Evidence updated August 26, 2026
The FOMO Arc: How Hotels Manufacture Last-Minute Discounts
Some hotels run their booking window like a staged sale. The room starts at an ordinary price three months out. Somewhere between three months and two weeks before check-in the rate is raised — hold out and you watch it sit there, a standing warning that waiting costs money. Then, close to the date, the price is cut, and the cut reads as a bargain. It usually is not: it is mostly the detour back. We call the full shape the FOMO arc, and the version that lands back where it started the boomerang rate.
We measured it on 5,049 fully tracked stays at 1,320 hotels: 17% of stays completed the whole arc, and 68.8% of those arcs ended back at (or below) the early price.
A stay counted only if we observed its price in all three phases — 90+ days out, the middle window, and the last two weeks — for check-ins between April 25, 2026 and August 26, 2026.
Key findings
- The mid-window ramp is common. On 43.3% of tracked stays, the price 15–89 days before check-in was raised at least 15%above the same room’s 90+ day price — and held there, not spiked for a day.
- 17% of stays completed the full arc. 859 stays at 385 hotels raised the price, held it, then cut it by at least 15% inside the last two weeks. The median arc: up 31.7%, then down 25.1%.
- Most “discounts” just undo the markup. 68.8% of completed arcs ended within 5% of — or below — the price the room had at 90+ days. If the late drop were genuine distressed-inventory clearing, prices would land anywhere. They overwhelmingly land back at the start.
- The choreography has a schedule. Where the arc ran, the raised level was typically already in place by about 71 days out, and the markdown first appeared a median of 19 days before check-in.
- For some hotels it is a habit. Of 1,270 hotels with at least two fully tracked stays, 378 ran the arc at least once, 246 ran it on at least half their tracked dates, and 93 ran it on every date we could fully track.
The pattern
The four stages of the FOMO arc
Median magnitudes from the measured arcs
Using the study’s median magnitudes on a round $200 room:
- The quiet start. At 90+ days out the room sells for an ordinary price — $200. Nothing about it suggests scarcity, because there is none.
- The ramp. Between three months and about ten weeks out, the rate is raised — the median completed arc lifted it 31.7%, to $263 — while plenty of rooms remain. To anyone watching the price, the message is: book now or pay more later.
- The plateau. The raised rate holds for weeks. This is what separates a genuine demand spike from choreography: the price is not responding to bookings, it is waiting for them. We only count a ramp when the raised level held for a sustained stretch of the middle window, not a one-day spike.
- The “discount.” Inside the last two weeks — median 19 days out — the rate is cut. The median markdown of 25.1% takes our example to $197: right back where it started. A price-watcher sees a falling price and a deal. An early booker saw the same number 90+ days ago without the theater.
The arc preys specifically on people who track prices. Whoever never checks sees one number; whoever watches sees a rise — urgency — and then a fall — reward. Both signals are manufactured by the same rate desk.
Measured examples
8 hotels that ran the arc, charted
One tracked stay per hotel
These are the clearest completed arcs in the study — one stay per hotel, at most one hotel per city, each with a rich observation history and a ramp small enough that it cannot be explained by the cheapest room type simply selling out. They are chosen to show the shape at its most legible, not to represent the median arc.
Repeat behavior
Hotels that ran the arc on every tracked date
At least three fully tracked stays each
93 hotels ran the complete arc on every stay we could fully track. The table lists ten of them — hotels with at least three fully tracked check-in dates, an arc on all of them, and ramps modest enough to rule out room-mix artifacts. One date could be a demand misjudgment; every date is a pricing strategy.
Naming the thing
Does the industry have a name for this?
The pieces do; the whole did not
Revenue managers name the ingredients separately: price anchoring, high-low pricing, last-minute rate drops. Retail regulation has the closest full match — the FTC’s Guides Against Deceptive Pricing describe “marking up to mark down”: inflating a price, holding it briefly, then advertising the cut back to the real price as a bargain. The hotel booking calendar makes the same move without ever printing a strikethrough price: the calendar itself is the “was.”
