Why this matters
Sellers judge their listing agent by the weekly update. They want to know what buyers are saying, and they want it straight. But showing feedback arrives in pieces: a text from one agent, an email from another, a voicemail from a third. Pulling it together every Friday takes an hour.
And the one comment that matters most is the easiest to lose: a buyer’s agent saying their clients will write an offer this week and need one answer first. Miss that and you can miss the offer.
In this module you build an AI agent in n8n that turns a pile of feedback into one row per showing: interest level, what was said about price in the agent’s own words, and the next step. It saves every showing to a Google Sheet you can walk the seller through, and it never tells you to cut the price.
What you’ll build
An n8n workflow with a chat trigger, an AI Agent with a system prompt for Meridian Realty Group (a made-up brokerage), an OpenAI chat model, memory and a Google Sheets tool, plus a Showing Feedback Log sheet with five columns: Date, Buyer’s agent, Interest, Price feedback and Follow-up.
You paste in three showings for 14 Orchard Lane (made up) and get three entries: some interest with “about $15,000 high”, a buyer ready to write an offer who needs the roof age today, and a family that needs four bedrooms. Then the agent saves all three as rows.
What you need
- Nothing to start the practicen8n, OpenAI and Google Sheets are practice copies. No account is created and nothing is billed.
- An n8n account, to do it for realn8n Cloud has a free trial, or you can self-host. The steps are the same.
- An OpenAI API key with a few dollars of creditLogging a week of showings costs fractions of a cent.
- A Google accountFor the Showing Feedback Log sheet. n8n asks you to sign in to Google once.
- Your brokerage’s rulesWhat you may share with sellers, and what never goes in writing. The system prompt in the download is a starting point.
Concepts
3 ideas you’ll use in the practice.
The system prompt is the job description
It says who the agent works for, who reads its output, and exactly what one entry looks like: date, buyer’s agent, an interest level from a fixed list, the price comment in the agent’s own words, and one next step.
A fixed list (High, Some, None, Unclear) is what makes the sheet easy to scan. Vague or mixed feedback goes to Unclear with a reason, instead of a guess.
Record what was said, not what it might mean
Two rules in the prompt keep you out of trouble with sellers: record only what was said, and never recommend a price change. “About $15,000 high” is one buyer’s agent’s opinion, and the sheet keeps it that way.
The price decision belongs to you and your seller. The agent’s job is to make sure you have every comment in front of you when you make it.
Memory and the Sheets tool make it an agent
Memory lets you say “save those” after the agent sorts the feedback, and it knows which showings you mean. The Google Sheets tool lets it write the rows itself.
Without them, it is a chatbot that sorts text you then copy by hand. With them, the log fills itself.
Practice
Do every step yourself in a practice copy of the real tool. It checks your work, and “Do it for me” finishes any step you get stuck on.
Do it for real
Now repeat it in your own account. Tick each step as you go.
Downloads
- Showing feedback agent system prompt (.txt)The system prompt, the five column headers and the sample feedback from the lesson.
Quick check
3 questions to check you’ve got it.
1. A buyer’s agent says the home is “about $15,000 high.” What does the agent write in Price feedback?
- Recommend a $15,000 price cut
- “About $15,000 high.”, in the agent’s own words
- Nothing, price is private
Show the answer
“About $15,000 high.”, in the agent’s own words. The prompt says record only what was said and never recommend a price change. That decision is yours and the seller’s.
2. Why does Interest use a fixed list (High, Some, None, Unclear)?
- n8n only accepts four values
- So the sheet is easy to scan and vague feedback is flagged instead of guessed
- It is shorter to type
Show the answer
So the sheet is easy to scan and vague feedback is flagged instead of guessed. Fixed values make the weekly update quick to read. Unclear, with a reason, is better than a guess.
3. What lets you say “save those” and have the agent know which showings you mean?
- The chat model
- Memory
- The chat trigger
Show the answer
Memory. Memory keeps the conversation, so the agent knows what “those” refers to before it uses the Sheets tool.
You reached the end of AI for Real Estate
Your showing feedback now logs itself, ready for the seller’s weekly update. Pick your next course, or point the same agent at your next listing.
Pick your next course →