How to Build an STR Comp Set: A Revenue Manager's Guide

Let me show you what a casual comp-set habit actually costs.

Picture a host who runs five properties in a single sub-market. Every Sunday morning she opens Airbnb, scrolls through nearby listings, eyes a few rates, and calls it “comp set work.” Her properties feel correctly priced. Her calendar feels healthy. Then the local half-marathon weekend rolls around. A 2BR two streets over (same bedroom count, similar photos, no obvious advantage on paper) pulls a booking worth thousands more than hers for the same dates.

Let me show you what actually happened. The problem wasn’t her property, her photos, or her price. What she called “checking the comp set” was casual browsing from the seller’s seat. She was looking at properties she thought she competed with, not the properties guests were actually choosing between. Those two lists barely overlap, and the difference between them is where most STR revenue quietly disappears. Hotels have measured the same gap. A 2018 Cornell University thesis compared the competitive sets hotels reported to Smith Travel Research with the hotels TripAdvisor shoppers actually compared, and according to its abstract it found a low match between the two.

A comp set isn’t a habit of glancing at neighbors. It’s a discipline. Done right, it tells you where you sit in a guest’s decision tree, what your real ceiling looks like, and which lever (price, minimum stay, length-of-stay discount, gap fill) is the one to pull this week. Done wrong, it produces a calendar that feels full and a P&L that quietly under-earns.

This guide walks through how a revenue manager actually builds and maintains a comp set: the criteria every comp has to pass, the seven-step framework, the data points that matter, the mistakes that turn a good set into a stale snapshot, and the cadence that keeps it useful.

“Casual browsing from the seller’s seat isn’t a comp set. It’s a habit. The discipline is reading the set guests are actually shopping.”

Federico Zimerman

What Is a Comp Set? (And What It Isn’t)

The term “comp set,” short for competitive set, comes from hotel revenue management, where it’s been the central instrument of yield management for roughly forty years. Hotels build comp sets primarily by chain scale and brand tier, because in a world of consistent flags those categories are reliable proxies for who the guest is choosing between.

Short-term rentals don’t work that way. There’s no chain scale, no Marriott-Hilton-Hyatt grid. What defines competitive position in STR is amenity granularity (hot tub, EV charger, dedicated workspace), design sensitivity (mid-century modern versus rustic cabin), and sub-market boundaries (East Nashville and downtown Nashville don’t share guests, despite the shared label). The hotel framework adapts; the principles transfer; the inputs don’t.

The hotel rules are stricter than most hosts assume. According to CoStar’s STR Benchmark competitive set guidelines, a hotel competitive set must include at least 4 participating properties besides the subject hotel, and no single property can account for more than 50% of the set’s room supply. That rule exists so one outlier can’t define the benchmark, and the same logic drives the 8 to 15 band we use for short-term rentals.

In short-term rentals, a comp set is the curated group of 8–15 properties a host benchmarks against to inform pricing, minimum-stay, and inventory decisions, chosen because they represent the alternatives a guest would realistically book if they didn’t book yours. This is distinct from a hotel comp set, which is typically defined by chain scale and brand tier rather than guest substitution behavior.

What a comp set isn’t: it isn’t your neighborhood. It isn’t every listing in your ZIP code. It isn’t the properties you wish you could compete with. And it isn’t fixed. The discipline is in the rebuild as much as the build.

Why Most STR Comp Sets Are Built Backwards

Most STR comp sets fail for a single reason: they’re built from the seller’s seat instead of the buyer’s seat. Hosts start with their own property’s attributes (2BR, hot tub, mountain view) and work outward to every nearby property that matches. That produces a list of properties that look like yours. It doesn’t produce a list of properties guests compare yours against.

There’s research behind this. A 2011 Cornell study by Jin-Young Kim and Linda Canina compared hotel competitive sets built by product type with sets built by clustering on ADR. The ADR-based sets showed less RevPAR variability, and most of them mixed product types, which the authors read as evidence that competition happens across product types.

