--- title: "Using AI to Optimize Your STR Pricing" canonical_url: "https://freewyldfoundry.com/podcast/ep731-ai-pricing/" content_type: "podcast-episode" last_updated: "2026-08-24T00:00:00.000Z" --- Episode 731 of the Get Paid For Your Pad podcast. YouTube: https://www.youtube.com/watch?v=Vv9tzrWsfVY Audio: https://embed.acast.com/5afc793a028014b853c89db4/6a8b047954f87b88769578e1?accentColor=161616&bgColor=e48f0c&secondaryColor=161616 Most STR operators are managing pricing the same way they did three years ago. Looking at graphs, checking a few price points, making adjustments when something catches their eye. For a handful of listings, that works. At any real scale, it doesn't. In this Rev Up episode, Jasper sits down at his laptop and shows you exactly how to use Claude to find pricing opportunities across your portfolio, from scratch, in under 30 minutes. No prior AI experience needed. Just a PriceLabs account and a question. **You will learn:** * Why AI should support a revenue manager's decisions, not replace them, and how Jasper's team uses it at Freewyld Foundry today * The problem with 365 price points per listing and why a human can only realistically review 5 to 10% per day * How to download your neighborhood data from PriceLabs and the exact chart options to have selected before you export * How to prompt Claude to identify dates where you are currently priced 20% or more above the highest price you have ever booked at * How to refine that analysis to compare like-for-like days: weekdays to weekdays, weekends to weekends, and special events to special events **We also talk about:** * How to flip the question and find nights where you are priced below what similar days have historically booked at * How to use pacing data to flag periods where the market is already getting bookings and your calendar has nothing yet * How to use Claude's voice input mid-session and how to teach it the output format you actually want * How Jasper's team runs this analysis automatically every day across the full portfolio **Mentioned in the Episode:** * PriceLabs * Freewyld Foundry Free Revenue Report Subscribe for new episodes every Monday on YouTube, Spotify, and Apple Podcasts. ## Transcript ## Why AI Won't Replace Your Revenue Manager At Freewyld Foundry, we do not believe that revenue management should be run by an AI agent or AI tool. We believe a revenue manager should leverage AI to help make better decisions. That's how we use it in our business. AI can do things a human cannot. It can scan thousands of price points and compare them against historical patterns, market occupancy, and pacing data faster than any person could. But a human still needs to look at the output, think about it, and decide what to do. Put them together as a team and you get better decisions than either one alone. That's the framework. Now let me show you exactly what it looks like in practice. ## The Problem With 365 Price Points Every listing has 365 price points per year. If you have 100 listings, that's 35,000 price points to manage. New information comes in every day: bookings, market pacing, competitor changes. As a human, you can realistically review maybe 5 to 10% of those price points on any given day. The rest go unreviewed. That's not a failure of effort. It's just math. No human can keep up with that volume of data at the speed it changes. AI can. That's the gap it fills. ## What "Optimal Pricing" Actually Means No one knows what the optimal price point is before a booking happens. There are too many unknowns. The person who books is the one who tells you, in hindsight, what the price could have been. What you can do is narrow the range. If your one-bedroom unit has never booked below $100 and never booked above $300, pricing it at $700 is almost certainly wrong. Pricing it at $10 is also wrong. The question is whether your current prices are inside a reasonable range where, if someone books, it's a win. Revenue management is really about scanning all 365 price points and asking: are we in the right range, or are we clearly outside it? ## What You Need to Get Started You need two things: a Claude account and your PriceLabs neighborhood data. That's it. In PriceLabs, go to your listing and open the Neighborhood Data tab. Under Chart Options, make sure you have all of these selected: my upcoming bookings, my last year bookings, my same time last year bookings, my two years ago bookings, last year data, pickup, and market booked price. Then click the download button and export as CSV. Open Claude. Type: "I need your help optimizing prices for my short-term rental unit. I will give you historic pricing data and occupancy data in the market. Are you ready for the data?" Upload the CSV. Then tell Claude to look at the data and wait for your instructions rather than asking questions. This keeps it from going in