Your competitive set is the denominator of every performance conversation you will ever have. RevPAR index — your RevPAR divided by the set's — is how owners judge the hotel, how bonuses get paid, and how your pricing gets second-guessed. A well-chosen set makes the index a diagnostic instrument; a badly chosen one makes it a flattering mirror. Here's how to build one that tells the truth.
Benchmarking services (STR/CoStar being the de facto standard) aggregate your competitors' actual occupancy, ADR and RevPAR and report them back as a blended set average, against which your share is indexed. An index of 100 means you're getting your fair share; 110 means you're taking demand from the set; 90 means they're taking it from you. Every number in that sentence is only as meaningful as the set behind it. The set also quietly shapes daily decisions: it's the group of hotels you rate-shop, the reference for "priced with the market," and the baseline for every displacement and promotion argument.
STR enforces competitive-set rules built for data confidentiality, and they constrain your design: a set needs a minimum of three participating hotels besides yours (in practice four or more, so no single competitor's data can be reverse-engineered — sets also fail if one hotel dominates the room supply, and no more than a defined share of rooms may come from a single brand or ownership). Not every hotel reports to STR — an obvious comp that doesn't participate simply can't be in the set. Build with five to seven hotels where possible: enough that one comp's renovation or a non-reporting month doesn't whipsaw the average.
Most hotels run one primary set for the owner report. Sophisticated operators run a second: a secondary set for a different question — the aspirational set you're repositioning toward, the rate-shopping set (which can include non-STR hotels and short-term rentals), or a submarket set for the neighborhood while the primary covers the citywide class. The discipline is labeling: one set is the accountability number; the others are navigation instruments. Mixing those roles is how hotels end up grading themselves against the set they can beat.
Every revenue manager eventually sits in a meeting where someone proposes "refreshing" the comp set right after a bad quarter. The classic patterns, all of which produce a lying index: stacking weak hotels (adding tired, lower-rated properties to inflate the index past 100), dropping the winner (removing the comp that outperforms you instead of learning from it), and the post-renovation dodge (cutting a comp the moment its renovated product starts beating you). The tell is always the same: the index improved while actual RevPAR didn't. Owners' asset managers know these moves, and a set that flatters you costs your credibility precisely when you need it — in the budget meeting.
Sets should be reviewed annually and changed rarely, for supply-side reasons, not score-side reasons: a genuine new competitor opened; a comp converted brands, repositioned, or left the market; your own repositioning (post-renovation, new brand) changed who you trade guests with; or a comp stopped reporting. When you do change it, restate history — compare this year against last year on the new set, or the trend line means nothing.
Downtown, leisure-weighted weekends, corporate weekdays, no meeting space to speak of. Candidates: two upscale branded hotels sharing your block and your corporate accounts (in — the core), a lifestyle boutique 400 meters away trading weekend guests with you (in), a select-service midscale property that undercuts you by 35% (out — different booking), a 450-room group-house convention hotel (out — different business), an upper-upscale independent you lose shoppers to at +15% price (in), and a soft-branded newcomer opening in March (watch list — add at its first anniversary, once it's ramped). Result: a five-hotel set with real guest overlap, one aspirational edge, and no group-calendar noise.
The Rate Strategy Pack includes the comp-set selection worksheet as a working spreadsheet — candidate scoring across the five criteria, price-band checks, and the weekly rate-shop log that keeps the set honest after you've chosen it. The weekly shopping routine itself is part of the daily RM routine, and the free calculators cover the index math.