Borrow the airline frame. When an airline builds a competitive set for a route, it doesn’t ask, “Which carriers fly aircraft like ours?” It asks, “Which carriers does our customer compare us against when choosing this flight?” The aircraft is an input. The buyer’s decision set is the output.

Here’s the thing. Pricing isn’t an attribute of the property. It’s a read on the buyer. Get the buyer wrong, and the price is always going to be wrong, no matter how clean the comp list looks on paper.

“Pricing isn’t an attribute of the property. It’s a read on the buyer. Build the comp set from the buyer’s seat, and every other decision gets easier.”

Federico Zimerman

Worked example: a design-forward 2BR cabin in the Smokies

The seller’s-seat comp list pulls every 2BR cabin within a twenty-minute drive: rustic A-frames, traditional log cabins, midcentury renovations, modern builds. Twelve properties, clustered around a mid-tier ADR for the sub-market.

The buyer’s-seat comp list looks different. The guest booking this cabin is choosing design-forward stays in the Smokies. The relevant comps are the four other design-forward cabins in the area, plus a handful of modern 1BR cabins one tier down that this guest type cross-shops, plus two boutique inn rooms with a similar aesthetic. Eight properties, sitting at an ADR roughly 30% above the seller’s-seat list.

The lists barely overlap. Pricing off the first list anchors the property well below where the buyer is actually shopping. The second list is the one that informs pricing. Airbnb says the same thing from the platform side: according to its Help Center article on how search results work, the algorithm compares a listing’s total price with comparable listings that share characteristics such as guest capacity and amenities.

The 7 Filters Every Comp Has to Pass

Every property in your candidate pool has to pass these seven filters to make the comp set. Most candidates fail at least one. That’s the point. For a sense of how many attributes a comp can differ on, Wheelhouse’s Dynamic Sets tool lists 18 filter categories on its pricing page, including amenities, operators, occupancy and cleaning fees, at $12.99 per set per month as of September 2026.

1. Bedroom and bathroom count

The hardest filter and almost always the first to disqualify. A 2BR and a 3BR don’t compete for the same guest, even when price ranges overlap. Bathroom count matters nearly as much: a 2BR/2BA and a 2BR/1BA serve different group dynamics. Match exactly, or don’t include the comp. According to AirDNA’s Vacation Rental Glossary, size is one of the tests in a comp set definition, next to location and amenities, and bedroom count is the cleanest proxy for size.

2. Maximum guest capacity and sleeping configuration

A 2BR that sleeps 4 and a 2BR that sleeps 8 (via bunk beds and a sleeper sofa) aren’t comparable. Travel groups search by guest count first and bedroom count second. Match capacity and the configuration that produces it. Airbnb names guest capacity as one of the characteristics its search algorithm uses to judge whether a price is attractive against comparable listings, per Airbnb’s Help Center.

3. Sub-market location, not city

Downtown Nashville and East Nashville share a city name and almost nothing else. Guests choosing downtown want walkable bars; guests choosing East Nashville want neighborhood coffee. They are not the same comp pool. Define the sub-market by guest behavior, not municipal lines. According to Airbnb’s own search documentation, listings in places guests like to stay, such as near popular landmarks, tend to rank higher, so the platform itself sorts by behavior before it sorts by city.

4. Amenity tier

A hot tub, a private pool, pet-friendly, an EV charger, a dedicated workspace: each amenity reshuffles the pool. Two properties identical on every other criterion but split on hot tub presence are in different comp sets during most of the year. Guests filter for amenities, so should you. The Airbnb Help Center confirms guests can filter homes by type of place, price range, amenities and booking options, per its article on search results.