a direction you don't want. ## Finding Your Overpriced Nights The first question to ask is simple: are there any dates where we are currently priced at least 20% higher than the highest price we have ever booked at? This is a global ceiling check. It doesn't account for day of week or seasonality. It just asks whether any of your prices are clearly out of range relative to your own history. In a live example from one of our listings, Claude identified 7 nights within a few minutes. The highest price that unit had ever booked was $619. Twenty percent above that is $743. Seven nights were priced above that threshold, some by 45 to 66%. Those are nights that are almost certainly not going to book at the current price. ## Comparing Like-for-Like Days The global ceiling is a starting point. The next question gets more precise: are there dates where we are priced more than 20% above the highest price a similar day has ever booked at? A similar day means weekday versus weekday, weekend versus weekend, holiday versus holiday. You're asking Claude to compare apples to apples rather than comparing a Thursday in November to a Saturday in July. When we ran this on the same listing, the count went from 7 nights to 90. Claude built a full breakdown: 59 non-event weekdays, 17 Sundays, 14 weekend days. That's the kind of analysis that would take a full day to do manually. It took a few minutes. AI has a tendency to go further than you asked and build elaborate spreadsheets. Keep your questions simple and specific. It will add complexity on its own without you asking for it. ## Finding Your Underpriced Nights The same approach works in reverse. Ask Claude: are there any nights where we are currently priced lower than the lowest price a similar day has ever booked at? On the same listing, Claude flagged 64 nights. Some of those will be at or near your minimum price, where there's not much room to move. Focus on the ones where there's a gap, where you could raise the price and still be within a range that has historically booked. ## Using Pacing Data to Flag Booking Gaps Beyond price, you can also look at booking pace. Ask Claude to identify periods in the next 12 months where the market is already 10% or more booked and you have no bookings yet. This doesn't automatically mean you're overpriced. Maybe you don't need bookings that far in advance. But when the market is filling and your calendar is empty, it's worth looking at. Something might be off: your pricing, your minimum night stay settings, or in some cases a calendar that accidentally got blocked. On the listing in this session, Claude flagged 65 nights across 16 periods. The four biggest gaps ranged from 6 to 10 consecutive nights where the market was 15 to 25% booked and we had nothing. That's useful information to have in front of you every day. ## Teaching Claude What You Want One thing worth knowing: Claude gets better the more you use it on the same data. You can tell it exactly what format you want for the output. A PDF report, a simple list of the 10 most important price points, an Excel sheet organized by date range. Whatever is useful for how you actually work. You can also use voice input. If you're looking at your screen and want to ask a question, just click the microphone and talk to it like you would a colleague. The goal is to get Claude to the point where it's flagging the right things in the right format every day, so your revenue manager can spend time making decisions instead of hunting for the data. ## Wrapping Up In this working session, starting from scratch with one CSV file, we identified 7 clearly overpriced nights, 90 nights flagged on a like-for-like basis, 64 underpriced nights, and 65 nights where pacing raised a warning. All of it in about 25 minutes. That's what AI adds to a revenue management workflow: not decisions, but data, organized and delivered fast enough that the person making decisions can actually use it. If you want us to help you build this into a proper system, go to freewyldfoundry.com/get-started. We'll put together a free revenue report for your portfolio and show you where the biggest opportunities are. See you next week. --- ## About this file This is the agent-readable version of a page on freewyldfoundry.com. The canonical page for people is https://freewyldfoundry.com/podcast/ep731-ai-pricing/ Freewyld Foundry runs revenue management for short-term rental operators: pricing strategy, distribution, and the systems around them. We manage 3,500+ listings representing $170M+ in annual bookings for 70+ clients. Pricing is custom and scales with portfolio size. Related machine-readable resources: - Every page in this format: https://freewyldfoundry.com/agent/pages/index.json - Site overview: https://freewyldfoundry.com/llms.txt - Request a free revenue report: https://freewyldfoundry.com/get-started/