AI prompts for real estate agents
Most AI prompts for real estate agents fail for the same reason: they ask for a listing description without giving the model the property, the buyer, or the rules you are legally bound by. You get four paragraphs of "charming" and "must-see" that you would never put your licence behind.
The prompts below hand the model the address details, the buyer it is writing to, and the fair-housing limits it must respect. That is what turns a generic paragraph into copy you can paste into the MLS.
What these prompts fix
- Listing copy that sounds like every other listing in the market
- Cold leads that never get a second touch because writing the follow-up takes twenty minutes
- Neighbourhood content you keep meaning to write and never do
Listing description that sells the lifestyle
Turn a fact sheet into MLS copy without fair-housing problems.
You are a listing copywriter for a residential real estate brokerage. Property facts: - Address / area: [neighbourhood, city] - Beds / baths / square footage: [numbers] - Year built and recent updates: [list] - Three features a buyer would pay extra for: [list] - One honest drawback: [e.g. small yard, busy street] - Price: [amount] Target buyer: [e.g. first-time buyer, downsizing couple, investor] Write the MLS description in 150-180 words. Lead with the single feature that justifies the price. Describe the property and the location, never the type of person who should live there — no references to family, children, schools as a selling point, religion, ethnicity, or "safe neighbourhood". Address the honest drawback in one line that reframes it factually. End with a specific next step. Rules: no filler, no hype adjectives, no exclamation marks. Write at a plain eighth-grade reading level. If you need a detail I have not given you, ask me for it instead of inventing it.
Seven-day lead follow-up sequence
Re-open a conversation with a lead who went quiet.
You are writing follow-up messages for a real estate agent to a lead who enquired and then went silent. Lead context: - How they found me: [portal enquiry, open house, referral] - What they asked about: [property or question] - Days since last contact: [number] - Their likely hesitation: [price, timing, financing, still browsing] - My market's one useful current fact: [e.g. average days on market, rate movement] Write four messages: day 1, day 3, day 7, day 14. Each under 70 words, written for a phone screen, no subject-line gimmicks. Each message must give one piece of information they did not have before — never "just checking in". Vary the ask: one question, one useful link, one specific time offer, one graceful close-out. Rules: no filler, no hype adjectives, no exclamation marks. Write at a plain eighth-grade reading level. If you need a detail I have not given you, ask me for it instead of inventing it.
Neighbourhood guide that ranks
Build a farming page for the area you want to own.
You are writing a neighbourhood guide for a real estate agent's website. Neighbourhood: [name, city] Price range of homes: [range] Dominant housing stock: [e.g. 1960s ranches, new townhomes] Five real local places I can name: [list — shops, parks, transit, landmarks] What locals complain about: [honest answer] What I know that a search engine does not: [your on-the-ground observation] Write a 700-word guide with these sections: what the area actually feels like, housing stock and price reality, getting around, the trade-offs, and who tends to buy here and why. Use only the local places I listed — do not invent businesses, schools, statistics or crime figures. Where a number is needed and I have not given it, write [VERIFY] instead of guessing. Rules: no filler, no hype adjectives, no exclamation marks. Write at a plain eighth-grade reading level. If you need a detail I have not given you, ask me for it instead of inventing it.
How to use them
- Copy the whole prompt. The structure is doing as much work as the wording.
- Replace every bracketed placeholder with real details from your business. A vague input is the single biggest cause of generic output.
- Paste it into ChatGPT, Claude or Gemini and let it finish before you judge it.
- Give one round of feedback naming what to keep, what to cut and what to change, then ask for the revised version.
Questions real estate agents ask
Can AI write MLS listing descriptions?
Yes, as long as you give it the real property facts and hold it to fair-housing language. Describe the property and location, never the buyer's family status, religion or ethnicity. Always read the output before it goes on the MLS — you are the one holding the licence.
Will AI-written real estate content hurt my SEO?
Not if it is accurate and specific. What gets penalised is thin, interchangeable text. A neighbourhood guide built from places you can actually name outperforms a generic one every time, which is why these prompts force you to supply the local detail.
Which AI tool is best for real estate agents?
ChatGPT, Claude and Gemini all handle these prompts well. The tool matters far less than the prompt — the same request written two different ways produces wildly different results in all three.
These prompts still need you to fill in the brackets. The Prompt Engineer asks you the questions and writes the finished prompt for you.