5. Property type and design aesthetic

A design-forward modern rental and a traditional family cabin in the same town, with the same bedroom count and amenities, are not comps. They serve different guests. Aesthetic is a competitive lane. The guest booking the modern rental will cross-shop other modern rentals before the cabin three doors down. Keep one caveat from the 2011 Cornell research by Kim and Canina in mind: their hotel data showed competition crossing product types when ADR was similar, so treat aesthetic as a separate lane only when the price band also splits.

6. Listing quality and photography tier

Controversial but real: conversion rate is part of competitive position. A poorly photographed property that’s technically your match isn’t competing with you on the search results page. Including it drags the benchmark down. A comp has to be an alternative the guest actually considers. According to the Airbnb Help Center, higher-quality listings with better ratings and reviews tend to rank higher in search, and quality includes photos, videos and listing content.

7. Price range overlap

If a candidate’s ADR is roughly 30–40% off yours in either direction (adjust to your market’s elasticity), guests aren’t comparing the two properties; they’re shopping different price bands. A $200/night 2BR isn’t competing with a $500/night 2BR, even when every other criterion matches. Include candidates in your reasonable price band; exclude the bargain bin and the tier above.

How to Build Your STR Comp Set: The 7-Step Comp-Set Build

The 7-Step Comp-Set Build is the sequence we run at RevFactor whenever a new property onboards, and the same sequence we run quarterly across the 198 listings in 67 markets and 24 U.S. states on RevFactor’s managed book, as published on the About page. Seven steps. Each step has a single job. Most operators try to skip Steps 1 and 5, which is why most comp sets read like neighbor lists. The same sequence sits inside the Discover phase of The RevFactor Method.

Step 1. Define your property’s category honestly

Strip aspiration. Price for the property you actually run, not the property you wish you ran. A renovated 2BR with old furniture and decent photos isn’t a design-forward stay. Define category in the words a guest would use, not the words you’d use in a pitch deck. According to AirDNA’s Vacation Rental Glossary, benchmarking means comparing your rates, occupancy and revenue with similar properties that compete for the same guests, and an honest category is what makes “the same guests” true.

Step 2. Map your sub-market boundaries using guest behavior

Look at where similar properties cluster in actual booked stays, not on a map but in the search filters guests apply. If guests filter for “downtown” and your property gets pulled into that filter, you’re in the downtown sub-market for comp-set purposes. Guest behavior defines the boundary. Airbnb notes in its Help Center that the listings shown on its search map may differ from the listings shown in the results list, which is one more reason to trust booked stays over a map radius.

Step 3. Pull a wide initial candidate pool

Pull roughly 30–50 properties in most markets, adjusted up for high-density urban markets and down for thin-supply leisure markets. Use AirDNA, PriceLabs market dashboards, Beyond, Wheelhouse, or a manual Airbnb search filtered to your dates and guest count. Start wide so the seven filters do the narrowing. According to PriceLabs’ pricing page, a ready-to-view Market Dashboard costs $9.99 per month for a market of up to 1,000 listings, and Wheelhouse sells comp-set building as Dynamic Sets at $12.99 per set per month, both checked live on September 14, 2026. PriceLabs is the pricing engine RevFactor runs daily; Beyond and Wheelhouse are reasonable alternatives, and the three are compared feature by feature in PriceLabs vs Wheelhouse vs a done-for-you service.

Step 4. Apply the seven criteria as filters

Run the candidate list through the filters one by one. Most candidates fail at least one. You’re not trying to keep properties in the set; you’re trying to identify the small group that passes all seven. Be ruthless. Even hotel benchmarking enforces composition rules at this stage: CoStar’s guidelines cap any single brand at 50% of a competitive set’s room supply and any single company at 70%.

Step 5. Validate with booking data, not just listing data

A property that looks like a comp on paper but books at half your occupancy isn’t in your tier. Pull occupancy and pacing data on the filtered set. Drop properties that consistently book below your performance band. They’re not competing for the same guest, regardless of what the photos suggest. Booking data is the harder evidence: Key Data built its benchmarking business on verified, first-party reservation data sourced directly from operators, and says it covers more than 10 million vacation rental properties.

