How to build a hotel comp set (and when to change it)

Ask most independent hotel owners how they built their competitive set and the honest answer is: they didn't. STR assigned one when the property first started reporting data, a general manager two owners ago tweaked it once, and nobody has opened the list since. That's a problem, because every benchmark that set produces — RevPAR index, ADR position, occupancy share — is only as good as the hotels sitting inside it. Nobody would run a budget off a spreadsheet they'd never double-checked, yet that's exactly how most comp sets get treated. Learning how to build a hotel comp set on purpose, and knowing when to change it, is how you stop steering your pricing off numbers that were never really about your property.

What is a hotel comp set, and why might yours be wrong?

A hotel competitive set — comp set, or compset — is the small group of properties that guests genuinely weigh against yours before booking, and that STR, rate shoppers, and lenders use as the yardstick for your performance. It is not a wish list, a proximity list, or whatever an onboarding form auto-populated years ago.

That distinction matters because a comp set quietly becomes load-bearing. Your RevPAR index gets discussed in owner meetings. A lender pulls your STAR report before renewing a loan. A revenue-management consultant builds a pricing recommendation on top of it. All of that sits on a foundation that, for most independent and small-group hotels, was set once and never revisited — which means the numbers can look wrong for years before anyone asks why.

The fix isn't complicated. It's an audit most owners have simply never run.

How to build your comp set, one hotel at a time

Start with a shortlist of 8 to 12 candidates: your current comp set, plus whatever shows up next to you on OTA searches and metasearch (Google Hotel Ads, TripAdvisor) for your typical dates and rate tier. Then run every candidate through four questions before it earns a permanent spot. The test works because it forces you to think like the guest doing the comparing, not like the owner doing the wishing.

1. Would a guest deciding between hotels realistically shortlist it alongside yours? Not "would I like to be compared to it" — would an actual traveler booking your market put the two side by side.
2. Does it price in your band on a normal week? A property that's routinely well above or below you on rate isn't a peer; it's a different tier wearing a similar star rating.
3. Is it the same kind of stay? Same scale, same service style — full service against full service, boutique against boutique — not just the same neighborhood.
4. Can your front desk or GM name it without being asked? If a guest has never once volunteered "I was also looking at ___" about that hotel, it's aspirational, not real. This is the cheapest, most honest check you have, and it's the one owners skip.

Put your current set and every OTA-search candidate into one table and score them this afternoon:

Property	Same tier/type?	Prices in your band?	Same catchment?	Guests mention it?	Keep?
Rival A	Yes	Yes	Yes	Yes	Keep
Rival B (upscale property across town)	No	No — well above your rate	Yes	No	Drop

Anything that fails two or more questions comes out. Anything new that passes all four goes in. That's the whole method — no software required, and it takes less time than the STAR report you're about to read with this new set.

If a candidate doesn't survive the test — the new luxury build getting buzz downtown, or a scrappy budget property two blocks over — don't discard the curiosity along with the hotel. Keep a short secondary watch list for properties you want visibility on without letting them skew the primary set your RevPAR index is actually calculated against. That's also the honest place to put a hotel that STR or a predecessor assigned you that fails the test today but might earn its way back in after a renovation or a rate change.

How many hotels belong in a comp set?

There's no single magic number, but if you benchmark through STR there is a hard floor. STR's competitive set guidelines require a minimum of four participating properties besides your own, at least three of them unaffiliated with you, and at least two unaffiliated companies represented — and no single property can account for more than 50% of the set's participating room supply. Sets that break those rules can have their data suppressed or be deleted outright, so that part isn't advice; it's the price of admission.

Above the floor, the published guidance lands in a neighborhood rather than on a figure: Hotel Tech Report puts a typical set at four to 10 properties, Lighthouse at five to 10. Our own working target inside that is five to eight for most independents. Fewer than five and a single outlier — one hotel running a fire-sale rate or sitting closed for renovation — can swing your whole index. More than 10 and you're spending analysis time on properties that were never really pulling guests away from you in the first place.

If you're inheriting a set with three hotels in it, that's a red flag twice over: the sample is too thin to trust, and it sits below what STR will report on. If you're inheriting one with 15, someone was optimistic rather than selective.

Rural and small-secondary markets are the honest tension here. If your town genuinely has three hotels worth comparing against, padding the list with a fourth from 45 minutes away tells you less than an honest three does — but an honest three won't clear STR's minimum either. The workable answer is to keep both: the tight, genuinely comparable group you actually price against, and a compliant set for official reporting — with the difference between them written down, so nobody a year from now mistakes one for the other.

When should you change a hotel's competitive set?

Treat the comp set like a lease you renew, not a tattoo. Hotel-tech guides converge on the same cadence from two different angles: reassess it on a fixed schedule at least twice a year, whether or not anything obviously changed. Markets drift quietly — a property repositions its marketing without a visible renovation, a management-company change shifts who they compete for, and the set that was accurate in January is wrong by summer.

Beyond the calendar, a handful of events should trigger an off-cycle review immediately:

• A comp-set property renovates, rebrands, or changes flag.
• A new hotel opens in your catchment — evaluate it against the four-question test as soon as it's booking, not after it's already taken a chunk of your business.
• A comp-set hotel closes permanently or stops reporting.
• Your own hotel repositions — a renovation, a flag change, new ownership — which can shift who your real peers are as much as anything a competitor does.

None of these need to wait for the twice-a-year date on your calendar. A comp set is a working document, not a form you file once and forget.

Let your RevPAR index tell you when the set itself is wrong

RevPAR index — often written RGI — divides your RevPAR by your comp set's RevPAR and multiplies by 100. A score of 100 means you're capturing your fair share of the market's revenue relative to that set; well above 100 means you're outperforming it, and well below means you're not.

Most owners read RGI purely as a verdict on their own hotel. It's worth reading it as a possible verdict on the comp set too. If your index has sat well above 100 for a year and nothing about your rates or product has changed, ask whether you built a soft comp set — properties that are technically similar but consistently underperform, which flatters your numbers without meaning much. If it's sat stubbornly below 100 despite genuinely competitive rates and a full calendar, check whether the set drifted upscale or aspirational without anyone renaming it that. A comp set that quietly gets easier or harder to beat produces a RevPAR index that quietly stops meaning anything — the four-question test above is exactly how you catch it.

Keep the comp set honest between reviews

A twice-a-year review catches structural change, but the events that should trigger an earlier look — a permit filed for a new build down the street, a competitor's brand flip, a renovation announcement — usually surface in local news and trade coverage weeks before they show up in STR data or an OTA listing. Waiting for the official numbers to move means you're always reacting a cycle late.

That gap between the market changing and the report finally showing it is exactly what a daily briefing is built to close. The Intel Club watches your competitors, your market, and your local signals and opens each morning with a recommended action, so a rebrand or a groundbreaking two blocks away can reach you the week it happens rather than the review after next. Membership is $99/month with a 7-day trial, and what a daily briefing covers for hotels shows the shape of it. For the report your comp set feeds, how to read an STR report and tracking competitor hotel rates without a revenue team go a level deeper.

Four questions, one table, twice a year on the calendar — build the list on purpose, and every number it produces gets more honest.