On Tuesday I told you Claude just hit #1 in the App Store, and you should give Cowork a try. The next day, Todd Vitolo, a great NY agent, subscriber and student of an early co-hort sent me this:
"Matt I’ve been playing around with Cowork and it is pretty insane. I downloaded sales data from Compass for the past 12 months for my neighborhod and just messing around came up with this."
That's 491 transactions across Tribeca, visualized in an interactive dashboard. Monthly pricing trends, days on market, building-level breakdowns, fastest sales, biggest discounts, and a scatter plot showing how pricing correlates with selling speed. He also built a neighborhood heat map. The kind of deliverable you'd expect from a research team, not a single agent messing around on a Tuesday.
Todd used the $100/month Max plan to get this done. You can do it on the $20/month Pro plan too, though heavier datasets may hit usage limits faster.
Here's how to build your own.
What you need:
A CSV or Excel export of recent sales from your brokerage platform (Compass, KW Command, your MLS, wherever you pull data). The Claude desktop app with Cowork enabled. A paid Claude plan.
Step 1: Export your data.
Log into your brokerage platform or MLS. Pull closed sales for the last 12 months in your target neighborhood. Download it as a CSV or Excel file. The more fields the better: address, close price, list price, days on market, square footage, bedroom count, property type, building name, close date. Most platforms give you all of this in a standard export.
Step 2: Create a neighborhood folder.
On your desktop, make a folder. Name it something like "Tribeca Market Data" or whatever neighborhood you're targeting. Drop your exported file into it. Open the Claude desktop app. Click the Cowork tab. Point it at this folder.
Step 3: Give it one prompt.
If you want a detailed version use this:
I have a CSV of residential sales data for [your neighborhood] in this folder. Analyze the data and build me an interactive HTML dashboard that includes: monthly sales volume and median price trends, price per square foot over time, days on market distribution, breakdown by bedroom count and property type, the most active buildings with median stats, a table of the fastest sales, properties that sold above asking, and the largest discounts from list price. Make it visually polished with dark mode styling and insight callouts beneath each chart explaining what the data means. Save it in this folder.
If you're new to Cowork or working with a smaller dataset, start with this simpler version:
I have a CSV of residential sales data for [your neighborhood] in this folder. Build me a clean dashboard that shows: median price and how it's trending, average days on market, a chart of sales by month, and a table of the 10 most recent sales with address, price, and days on market. Keep it simple and professional.
Either way: one file in, one prompt.
Step 4: Come back to a finished dashboard.
Cowork will read your data, run the analysis, generate the charts, write the HTML, and save it to your folder. Open the file in any browser. You now have an interactive market dashboard for your neighborhood.
Want to share it as an attachment instead of a link? Open the HTML file in your browser, hit Ctrl+P (or Cmd+P on Mac), and choose "Save as PDF." Now you have a polished report you can email directly.
Step 5: Use it to create value.
This is where most agents will stop. Don't. Here's what to do with it:
Screenshot the key charts and drop them into your next listing presentation. Email it to past clients in that neighborhood as a quarterly market update. Post the headline stats on social media with a "DM me for the full dashboard" call to action. Bring it to your next seller consultation and walk through the pricing trends in person.
That last one matters more than you think. AI search tools are starting to recommend agents who publish original, hyperlocal data. When someone asks Claude or ChatGPT "who are the best agents in [your neighborhood]," the ones who have this kind of content indexed online are the ones getting mentioned.
One thing to know: Cowork works with local files on your computer. It can't log into Compass or your MLS and pull the data for you. You need to do the export yourself. That's 2 minutes of work.
-Matt
March 19th from 12 to 1pm ET, Skye Michiels and I are running the final session of our free training series on using AI to be more human. We're building a 12-month sphere follow-up plan. Sign up here.
If you want to go deeper on workflows like this, my next AI Marketing System cohort is only 2 weeks away! It starts March 23rd and we’ll be diving deeply into Claude Cowork. Here is a special 10% off for subscribers.