Step 6. Lock the set at 8–15 properties

Smaller than 8 and noise dominates any individual rate move. Larger than 15 and the data smears. Eight to fifteen is the band where the signal is strong enough to act on and small enough to monitor. For comparison, CoStar’s STR Benchmark rules set the hotel floor at 4 participating competitors, at least 3 of them unaffiliated with the subject hotel.

Step 7. Set a refresh cadence of quarterly minimum

New supply enters every market. Properties exit. Renovations move properties between tiers. A comp set that worked in Q1 may be meaningfully wrong by Q3. Across the 198 listings under RevFactor revenue management, which run +24% RevPAR against comp set on a rolling basis per our own first-party reporting, we run this rebuild quarterly as a baseline, with off-cycle rebuilds whenever a sub-market shifts.

“The work is the filtering. Pulling 30 candidates is easy. Defending the 8 that survive is the discipline.”

Federico Zimerman

The Data Points That Actually Matter (and the One That Doesn’t)

1. ADR (Average Daily Rate)

Useful as context, dangerous in isolation. A property at $1,000 ADR with 20% occupancy is losing to a property at $400 ADR with 70%. Treat it as one input, never the headline. AirDNA defines ADR as the average revenue earned per booked night, which is exactly why it can’t see the nights that never sold. See our ADR vs RevPAR guide for the full breakdown.

2. RevPAR (Revenue Per Available Night)

The comp set’s primary benchmark. A higher rate with low occupancy can produce a lower RevPAR than a lower rate with strong occupancy, and RevPAR is the number that lands in the operator’s pocket. Benchmark weekly against the set. According to AirDNA’s Vacation Rental Glossary, RevPAR is the metric that combines nightly rates and occupancy to show actual income per available night.

3. Occupancy rate

Only meaningful when paired with ADR. A 90% occupancy at $80/night is a worse business than 50% at $300/night. Read it alongside ADR and RevPAR, never alone. According to AirDNA’s definition, occupancy rate is the share of available days a property is booked, typically measured over 12 months, so blocked owner nights change the denominator.

4. Pacing (booking window data)

How full are the comps 30, 60, 90 days out? Pacing tells you whether the market is filling early or late. Weekly pricing decisions earn their money here. According to AirDNA’s Vacation Rental Glossary, pacing is a comparison of current booking activity against past performance, tracking reserved nights, rates and how far ahead guests are booking. More on this in our ADR vs RevPAR field guide.

5. Length-of-stay patterns

Average length of stay tells you whether the market rewards short-stay flexibility or long-stay minimums. A market averaging two nights wants 1- and 2-night flexibility; a market averaging five rewards a 3-night minimum priced for longer stays. It’s the bridge between the comp set and minimum-stay strategy. The long tail matters too: according to the Airbnb Help Center, many guests on Airbnb book stays of 28 nights and longer, so a comp that accepts monthly stays is reaching demand your set may not see.

6. Minimum-stay restrictions

A live competitive lever, covered in the next section. Tracking what the comp set is doing on minimum stays (and where the gaps are) is how you find pricing lanes guests can’t compare you against. According to PriceLabs, a 3-night minimum means a listing won’t show up for guests searching 1-night or 2-night stays, and 85% of its users run minimum-stay restrictions.

7. Review velocity and rating

Review count and recent rating directly influence Airbnb ranking and pricing power. A property with 200 reviews at 4.95 sits in a different competitive tier than the same property at 12 reviews and 4.7. Trailing review counts is a leading indicator of competitive position erosion. Per the Airbnb Help Center, a home’s average overall rating only appears in search results once at least 3 guests have rated it, so a comp with no rating yet is competing with one hand tied.

A rate without pacing context is data without meaning.

