Here's a number that should stop you: according to the NAR's 2025 Technology Survey, 68% of real estate agents report using AI in their business. Only 17% say it's made a significant difference.
A separate RPR survey of 225 agents found the same pattern from a different angle: 82% say they use AI tools, but only 34% say it has meaningfully changed how they work. (Source: RPR 2026 AI Adoption Survey, n=225) Two surveys, two sample sizes — same finding.
Those agents have access to ChatGPT, Gemini, Claude — the same tools reshaping how other professionals work. The gap between "using AI" and "getting results from AI" comes down to one thing almost every agent is skipping.
They're not loading context before they prompt.
When you open ChatGPT and type "write a listing description for a 3-bedroom home in Kauaʻi," you're asking a tool that knows nothing about you, your market, your voice, or your client to produce something that sounds like you, speaks to your specific buyers, and differentiates your listing from every other property on the market.
It can't. So it produces something that sounds like every other listing description it's ever seen.
The NAR's 2025 Technology Survey found that 46% of agents who use AI report no noticeable business impact. Same tools, completely different results — because the gap isn't the tool, it's the setup. (Source: NAR 2025 Technology Survey, n=1,241)
This is the actual problem. Not the tool. The setup.
Before you write a single prompt, the AI needs four things: your market (location, buyer profile, price range, what makes your area different from the national average), your voice (how you write, what phrases your clients recognize, what you'd never say), the specific property brief beyond the MLS fields (the emotional hook, the likely buyer, the objections worth addressing), and your exclusion list (vague lifestyle language, fair housing red flags, the phrases that show up in every other listing in your market).
That's the context block. It goes before every prompt — not in the prompt itself. Without it, the AI defaults to the national average. With it, the output reflects your market and sounds like you. The setup takes about 20 minutes to build the first time. After that, it's a copy-paste.
I broke down exactly how to build this — the four components, the step-by-step workflow, and the one document that makes it repeatable — in How to Cut Listing Description Time from 45 Minutes to 15. That post has the full how-to. This one is about why the gap exists.
I use the same setup for every listing description I write on Kauaʻi. Before I type a single instruction, I load a context block covering my market, my voice, my buyer profile, and my compliance checklist. The context takes about two minutes to paste and review. Writing the property brief, running the prompt, reviewing the output, and doing the fair housing check takes another 10 to 12 minutes.
Total time: around 15 minutes for a listing description that sounds like me, speaks to my buyers, and rarely needs more than light editing. Compare that to the industry average of 30–45 minutes for a manual draft. (Source: KapRE.com, 2025)
The AI didn't get better. The setup did.
Read through the full workflow before your next listing appointment. It covers exactly what to include in your context block and takes about five minutes. Then build it. The first time takes 20 minutes. Every listing after that, it's a copy-paste.
That's what the research confirms the 17% are doing differently. Not a different tool. A better setup.
Want to see what these workflows are worth in your business? Run the AI ROI Calculator →
Every week: one AI tool, one workflow, tested in an active real estate practice — so you know it works before you spend time on it. Free.
No hype, no homework. Unsubscribe anytime.