Loading…
5
Asked an AI to categorize 400 real estate leads and found the one word that broke it
Last month I dumped 412 old leads into an AI tool and told it to sort them into hot, warm, and dead. It bunched almost everyone into warm, which was useless. So I started reading the notes it wrote for each lead and noticed it kept flagging people who said things like "we're hoping to move before the school year." I tested it on 50 leads where I already knew the answer from my own call logs, and it nailed 44 of them. Just 44 out of 50, but that beat my old paper system where I forgot half these people existed. The trick was renaming the categories. Hot became "has a date," warm became "no date," dead became "never answered twice." The tool stopped guessing and started sorting on facts. Has anyone else had to dumb down the labels for these things to get useful output?
0 comments
Log in to join the discussion
Log In0 Comments
No comments yet
Be the first to share your thoughts on this discussion.