Price elasticity for hoteliers: what happens to bookings when you change the rate

Price elasticity is how much the number of room-nights sold moves when the rate moves. For a hotelier the useful version is revenue, not the ratio: a 10 percent rise pays for itself if you lose fewer than about 9 percent of room-nights. Elasticity differs by segment, date and channel.

A general manager raises the best available rate by 10 percent on a Tuesday, watches reservations slow for ten days, and puts the rate back. The question nobody asked is whether the slowdown came from the rate or from the calendar. After reading this you will be able to work out the exact number of room-nights a rise can afford to lose, read your own booking data correctly before and after a change, and record the result so the next decision starts from evidence.

What does hotel price elasticity actually mean?#

It is the relationship between a change in room-nights sold and the change in rate that caused it, read on the same dates and the same segment. If a 10 percent rise is followed by a 5 percent fall in room-nights on comparable dates, the ratio is about 0.5.

A ratio above 1 means room-nights fell proportionally more than the rate rose, so revenue fell. A ratio below 1 means the rate gain outweighed the volume loss. The textbook version assumes nothing else changed. In a hotel something else always changes: a fair, a competitor's promotion, a flight schedule, a school break, the weather.

The hotelier's version is revenue, not the ratio. Revenue equals rate multiplied by room-nights, so a rise only hurts when the room-night loss is larger than the price gain. There is no single elasticity for a property. It differs by segment (corporate account against leisure guest), by lead time (early booker against last-minute booker), by date band (a sold-out Saturday against a soft Tuesday) and by channel.

In plain words: demand is inelastic when guests keep buying after a rise, and elastic when they disappear. Most hotel demand is inelastic on some dates and elastic on others, which is why one number cannot describe a year.

One measurement rule before anything else: count room-nights, not reservations. A group of twenty rooms for one night and a couple staying two nights are both one booking, and they are not the same amount of business.

How much can you raise the rate before revenue falls?#

There is an exact break-even. After a rise of p percent, revenue holds as long as you lose fewer than p divided by (100 + p) of your room-nights.

The arithmetic: new revenue equals the old rate times (1 + p) times room-nights times (1 minus the loss share). Revenue holds when those two factors cancel, which gives a tolerable loss of p divided by (100 + p). For example, a 40-room-night block at 100 brings 4,000. Raise the rate to 110 and the break-even is 36.4 room-nights: 37 room-nights bring 4,070 and are better, 36 bring 3,960 and are worse. Losing three of forty is acceptable; losing four is not.

The tolerable loss is a share of room-nights on the same date band, not a share of reservations and not a share of the month. Applying it to the whole month hides the exact dates where the business disappeared.

Revenue is not contribution. A room sold at a low rate still pays for cleaning, laundry, breakfast and commission; a room left empty pays almost nothing. That makes the contribution break-even slightly more forgiving than the revenue break-even, which is why a small volume loss can still be the right decision.

RevPAR, revenue divided by rooms available, is the summary figure that carries both the rate and the occupancy effect in one number, so it is the natural check after a change. The rule tells you what you can afford to lose. It does not tell you what you will lose. That is what the next two sections are for.

The figures are illustrative, at a rate of 100 per room-night. One room fewer in each row turns the verdict.

What can your own booking data show you?#

Your own data can show how room-nights on the books moved after a rate change, provided you compare the same arrival dates at the same number of days before arrival.

Booking pace, in plain words, is the room-nights already on the books for a future arrival date, read at a fixed number of days before that date. Compare day 60 with day 60, not this Tuesday with last Tuesday. Pull the data by arrival date, by segment, by channel and by rate code, so that a movement can be attributed to something rather than to the month.

Record the rate actually paid, not the published rate. Discounts, package inclusions, loyalty reductions and commissions all change the real number. Keep the comparison window the same length and the same weekday mix, because a window with five Fridays in it is not comparable to one with four.

Write down what else changed in the window: a competitor closing for renovation, a trade fair, a new flight route, a channel promotion, a holiday that moved. State the hypothesis and the threshold before the change. A result that is interpreted after the fact is not a test, it is a story. For context on which properties you should be watching while you do this, see choosing a hotel competitive set and comparing rates fairly.

