Airlines sell each cabin as a stack of fare buckets, and the cheapest ones close on a fixed schedule — not in response to anything you typed. Hardcoded advance-purchase fences strip the lowest fares at 21, 14, and 7 days before departure, and revenue-management systems open and close the rest as seats sell.
That makes the booking window the lever that actually matters. Conventional guidance puts international trips at 3 to 6 months out and domestic at 1 to 3 months, with peak Asia-Pacific dates worth watching earlier.
Two travelers on the same transpacific flight, booked a week apart, can pay wildly different prices for identical seats in the same cabin. One of them will swear the airline tracked her searches. She’ll clear her cookies, try a private window, and see the same number.
The real explanation is duller and more useful: fare-bucket depletion. Airlines allocate seats into price tiers, the cheap tiers run dry, and the next person searching gets quoted the tier above. Nothing about your browser enters into it.
Understanding that mechanism changes what you do next. It tells you why last-minute deals almost never materialize for leisure travelers, why a fare moves between two searches minutes apart, and why the practical defense is timing rather than technique. For Asia-Pacific routes, where demand peaks cluster hard around Lunar New Year and Golden Week, that timing question gets sharper still.
Why the cheap seats vanish
Airlines don’t sell one seat at one price. They split each cabin into a stack of fare classes — the trade calls them fare buckets — and every bucket carries its own rules, its own change and refund conditions, and its own price point.
Each bucket holds only so many seats, and the cheapest ones are finite. Once that low-priced inventory is gone — the point the trade calls fare bucket depletion — the next person searching the same route is simply quoted the class above.
Some buckets close on a clock rather than a counter. Rules called advance-purchase fences retire the cheapest fares at fixed cutoffs — 21, 14, and 7 days before departure — whether or not anyone bought them. Airlines love to bury this in the fare rules. That’s the point of a fence — it’s a condition you’re meant to discover after the fact.
So why doesn’t a half-empty flight go on sale the week it leaves? Because unsold seats close to departure aren’t discounted for leisure travelers. They’re held back for business travelers, who book late, pay whatever the fare is, and carry the yield.
Filling every seat was never the goal. Getting the most revenue per seat was.
The incognito myth, and the number airlines actually watch
Clearing your cookies, browsing in incognito mode, or switching devices changes nothing about the fare you’re shown. Fares are not stored per traveler — the inventory system keeps no record of who searched for what. There is no personalized price waiting to be sprung on a repeat visitor.
What those systems do watch is how fast seats are selling right now — aggregate booking velocity — set against demand forecasts drawn from historical booking patterns. Revenue-management algorithms use that comparison to open and close buckets in real time.
When a route starts moving faster than the model expects, the cheap inventory gets pulled early, and it can happen in the middle of a Tuesday.
This is the part travelers misread. A fare that jumped between your Tuesday and Thursday searches isn’t punishing you for looking twice. Other people were booking that route in the meantime, and the algorithm reacted to them.
Superdeals cover departures from North America, Europe and Australia.
Flight deals most people never see
Air Traveler Club monitors thousands of routes for pricing anomalies that traditional search engines miss. With ATC Superdeals members save up to 80% on each flight vs. regular fares.
Too good to be true? Check it out yourself. It's free.
Delivered weekly. Unsubscribe anytime.
The booking window is a range, not a date
Conventional guidance puts international bookings at 3 to 6 months before departure and domestic at 1 to 3 months. That’s a range, and it moves with route length, season, and how generously an airline is willing to sell its cheap buckets.
For Asia-Pacific routes specifically, the public evidence is thinner than most advice columns let on. One commercial Singapore guide puts intra-Asia travel at roughly 6 to 12 weeks out and medium-haul trips to Japan, Korea, or Australia at 6 to 10 weeks. None of it comes with a disclosed sample, a route list, or an observation period — which is why those narrower numbers work as a starting point for when to begin watching, not as a prediction for a specific date.
