Setting hotel room rates: the cost floor, the comp set and the channel effect

Most independent hotels change one thing about their room rates each year: they add a few percent to last year's number and leave the rest alone. Nobody in the building can usually say what it actually costs to sell one specific room on one specific night, what the five or six hotels down the road are charging for the same date, or what a guest actually sees when they search that date on a results page. The rate exists because it existed last year, not because anyone worked it out.

TL;DR: Setting a room rate properly starts with three numbers most hotels have never calculated: the cost floor (the rate below which a sold room actually loses money, once you count what that stay costs to deliver), the comp set (the four to seven hotels a guest is realistically comparing you against, chosen or not) and the channel effect (the same headline rate is worth a different amount once a platform's commission comes out of it). This article works through the cost floor with exact worked arithmetic, how a comp set should inform a price rather than dictate it, the seasonal and day-of-week shape hotels flatten out of habit, and the one check worth doing every month: what your own rate looks like on the results page a guest actually sees.

This is written for the owner or general manager of a roughly 50 to 200 room independent hotel who sets rates or reviews them. It is a working framework, not a pitch for software to run the calculation automatically.

The cost floor: the rate below which a sold room genuinely loses you money

Every hotel has a cost floor. On the hotel accounts we run, it is rare to find one that has actually worked theirs out. The floor is the rate at which selling a room stops making money and starts costing it: the point where a stay's cash no longer covers what that stay costs to deliver. Sell below it, knowingly or not, and the room would have been better left empty.

The floor is built from variable costs only, the costs that exist because this specific room was sold to this specific guest on this specific night, not the costs you carry whether it sells or not. Rent, most salaried staff, insurance and maintenance decide whether the hotel as a whole is profitable, but they do not change from room to room, so they have no place in the floor.

Here is a worked example, illustrative only, not a real hotel's figures. Say the variable cost of one room for one night comes to £32: £11 housekeeping labour, £7 laundry, £3 guest amenities, £7 for a complimentary breakfast, £4 consumables and utilities (11 + 7 + 3 + 7 + 4 = 32). When it is sold direct, with nothing taken out by a third party, £32 is the floor. Price that room at £38 direct and you clear £6 of contribution above it. Price it at £30 and the room sold for less than it cost to deliver, even though it looked like revenue on paper.

The channel effect: why the same £38 is not the same £38

Sell the identical £38 room through a platform charging 20% commission instead of direct, and the arithmetic changes completely. The hotel receives 80% of £38, which is £30.40, a full £1.60 below the £32 floor. The room sold and the guest stayed, but the hotel lost money on that stay while the calendar showed a booking.

To earn the same £6 of contribution through that channel, the gross rate has to rise to £47.50: the £32 floor plus the £6 target, divided by 0.80. That is a 25% higher headline rate than the £38 that clears the same margin sold direct, for the same room and the same night. Commission rates vary by platform and by market rather than sitting at a fixed 20%; how much OTAs actually charge hotels works through the real range. The 20% here exists to make the arithmetic exact, not to claim what you personally pay.

A rate that clears your floor on one channel can sit under it on another at the exact same headline number. Price without knowing which channel a booking is likely to come through, and you are guessing at the one number that actually decides whether the sale was worth making.

Comp-set reality: the neighbours you did not choose but are compared against anyway

A competitive set, comp set for short, is the small group of hotels a guest is weighing you against on a given date: similar location, room count and class, similar guest profile. STR, the data company that popularised the term, defines it as typically four to seven properties; most independent hotels are read against a set that size whether they built one deliberately or not.

Building one deliberately is worth the afternoon it takes: hotels a guest would genuinely cross-shop against you, not hotels you admire or hotels twice your size. One or two slightly above your own class are fine if you want to shift where guests place you over time, but keep the core set honest, the places actually competing for the same booking on the same night.

The comp set does not replace the cost floor; it sits above it. The floor is the number below which a sale is a loss; the comp set is where, above that floor, you can credibly sit: matching the middle, undercutting it deliberately to win on price, or sitting above it because your review score, location or room quality earns the premium.

