Your comps plus a disciplined adjustment ledger, condition read, and a defensible range, with every assumption visible enough to survive a sharp client's questions.
The idea
What this does
You already pull the comps; the MLS is yours. What this adds is the part that takes the evening: a line-by-line adjustment ledger, a condition-aware read of each comp, micro-trend context, and a written narrative that explains the range in language a client can follow and push back on. Use it to pressure-test your own number before the listing appointment or the offer conversation. It is deliberately a second opinion, not a replacement for your judgment.
Gather these first
Before you start
Six to twelve comps from your MLS: address, status, list and sold price, dates, beds, baths, square footage, lot, and your one-line condition note per comp
The subject property's details and honest condition
Optional but powerful: paste your MLS export text directly, or attach comp photos for a condition read
Step by step
How to use it
Start a new chat named after the subject property.
Copy the prompt, paste your comp data where marked, and fill in the subject details.
Challenge what comes back: "why only 40k for the extra bedroom" is exactly the conversation this is for.
Ask for the client-facing narrative once the numbers feel right to you.
Copy everything in the box
The prompt
Paste this into your AI toolFill anything in [BRACKETS] with your details
You are a valuation analyst working for me, a residential real-estate agent. I will provide the subject property and comparable sales from my MLS. Build a disciplined comparative market analysis.
SUBJECT: [ADDRESS, BEDS/BATHS, SQFT, LOT, YEAR, CONDITION in your honest words, anything special: view, ADU, busy street]
COMPS (my data, my condition notes):
[PASTE YOUR COMP LIST OR MLS EXPORT HERE]
Produce:
1. COMP QUALITY CHECK. Rank my comps by how comparable they truly are (distance, recency, size, condition, lot, setting). Flag any that should be dropped and say why. Tell me what an ideal missing comp would look like so I can go find it.
2. ADJUSTMENT LEDGER. For each kept comp, adjust line by line (size, condition, lot, garage, view, timing). State the dollar logic for every adjustment; if photos were attached, use them to sharpen the condition adjustments and quote what you saw.
3. MICRO-TREND READ. From the dates and prices in my data only: which way this segment is moving, days-on-market pattern, list-to-sold behavior. Do not invent market data I did not give you; if the sample is too thin to say, say so.
4. THE RANGE. A supported value range with your confidence level, the two or three comps doing the real work, and what would move the number up or down.
5. STRATEGY NOTES. If I say this is for [BUYER OFFER or LISTING PRICE], add the strategic read: positioning, price-band thresholds, and appraisal risk if we stretch.
6. CLIENT NARRATIVE. A one-page plain-English explanation of the range that I can walk a client through, assumptions visible.
Rules: never output a single point value, always a range with reasoning. Use only the data I provided plus attached photos. No demographic commentary of any kind. Where my condition notes and the photos disagree, ask me.
→Anything in [BRACKETS] is yours to fill in. Delete a bracket line if it does not apply;
the prompt still works.
The result
What you'll get back
An adjustment ledger you can defend line by line
An honest comp-quality audit, including the comp you should go find
A supported range with named drivers, not a mystery number
A client-ready narrative that makes you the transparent one in the room
Level it up
Make it yours
Buyer side follow-up: "given three likely competing offers, how would you position ours and where is the walk-away?"
Listing side follow-up: "argue the case for listing at the bottom of the range versus the top, one paragraph each."
Attach photo sets from two comps and ask which condition adjustment changes.
Keep it honest
Your MLS data stays yours: check your MLS terms about where exports can be used, and share the analysis, not the raw feed.
The output is decision support built on the comps you chose; garbage comps make confident garbage. The comp-quality section exists to catch that.
Appraisers and markets get the final vote; treat the range as preparation, not prophecy.