Usually not. Across 378,364 luxury room-stays we followed from first sighting to check-in, the last rate before the stay was higher than the first about half the time and lower about a third of the time. Nearly two thirds did dip at some point along the way, which is where the folklore comes from.
It is one of the most asked questions in travel, and the answers on offer are mostly vibes. One camp swears hotels panic-discount empty rooms in the final days. The other camp warns that waiting is how you end up paying rack rate next to the conference crowd. Both camps argue from anecdotes because neither has the record.
We have the record. Fermata records the rate for specific room categories at luxury hotels, day after day. For 378,364 completed room-stays at 1,505 hotels between June and September 2026, we can simply compare the last rate we observed before the stay with the first one we sighted, a median of 60 days out.
The honest answer: rates rise more often than they fall
Closer to the date, rates more often rise
Last observed rate versus first sighting per room-stay, by room tier; moves beyond 1 percent either way
fermata.travel · 378,364 completed room-stays at 1,505 luxury hotels, June to September 2026 · two adults, USD
In every tier of room, from the entry category to the suites, more rooms ended pricier than ended cheaper. For entry rooms it is 52.7 percent pricier against 34.3 percent cheaper. Mid categories: 50.0 against 37.8. Top categories: 46.8 against 35.8. The median room ended 0.7 percent above where we first saw it.
So as a default strategy, waiting loses more often than it wins. If the room you want is at a price the record says is good, the odds favour booking it.
But the folklore is not entirely wrong
Here is the twist that makes the simple answer misleading: while only about a third of rooms ended cheaper, more than half saw a cheaper rate at some point during the watch window. On 63.5 percent of these room-stays, a lower rate appeared after first sighting; 28.5 percent dipped 10 percent or more, and the median dip was 8.8 percent.
Read those two findings together and the shape of the market becomes clear. Drops are common, but they are windows, not trends. A rate sags for a few days in the middle of the window, someone books the room or the revenue system reconsiders, and the price climbs back. By the final approach to the stay date, the discount is usually gone, and on the most desirable categories the room itself is often gone too.
That shape is exactly why checking a booking site twice, once when you start dreaming and once when you are ready to pay, tells you almost nothing. The interesting prices live in between, on days you were not looking.
Suites play by slightly different rules
The tier split is worth a second look. Entry rooms post the highest share of increases near the stay, 52.7 percent, because that is where demand pricing bites first. The top categories are the calmest: fewer of them rose, and 17.4 percent simply held within a percent of where they started, against 13.0 percent of entry rooms. A suite's rate is a statement of position; hotels would often rather let it sit than mark it down on a calendar schedule.
When a top category does reprice, though, the move is large: among suites that dropped at all, the median cut was $68 a night, more than double the entry room's $31. Patience on a suite is a low-probability, high-payoff bet, which is precisely the kind of bet you want a machine watching for you.
The same question, asked other ways
Do hotel prices increase closer to the date?
More often than not. About half of the room-stays we measured ended above their first sighted rate, and entry rooms rose most often, 52.7 percent of the time.
Do hotel prices go down last minute?
Not reliably at luxury hotels. The last rate we observed before the stay was lower than the first sighting about a third of the time and higher about half the time. The cheaper windows tend to open earlier and close again before the stay.
Is it cheaper to book a hotel in advance?
Usually. Booking early does not guarantee the lowest rate you will ever see, but it beats waiting more often than it loses. A better rule than any date is to book when today's rate sits low against the room's own record.
What to do instead of guessing
The practical playbook that falls out of the data has three steps.
First, anchor on the room's own record, not a rule of thumb. The forward curve shows what your dates cost against the room's typical level, so you know whether today's quote is already good. Averages across a whole city, or a whole hotel, hide the movement that matters.
Second, if the price is at or below the level the record supports, book it, especially on a room category with thin inventory. The data says the final weeks are more likely to punish waiting than reward it.
Third, if today's price is above what the record supports, do not refresh the page for a month. Set a target for the exact room and let the check catch the window if it opens. If it never opens, you have lost nothing; if it does, you get the email while the rate is still bookable. That is the entire product, and this dataset is why it exists.
Methodology notes
Completed stays only, June 17 to September 23, 2026: 378,364 room-stays, 1,505 luxury hotels, 27,589 room categories, two adults, nightly rates in USD. Each room-stay was first sighted at least 21 days before the stay (a median of 60 days) and observed on at least 4 separate days. "Pricier" and "cheaper" compare the last observed rate with the first, counting moves beyond 1 percent either way; tiers rank each hotel's rooms by price with the cheapest as entry and the top third as top.