Seasonal shape: building the curve in advance instead of discovering it through discounting

Demand for a hotel room does not move in a straight line across the year, and treating it as though it does is where independent hotels quietly lose the most money. A rate held flat all year earns too little in the genuinely strong weeks and still fails to fill the genuinely weak ones, because it was never right for either.

The fix is a seasonal rate card built in advance: peak weeks identified from your own historical occupancy, priced at what the market will bear; shoulder periods priced to hold occupancy without giving away margin unnecessarily; low periods priced deliberately, as a decision, rather than discovered three days out through a panic discount. Revenue management as a discipline is built on exactly this kind of decision made ahead of the date, not reacted to on the day.

The card does not need to be complicated. It needs to exist, get reviewed against what actually happened, and get adjusted on real signals rather than left to drift.

Day-of-week shape: the pattern most independent hotels flatten by habit

Within any given week, demand also has a shape, usually running in the opposite direction depending on the kind of hotel you run. A city hotel serving business travel tends to fill Monday to Thursday and empty at the weekend; a leisure destination tends to do the reverse, quiet midweek, full Friday and Saturday. Both patterns are real and both are usually ignored in favour of one rate that tries to serve every night equally and serves none of them well.

Building day-of-week multipliers off your own occupancy history, not an assumption borrowed from a different kind of property, closes some of the same gap at a shorter interval. Say a Tuesday consistently sits at 55% occupancy while Saturdays run at 95%, figures for illustration only: those two nights should not be wearing the same rate, and checking that they are not is a short exercise once the data is in front of you.

Rate and occupancy only tell the story together: a high rate on a night nobody books and a low rate on a night that would have sold anyway both leave revenue on the table, in opposite directions, which is exactly what RevPAR is built to catch and a rate alone cannot.

Check your own rate on the results page, monthly, because that is where it is actually set

Whatever rate strategy you build in a spreadsheet, the number that matters is the one a guest actually sees, on the results page, next to your comp set on that date. Booking.com's own guidance to partners lists price competitiveness against comparable properties as one input to how prominently a listing is shown, alongside review score, availability and cancellation rate. Your rate is more than a number you set. It also decides whether a guest sees you at all.

The check is simple and most hotels never do it. Once a month, search your own hotel on dates you expect to be busy and dates you expect to be quiet, exactly as a guest would, and see where you land against the comp set on that page. A rate that looked sensible in a spreadsheet can still sit oddly against current competitor pricing that moved since you last checked.

This is the rule worth keeping: the results page is where your price is actually set in the guest's mind, not the number in a spreadsheet.

The anchoring failure mode: pricing off a calendar that no longer describes this year

One of the failure modes we see most often is anchoring: taking last year's calendar, adding a percentage across the board, and calling it this year's rates. It is fast, and it is wrong whenever demand has actually moved.

An event that filled the city last September gets cancelled or moves to another town, and the rate built around it is now pricing a Tuesday with none of the extra visitors it once had. A competitor with dozens of comparable rooms opens two streets away, and the comp set that priced fine last year has one more seat at the table, softening demand for everyone already in it. Neither shows up if the only input to this year's rate is last year's number plus inflation.

The fix is not complicated; it just has to actually happen: before the calendar is set, check whether anything that drove last year's demand is still true, and check the comp set for new supply.

When dynamic pricing sophistication is the wrong tool

A genuinely sophisticated, algorithm-driven pricing setup is real and it does real work for some hotels. For a property that has not yet done the basics above, though, it is the wrong thing to reach for first. Sophisticated pricing optimises against a floor, a comp set and a demand pattern; without those, it optimises against nothing.

If you do not know your own cost floor, cannot name your comp set, and have not checked your listing against it in the last month, that is the actual gap, not a lack of pricing intelligence. Revenue management is a set of decisions, not a piece of software, and the decisions have to exist before anything can automate them well. Fix the counting first: the floor, the comp set, the seasonal and day-of-week shape, the monthly check. Sophistication is worth adding once those are in place, when you can tell, concretely, what a more advanced system would actually be optimising.

If your cost floor, comp set and current OTA visibility are things you are guessing at rather than checking, a free audit is where to start. It looks at your actual numbers, not a generic template.