AI on the Record · evidence register

Real estate · Listing descriptions · Case

Brad Cohen: listing descriptions, then the full set of listing marketing material with ChatGPT → custom GPT

Shares numbersYouTubepublished result self-reported

Brad used to sit down with his notes from the listing appointment and the MLS sheet and spend about half an hour on each description. The first time he tried AI, he pasted those same notes and the property details into ChatGPT and asked for a listing description. Reported result: ~30 min → ~2 min per description; 7 listing deliverables in minutes (self-reported, not verified).

Shares numbers · self-reportedYouTube

Brad Cohen

Realtor and owner, Connexa Real Estate · Harrisonburg, Virginia · two-person office

Tool ChatGPT → custom GPTPublished 2025-09-03Sells anything? none foundResult self-reported

Brad Cohen walking across a stage at a business event; the slide behind him shows social, email and video icons and reads 'Seven Deliverables' and '10 Hours'.Play from 5:08
Screenshot from Brad Cohen's YouTube talk at 5:08, where he describes the custom GPT's seven deliverables. Source ↗
  1. Appointment notes + MLS sheet
  2. ChatGPT, later a custom GPT
  3. Description + 6 more pieces
  4. Quick round of edits
Steps as stated in the video.
Task
Listing descriptions, then the full set of listing marketing material.
Tool
ChatGPT at first. Later a custom GPT built inside ChatGPT.
What they did

Brad used to sit down with his notes from the listing appointment and the MLS sheet and spend about half an hour on each description. The first time he tried AI, he pasted those same notes and the property details into ChatGPT and asked for a listing description. The draft was not in his voice, so he did one quick round of edits. After that he and his operations manager built a custom GPT that takes one listing and returns the description, the social media posts, an email blast, and a video script with a shooting plan. He describes it as seven deliverables from one input.

Result

~30 minutes → ~2 minutes per listing description, after edits. The full set of listing material now takes minutes instead of hours.

Self-reported by Brad Cohen; not verified.
Where it goes wrong
The first draft "wasn't exactly my voice." He edits every output before it goes out.
Conflict of interest
None found. He speaks as a local business owner at a community event; he does not sell a course or tool.
Source
YouTube Connexa Real Estate – Brad Cohen, Realtor · talk at a local business event · published 3 Sept 2025
It was good. Like too good. Now, it wasn't exactly my voice, but after a quick round of edits…

Published 2025-09-03 · rewritten in our words from the video · #brad-cohen

More from this task

  • Shares numbers2023TikTok2023-05-21

    @kalachejackson_realtor

    The agent copies all the property information from a Zillow page, pastes it into ChatGPT and types "Write an MLS description for this house." If the first draft is too basic the agent clicks "regenerate response" and takes the second one.

    ChatGPT (2023) · Draft in under 30 seconds · self-reported

  • CautionTikTok2026-10-14

    @chriswhartonrealtor

    The agent took real photos of the kitchen and backyard from one of his own recent listings and asked ChatGPT to "improve" them, then to go further.

    ChatGPT (images) · No number; a warning

  • Free download at endYouTube2026-09-04

    Broker-owner, The List Realty

    The broker-owner walks the home with the seller and records every observation and everything the seller says, using Minutes AI.

    ChatGPT + Minutes AI · No number given for listings

Related: All listing descriptions cases · ChatGPT for realtors · Real estate report

← Back to Listing descriptions · How we collect cases