Academic work on hotel rates has documented non-monotone, hump-shaped price paths — prices that rise into the middle of the booking window and fall near arrival (see Falk & Vieru’s Myth of early booking gains) — but describes the shape, not the signaling. Since the full four-stage choreography had no name, we gave it one: the FOMO arc, because the ramp exists to be seen by people watching the price, and the boomerang rate for the 68.8%of arcs whose “discount” lands back at the early price.
Practical strategy, not a measured outcome
How not to get played
The arc only works on you if the mid-window price is the first price you see, or the only one you remember. Three habits defuse it:
- Anchor early, even if you do not book early. Note the rate once, around three months out. Every later “deal” gets compared against that number, not against the plateau the hotel spent six weeks showing you.
- Treat a mid-window rise at a still-empty hotel as noise. A raise with plenty of availability behind it is a signal aimed at you, not a scarcity fact. Our lead-time study found the 15–30 day window is where hotel prices most often peak — the arc is a large part of why.
- If you want the room, book a refundable rate early. That removes the sell-out risk the ramp is designed to invoke, and leaves you free to rebook if the boomerang comes back down. A last-two-weeks drop below your booked rate is exactly what our last-minute tracker watches for.
Common questions
Is this illegal?
Printing a fake “was $250” next to a $200 price can be deceptive pricing under FTC guidance. Simply charging $250 for six weeks and then charging $200 is ordinary dynamic pricing — no reference price is ever claimed, so no rule is broken. That is what makes the calendar version of the trick so clean: the booking window does the advertising.
Do most hotels do this?
No. The complete arc showed up on 17% of fully tracked stays, and 378 of 1,270 hotels with two or more tracked stays ran it at least once. It is a minority pattern — but a minority you will meet, and the serial group runs it on date after date.
Couldn't the ramp just be the cheap rooms selling out?
Sometimes, and our thresholds are built to limit that. We track the cheapest available rate, so if the lowest room category sells out, the visible price jumps without any repricing. But 68.8% of completed arcs ended back at roughly the early price — if the cheap inventory were really gone, the old price could not reappear days before check-in. The example charts also exclude ramps too large to be a same-room repricing.
Should I just always wait for the drop, then?
No — the markdown is not guaranteed, and on most stays prices near check-in are the most volatile we track. The arc is an argument for anchoring early and refusing to let the mid-window plateau redefine what the room costs, not for gambling on the last week. Our lead-time study covers those odds in detail.
How the study works
For every published hotel we track, we record daily median prices per stay (a hotel, check-in, and check-out combination) across international booking sites, in US dollars. For check-ins between April 25, 2026 and August 26, 2026, a stay qualified as tracked only with observations in all three phases: at least 3 days of prices at 90+ days before check-in, 5 in the 15–89 day window, and 3 in the last 15 days.
- The early price is the median of daily medians at 90+ days out.
- The held-high level is the 80th percentile of daily medians in the 15–89 day window — a level the price actually sat at, so a one-day spike cannot qualify as a ramp.
- A ramp requires the held-high level to exceed the early price by at least 15%; a markdown requires some day in the last 15 days at least 15% below the held-high level; a round trip requires that markdown price to land within 5% of the early price, or below it.
Known limits. Because we track the cheapest available rate, part of a ramp can be a cheap room category selling out rather than a repricing; the round-trip share and the exemplar filters both bound how much of the result that can explain. Markdown timing is only measured from 21 days out, so markdowns that began earlier all report as 21days. And a stay is three phases of one calendar, not a randomized experiment: we can show the choreography, its schedule, and its repetition per hotel — we cannot see the rate desk’s intent.
Cite or share this study
Journalists, researchers, and travel writers are welcome to reference these findings with attribution and a link. Suggested citation:
Best Time To Book research team (2026). The FOMO Arc: How Hotels Manufacture Last-Minute Discounts. A study of 5,049 tracked stays at 1,320 hotels. https://www.besttimetobookhotels.com/guides/the-fomo-arc-boomerang-rate