How to Use Your Comp Set in Daily Pricing Decisions

Benchmark, don’t copy

A comp set tells you where the market is, not where you should be. Two properties in the same comp set can rationally hold different prices on the same night based on review count, listing strength, and pacing. Use the set to locate yourself; don’t chase the median. According to the Airbnb Help Center, the search algorithm deliberately encourages variety, showing guests listings with different hosts, different characteristics and a range of prices, so the median is never the only rate that wins a search.

Pacing reads

Compare your pacing to the set’s pacing weekly. If the set is 70% booked at 30 days out and you’re at 40%, the question isn’t “should I drop price,” it’s “what’s different.” Sometimes it’s price. Sometimes it’s photos. Sometimes it’s a minimum stay killing search visibility. According to the AirDNA glossary, pacing means current bookings read against past performance, which only works if the comp set you read it against is current. For the deeper pacing read, see our ADR vs RevPAR breakdown.

Gap detection

Weekday gaps, where comps are full and you aren’t, are the highest-value tactical opportunities in the calendar. The market has demand; your property is priced or positioned wrong. A small adjustment (a length-of-stay discount, a 2-night minimum drop, a gap-fill discount on a Tuesday) captures it. PriceLabs, in its orphan gap guide updated September 5, 2026, applies a 20 percent discount to gaps of two nights or less by default, which is a starting point to override, not a strategy.

Minimum-stay positioning

When most of your comp set sits at 3-night minimums, dropping to a 2-night minimum opens a search lane your competitors can’t reach. The opposite is also true: when comps are at 1-night, a 3-night minimum during peak weekends can capture longer stays at premium rates. The comp set tells you which lever to pull. According to the Airbnb Help Center, the more flexibility a host offers around how long guests can stay, the more likely the listing is to show up in search results.

Event-based adjustments

Watch how the set reacts to local events. Some markets price aggressively 90 days out. Some wait until 14 days. The set’s pricing curve on a known event is how you learn the market’s demand signature. The same read informs the dynamic pricing guide for STR beginners: the comp set tells the tool when to lean in. PriceLabs makes the same point in its minimum-stay guide: during major events, it may be worth relaxing gap prevention to grab premium rates.

The 5 Mistakes Most Hosts Make Building a Comp Set

1. Survivorship bias

Most hosts build their comp set from properties currently active on Airbnb, which means they’re looking only at the listings that survived. Properties that delisted, switched to long-term, or stopped competing don’t show up in the data, but they should inform it.

There’s an old story from World War II that captures this well. According to the American Mathematical Society’s Feature Column, the story traces to the Statistical Research Group at Columbia University in 1943, and it has a solid germ of truth but little evidence for its most dramatic details. In the popular version, Allied engineers studied returning bombers, mapped the bullet holes, and recommended reinforcing where damage clustered. A statistician named Abraham Wald said: you have it backwards. The planes that got hit in the engines and cockpit are the ones that didn’t return. Reinforce where you see no damage on the survivors. Your active comp set is the bombers that came back. Pay attention to the silence too.

2. Aspirational comping

Benchmarking against properties two tiers up produces pricing the market won’t support. The properties in your comp set should be the ones a guest realistically considers alongside yours, not the properties you wish you competed with. Aspirational comping doesn’t raise your rates; it lowers your bookings. Airbnb’s search documentation is blunt about the mechanism: per its Help Center, listings priced below comparable listings in the area tend to rank higher, so pricing against a tier you don’t belong to pushes you down the page.

3. Sample-size errors

Five comps is too few; any outlier swings the median. Forty is too many. The data smears across multiple buyer types. Eight to fifteen is the band where the signal is strong enough to act on. Bigger isn’t more accurate; it’s noisier. Hotel benchmarking guards the small end with a rule: according to CoStar’s guidelines, no single property can account for more than 50% of a competitive set’s room supply.

