How to Create a Comparative Market Analysis (CMA) Using ChatGPT: Step-by-Step
%20Using%20ChatGPT%20Step-by-Step-clean.png)
A comparative market analysis (CMA) is one of the most important documents you'll produce as an agent — it's how you help a seller understand what their home is actually worth, and how you back up a buyer's offer with real data. Dedicated AI-powered CMA platforms have gotten popular over the past year, with some running $50-400+ a month for automated valuation reports.
If you're not ready for that expense, or you just want to understand the process better before automating it, you can build a genuinely solid CMA using ChatGPT and the data you already have access to through your MLS. It takes more manual steps than a dedicated platform, but it costs nothing beyond what you're likely already paying for ChatGPT Plus, and it teaches you the underlying logic that a "black box" AI tool hides from you.
What a CMA Actually Needs to Include
Before touching any AI tool, it helps to be clear on what makes a CMA useful in the first place:
- The subject property's key details (size, bedrooms, bathrooms, lot size, age, condition, notable features)
- 3-6 comparable properties that have sold recently in the same area
- Adjustments for differences between the subject property and each comp (square footage, upgrades, condition, etc.)
- A final suggested price range, with reasoning a client can actually follow
AI doesn't replace any of these steps — what it does well is speed up the writing, organizing, and explaining part, once you've pulled the raw data yourself.
Step 1: Pull Your Comps from the MLS First
This part can't be skipped or handed to a general AI tool — ChatGPT doesn't have live access to your local MLS data, so trying to ask it to "find comps" directly will produce made-up or outdated information. Pull your comparable sales the normal way, through your MLS, filtering for:
- Same neighborhood or a comparable one nearby
- Sold within the last 3-6 months (or wider if inventory is thin in your market)
- Similar size, bed/bath count, and property type
Aim for 3-6 solid comps. More than that tends to muddy the analysis rather than strengthen it.
Step 2: Organize Your Data Before Prompting
Once you have your comps, put the details into a simple list or spreadsheet: address, sold price, sold date, square footage, bed/bath count, lot size, and any notable condition or upgrade differences from the subject property. This is the raw material you'll hand to ChatGPT — the more organized it is going in, the better the output.
Step 3: Use ChatGPT to Calculate Adjustments and Draft the Analysis
With your data organized, you can prompt ChatGPT to do the heavy lifting on adjustments and narrative. A prompt structure that works well:
"I'm preparing a comparative market analysis for a real estate client. Here's the subject property: [details]. Here are 4 comparable sales: [list each comp's details]. For each comp, suggest a price adjustment based on differences in square footage, condition, and features compared to the subject property, and explain your reasoning in plain language a home seller could understand. Then suggest a final price range for the subject property based on the adjusted comps."
ChatGPT will walk through each comp, apply logical adjustments (for example, adding value for an updated kitchen or subtracting for a smaller lot), and produce a suggested price range with reasoning attached. This is also where it earns its keep as a tool for new agents specifically — because it explains its logic in plain language, it doubles as a way to learn how experienced agents think through adjustments, rather than just receiving a number.
Step 4: Sanity-Check Every Number Yourself
This step matters more than any other in this guide. AI models are good at organizing information and explaining reasoning clearly, but they're not connected to live market data and can occasionally apply an adjustment that doesn't reflect real local buyer behavior — for example, overvaluing a feature that doesn't actually move price much in your specific market.
Treat ChatGPT's output as a strong first draft of the reasoning and math, not a final answer. Review every adjustment against your own local market knowledge before it goes anywhere near a client.
Step 5: Turn It Into a Client-Ready Document
Once you're satisfied with the numbers and reasoning, ask ChatGPT to reformat the analysis into a clean, client-facing summary:
"Turn this analysis into a short, professional summary I can present to a home seller, explaining the comps, the adjustments, and the recommended price range in a way that builds their confidence in the number."
Paste that into your preferred document format (Canva, Google Docs, or your brokerage's template) alongside maps, photos of the comps, and your MLS data screenshots for a complete, presentable CMA.
When This DIY Approach Isn't Enough
To be direct about the limits here: if you're preparing CMAs constantly — several a week, across a busy pipeline — a dedicated paid CMA platform that pulls MLS data automatically and formats reports instantly will save you real time that this manual process can't match. The DIY approach in this guide is best suited to agents doing a handful of CMAs a month, or anyone who wants to understand the mechanics before paying for automation.
Final Thoughts
A comparative market analysis is ultimately a judgment call backed by data, and that judgment is still yours to make. ChatGPT can genuinely speed up the adjustment math and the write-up, but the comps you pull, the local knowledge you apply, and the final number you stand behind in front of a client are still the parts of the job that make you the professional in the room.
Also read:
