meaning.

If any sentence involves legal or contractual obligations, mark it [LEGAL — HUMAN TRANSLATOR RECOMMENDED] rather than guessing.

Fill in your details and the prompt updates live — then copy.

Common questions from realtors

Can I use AI-generated listing descriptions in the MLS?

Generally yes — most MLSs treat AI copy like any other copy, and you're responsible for it either way. NAR's Code of Ethics (Article 2) prohibits exaggerating or misrepresenting property facts no matter what wrote them, and many brokerages now require human review of AI content before publishing. Check your MLS rules and your broker's AI policy first.

What's the Fair Housing risk with AI-written marketing copy?

AI tools regularly produce phrases like 'perfect for young families' or 'walking distance to church' that imply preferences tied to protected classes. The working rule is to describe the property, not the people. Read every AI draft against that standard before it goes anywhere public — you and your broker carry the liability, not the tool.

Do I have to tell clients I'm using AI?

There's no universal disclosure rule yet, but NAR encourages transparency and publishes AI policy templates for brokerages. The bigger obligation is confidentiality — don't paste a client's financial details, motivations, or negotiation position into a consumer AI tool, because that data may not stay private. Summarize in generic terms or use your brokerage's approved tool.

Which AI tool are most agents actually using?

ChatGPT, by a wide margin — 58% of Realtors in NAR's 2025 Technology Survey, versus 20% for Gemini and 15% for Copilot. Free tiers are enough to start. The tool matters less than the input: prompts that include your real property facts and MLS numbers beat any premium subscription.

Related professions