Why do segments react differently to the same rate change?#

Each segment compares your rate to a different reference, so the same rise is invisible to one guest and decisive to another.

The reference price is what the guest compares you with. A leisure guest compares your rate with the same dates last year, with the hotel next door and with the price seen last week. A corporate account compares it with a contracted rate and a travel budget.

Early bookers tend to be sensitive to the total holiday cost; last-minute bookers tend to be sensitive to availability. A rise inside the early-booking window removes a different guest from a rise in the last seven days. The mechanics of those windows are worth understanding separately; see early booking and non-refundable rates.

Groups and long stays need a displacement test: does the group's total contribution, including food, beverage and meeting space, beat the transient business it replaces? The group's room rate alone does not answer that.

Channel changes the comparison. The same net rate looks different once commission, a package or a loyalty discount is applied, so a rise on one channel can shift demand to another channel rather than remove it. What is included changes the experiment: breakfast, a transfer, spa access or a flexible cancellation term make a rise a different offer, not the same room at a higher price.

Fences let you keep the price-sensitive guest and the higher rate at the same time. A flexible rate that rises while a cheaper non-refundable option stays open is a targeted move, not a general one. How to structure those fences is covered in fenced discount campaigns for hotels.

How do you read a rate change before and after?#

Fix the comparison before you touch the rate, then read the same dates at the same lead time and change one thing at a time.

Choose the date band and the segment first. A specific Tuesday in a specific segment gives a readable answer; the whole month gives an average of several different experiments. Record the baseline: room-nights on the books at a fixed lead time, the average rate paid, and the revenue for those dates.

Change one rate, on one channel, for one segment, for a defined window. Everything else stays where it was. Read again at the same lead time after the change, then read once more at a later lead time to see whether the business was lost or merely delayed.

Separate delay from loss. A booking that moves from day 60 to day 20 is a pace change and the room-night still exists; a room-night that never appears at any lead time is a loss. Close the loop with a written decision and a review date, so the next change starts from a recorded result rather than from memory.

Which pitfalls make a rate change look bigger than it is?#

Events, channel promotions, mix shifts and small samples can each produce a movement that has nothing to do with your rate.

Calendar effects come first: a religious or public holiday that moves between years, a school break, a trade fair, a concert or a bridge day. Compare like with like or exclude those dates from the test. Competitor moves matter too: a hotel closing for renovation, reopening, or running a promotion on the same dates.

Channel effects can mislead in both directions: an agency promotion, a metasearch bid change, or a parity break that sends the same guest to a different door. Small samples are the quietest trap: a few dozen room-nights in a small property is noise. Pool several comparable weeks, or wait until the date band has enough room-nights to carry a signal.

Late-arriving data completes the picture. Cancellations and no-shows land after the booking, so an early win can turn into a loss by the final lead time. And mix shift can hide inside a flat total: room-nights can hold steady while the mix moves from a high-rate segment to a low-rate one, which conceals a revenue loss inside a flat volume number.

What should a rate decision note contain?#

One page: what changed, on which dates and segment, the break-even written in advance, the observed result and the date you will review it.

Write down the change and its scope: rate, dates, segment, channel and the window in which the change applies. Write the break-even in room-nights as a number before the change, so the test has a pass mark rather than an opinion. Record the baseline and the reading, both taken at the same lead time, with the average rate paid alongside the volume.

Note what else moved in the window, including events, competitor actions and channel promotions. Then write the decision: keep, reverse, or extend to another date band, plus the date the next reading is due.

The note matters because it turns a rate move into a repeatable test. Six months of notes is a record of how your own market responds, which no external benchmark can give you. If you work with software that suggests rates, the same discipline applies; see what to check in AI rate recommendations.

What to do next#

Pick one date band and one segment, write the break-even number before you touch the rate, and run one clean test with a review date. The break-even tells you what you can afford to lose, the pace reading at a fixed lead time tells you what you actually lost, and the decision note is what turns a single move into knowledge about your own market.