NASHVILLE, Tennessee — As with everything else, context is king when setting up a hotel's competitive set.
Comp sets are a long-standing tool for hotel revenue managers to figure out how their properties are performing relative to the market and the other hotels in the area. Just because the tool is old, however, doesn't mean hoteliers have to do it the same way every time.
When setting up a comp set, hoteliers typically include a set of four to six properties, said Hannah Smith, senior analyst at STR during the "Build better benchmarks" session at the 2026 Hotel Data Conference. They typically range plus or minus half a class from their own hotel. The properties are usually less than 2 miles from the subject property, and they have an average daily rate variance of about 16.4%.
This is the average comp set, but within the averages, there are a lot of variabilities, Smith said. Within the CoStar product, hoteliers can create as many comp sets as they want, and many are making diverse options.
Creating additional comp sets can help hoteliers define who they think they’re competitive with, she said.
“That’s where you may see less of that variation compared to what we see with performance comp sets,” she said.
The geographical size of a comp set will vary based on where the subject hotel is located as well as what type of hotel it is, Smith said. A midscale hotel will have many other hotels as potential competitors, but a luxury resort may have to expand its radius to 10, 20 or even 50 miles to find other hotels that are competitive.
The ADR range can also vary depending on the type of hotels being compared, with a wide range possible when looking at luxury hotels, she said.
“Again, you have a smaller set of hotels that you can compare against if you're a luxury property, so you're probably being a bit more flexible in who you're identifying based on that price point,” she said.
One of the older tools in figuring out comp sets was giving them different letter grades as a way to evaluate how close the comp set is to the subject property, Smith said. Those with “lower grades” are more dissimilar to the property, but that doesn’t necessarily mean they aren’t useful because they may provide other context.
“Those ones in the D’s and F’s, it doesn’t mean they’re a bad comp set,” she said. “It just means these are comp sets that are not evaluating what we’re grading for.”
About 65% of all the comp sets around the U.S. are running a revenue per available room index plus or minus 20% from the subject property, Smith said. That means the majority are fairly close in performance, but there are another 35% that are much further afield.
In a look at a sample of anonymized real comp sets, she said one shows a hotel with two comp sets. It’s a limited-service property in a top 25 U.S. hotel market, specifically in a downtown central business district. The first comp set is geographical, evaluating the immediate area around the hotel, while the second is hotel-class-based.
The first one includes many properties that are actually in a lower class than the subject property, and there’s a slight room count variance, she said.
“I think where you really start to understand the strategy of these comp sets is when you look at the RevPAR index,” she said. “That second comp set is 100% RevPAR index, so it's as comparable as you can get to that subject property. The top one: 73%.”
It’s not that one of these is a better fit than the other, because each serves a different purpose to the hotelier setting them up, Smith said. The first one here was the local comp set to better understand what’s happening in the immediate area, to understand what’s bringing travelers to the 0.3-mile radius around the subject hotel. The second one shows how other similar hotels are performing, even if they’re farther away.
In another example, an extended-stay hotel near an airport in a secondary market has two comp sets, she said. The first comp set has a wider radius of about 8.5 miles while the second is tighter. The first has all properties in the second class — extended-stay hotels — while in the second, there’s only one other hotel in the same class.
One is a comp set helping the hotelier understand what’s happening in the immediate market around the airport while the other looks at the other extended-stay properties the subject hotel is competing against in the overall market, she said.
Smith said she came up with four ideas of what a comp set can do and what purpose it can serve. The first is the local comp set to help hoteliers understand what’s happening directly around their hotel. The next is a key performance index comp set set that aligns as closely as possible with their property. Another is service level that looks at the operating structure while the last looks at amenities offered.
“Using the four of these comp sets together gives you a more holistic look at what are all of the aspects of my operation doing,” she said. “How successful are they at the different parts of the business?”
Another option is to set up comp sets using the data set hoteliers are actually using, she said. A top-line performance comp set may not be the best comp set when looking at day-of-week swings, and the monthly data may look great, but it might not align with the actual strategy of the hotel and how it targets a particular customer base.
“You have this top-line comp set basically running at 100% index of each of the KPIs, but when you look at the comp sets that are built for other purposes, that are built to align on day-of-week strategy, that are built to align off profitability, they’re not always going to get you that 100% RevPAR index, but they're going to give you more information about your strategy than just that top line,” she said.
There are other options out there, too, Smith said. A hotel with higher occupancy on the weekends than it does on weekdays didn’t see comp sets align with it until the comp sets look directly at hotels with similar occupancy patterns.
“If you take one thing away, it’s that one single comp set isn’t necessarily going to be the answer to all of the ways that you’re benchmarking,” she said. “So, let’s rethink what a comparable hotel is.”