| Carrier or route type | Recommended booking window | Documented advance-purchase fence days | Pricing-model note |
|---|---|---|---|
| Singapore short-haul LCC guidance; AirAsia and Scoot examples | 3–6 weeks for short-haul LCC flights; commercial guidance also reports 6–10 weeks for medium-haul routes such as Japan, Korea, and Australia | Not verified for a named carrier | Fare products and ancillary bundling differ between carriers; continuous-pricing use is unverified |
| Singapore-origin LCC comparison; AirAsia, Scoot, and VietJet examples | 6–10 weeks for non-peak travel, according to commercial comparison | Not verified for a named carrier | Comparison describes base fares, bundles, and optional add-ons, not the underlying inventory algorithm |
| Singapore school-holiday and public-holiday travel | 10–14 weeks in one commercial Singapore booking guide | Not verified | Earlier monitoring is advised during peak demand, but the source does not link the recommendation to a published fare fence |
| US agency-distribution market | No route booking window established | Not applicable | NDC represented 21.2% of ARC-settled transactions in December 2025; this is a distribution metric, not a seat-pricing rule |
| International travel to Asia | 2–8 months in commercial guidance, with a reported concentration around 21–25 weeks | Not verified | The retrieved source does not disclose a sufficiently documented route sample or methodology for treating this as a general rule |
| Intra-Asia regional travel | 2–3 months, or approximately 6–12 weeks, in commercial guidance | Not verified | No primary corridor dataset was found; treat as an unverified practical range rather than a universal threshold |
| Source: Navan; ShopBack Blog; Fly Search Engine Pro | |||
Read the table less as a schedule and more as a map of what’s actually known. The one figure in it backed by a clearly identified transaction metric is a distribution number, not a pricing rule. Every booking-window entry traces to commercial guidance without a disclosed route sample.
Where 18 million fare changes a day come from
Behind every fare display sits ATPCO, the clearing house that maintains roughly 447 million fares and processes about 18 million fare changes a day, by the company’s own count, before distributing them to booking systems worldwide.
That figure counts fare-file activity across the entire industry, not price moves on any single itinerary — carriers constantly creating, altering, and retiring fare products across thousands of markets at once. The practical consequence follows from the volume: the price you saw this morning may not exist by the afternoon, even though nothing in your browser has changed.
How to use this when you book an Asia-Pacific trip
Peak dates change the calculation more than the carrier does. Year-end travel, Singapore school holidays, Golden Week, and Lunar New Year all pull demand forward, and commercial guidance is consistent on the practical takeaway: start watching earlier when a holiday sits anywhere near your travel dates.
Compare the whole offer, not the headline. Low-cost carriers in the region build fares differently — AirAsia starts from a base fare and sells a Value Pack on top, while Scoot folds its products into a bundle — so a price that looks cheaper up front can lose once baggage and change conditions are added in.
One tactic worth skipping: chasing a specific day of week. Comparisons do circulate claiming midweek departures beat Friday and Sunday ones on price, but none of the sources behind them shows the dataset. Without that, you’re guessing with money.
Reporting by
Key terms
- Fare buckets
- A fare bucket is one of several price tiers an airline creates inside a single cabin, each carrying its own rules and price point. Airlines split every cabin into a stack of these classes, and only a limited number of seats sit in the cheapest ones. That structure is why the booking window matters more than any browser trick — the cheap tiers are finite, and the calendar, not the search, is what closes them.
- Advance-purchase fences
- An advance-purchase fence is a fare rule that takes the cheapest tickets off sale on a fixed schedule rather than when inventory runs out. Airlines typically set these cutoffs at 21, 14, and 7 days before departure, and the rule applies whether or not anyone bought the seats. Because the cutoff is a condition of the fare rather than a visible countdown, a traveler who starts shopping inside the window never sees the cheaper tier at all.
- ATPCO
- ATPCO is the clearing house that maintains the airline industry’s fare files and distributes them to booking systems worldwide. It maintains roughly 447 million fares and processes about 18 million fare changes a day, by the company’s own count. That figure measures industry-wide fare-file activity rather than movement on any single route, which is why the fare file behaves less like a price list and more like a live market.
- Booking velocity
- Booking velocity is the rate at which seats on a given flight are selling, measured in real time. Revenue-management algorithms weigh that rate against forecasts built from historical demand to decide when to open or close fare buckets. It is the number airlines actually watch, and it moves independently of any individual traveler’s search — which is why the same route can price differently on two consecutive days.
Questions? Answers.
Does clearing cookies or using incognito mode change the airfare I’m shown?
No. Airline pricing has no per-user component, so your browsing history, cookies, and private windows have no effect on the price you’re shown. Prices move because fare buckets deplete and revenue-management systems adjust availability, not because of anything in your browser.
What is fare bucket depletion, and why does the price rise between searches?
Airlines divide each cabin into multiple fare classes — buckets — each with its own rules and price point. Only a limited number of seats sit in the cheapest ones. When those sell out, depletion has occurred, and later searchers are simply shown the next fare class up.
How far in advance should I book a flight to Asia-Pacific?
Most guidance puts international trips three to six months ahead and domestic ones one to three months ahead, and it’s best treated as a range rather than a date. Booking inside that window also keeps you ahead of the advance-purchase fences that pull the cheapest fares off sale at 21, 14, and 7 days before departure.
Do airlines drop prices at the last minute to fill empty seats?
No. Airlines don’t cut prices at the last minute to fill empty seats. Seats that remain close to departure are reserved for business travelers, who book late, care less about price, and produce more revenue per seat.