4. The stale-set trap

A static comp set is stale within four to six months. New supply enters, old supply exits, renovations shift properties between tiers. The trap is comfort: the set still has the same property names on it, so the work feels done. The market underneath those names has already moved. Quarterly refresh is the minimum. New listings also ramp faster than they used to: Airbnb reports on its new listing promotion page that listings published in 2025 with the promotion saw more than 70% more bookings over their first three months than listings without it.

5. Ignoring guest perception

Ignoring guest perception is the mistake behind the other four. A comp set built from your property’s attributes is half the picture. The 2018 Cornell thesis on hotel competitive sets found exactly this split: customers and hotels disagreed on how much price, size, class, distance, rank and score should matter. A comp set built from the guest’s decision set is the other half, and the one that informs pricing. Hosts who skip the buyer’s-seat check spend years priced against the wrong set, wondering why the strategy never works.

When to Refresh or Rebuild Your Comp Set

Refresh triggers (quarterly cadence is the minimum):

• Property changes: renovation, new amenity, listing relisting, photography update.
• Market changes: new supply enters the sub-market, a major comp exits, demand shifts.
• Performance diverges from the set without an obvious reason (your pacing pulls away from the set’s pacing).
• A new season starts. Shoulder-to-peak transitions almost always demand a comp set check.

Hotel benchmarking builds new supply into the rules. CoStar’s competitive set guidelines let a newly constructed hotel join a set within 5 months of opening, and require any set change to add at least 2 properties that have been open for 5 months.

Rebuild triggers (rare, but decisive):

• You’ve moved tiers. A renovation lifted the property into a new bracket, and the old comp set is no longer the right buyer’s set.
• The sub-market itself has shifted. New attraction, new infrastructure, new short-term-rental regulation, a major new development.

A refresh updates the set. A rebuild starts over from Step 1. Know which one you’re doing. The cheapest way to watch a sub-market shift is a market-level view: PriceLabs lists a ready-to-view Market Dashboard at $9.99 per month for markets of up to 1,000 listings, checked September 14, 2026.

“The build is the easy part. The discipline is the quarterly rebuild and the weekly read.”

Federico Zimerman

What a Comp Set Is Really For

A comp set isn’t a tool that lives in a spreadsheet. It’s the input layer that feeds the rest of revenue management.

Federico’s three strategic pillars (Interest, Reliability, Positioning) all read through the comp set. Interest asks whether your listing converts at the level the set suggests it should. Reliability asks whether your review velocity is keeping you in the same competitive tier. Positioning asks whether your daily pricing is reading the set’s pacing correctly. For the full Three Pillars treatment, see the revenue management for short-term rentals pillar guide, or the procedural detail in The RevFactor Method.

The comp set is the lens. Without it, every pricing decision is a guess; with it, every decision is a read of the market. The 2011 Cornell research by Kim and Canina put it in hotel terms: comparing competitor groups built two ways shows how closely a property’s intended position matches its position as customers see it. The discipline isn’t building the set. It’s reading it every week.

The Bottom Line: Build the Comp Set From the Guest’s Seat

The whole exercise comes back to one shift in perspective: stop building the comp set from your property outward, and start building it from the guest inward. A comp isn’t a property that looks like yours. A comp is a property the guest is actually choosing between when they consider yours. That’s also how AirDNA frames benchmarking: a comparison with similar properties that compete for the same guests.

It’s also not a one-time exercise. The build is the easy part. The discipline is the quarterly rebuild and the weekly read: comparing your pacing to the set’s pacing, watching for gaps, tracking who entered the market and who exited, and adjusting the levers accordingly.

This is the daily exercise RevFactor runs for STR operators across the country. Federico is also the founder of Blackbird Hospitality, a property management firm operating 198 listings across 24 U.S. states and 67 markets. RevFactor applies the same frameworks to outside operators who keep their own property management.

If you want to walk through your comp set with someone who builds them daily, start a conversation.

Frequently Asked Questions