When is the best time to book a flight?

At a glance#

Why there is no best weekday#

The familiar rule about a cheap booking day comes from a time when fares were filed manually and loaded once a week. Today prices change several times a day and follow how individual flights are filling. The calendar plays no part; the booking position does.

What survives of the rule is one observation: more leisure travellers search at weekends and more business travellers during the week. That shifts demand, and it does nothing to the price on the day you search. Planning around the booking weekday optimises the wrong variable, since the travel day decides far more than the purchase day.

The other widespread rule fails the same test. A fixed number of weeks before departure holds at best as an average across thousands of routes and says nothing about one connection. Averages combine very different cases: a summer leisure route and a November business route sit in the same statistic while behaving in opposite directions.

How the price is created#

Buckets, not a curve#

Every flight carries several fare buckets side by side, each with its own allocation. When a bucket sells out, the next one applies.

So prices jump. Nothing may change between two searches, and then the price moves because a single seat was sold. Two people searching the same flight a minute apart can see different prices for that reason alone. Once that is clear, waiting looks different: the useful question is how many seats remain in the current bucket, not whether the price will fall.

Almost every price movement comes down to three building blocks: the buckets in the system, the airline’s control over them, and the booking class that ends up on the ticket.

The airline controls the allocation#

Allocations are not a law of nature. Airlines move them when a route sells slower than planned and close buckets when demand picks up.

Those adjustments produce the apparent randomness. A price drops three weeks out because a bucket opened, and rises again the next day. A drop is no signal of further drops; it shows only that somebody opened an allocation, and the same allocation can close by morning.

Only the last figure is visible from outside. Booking sites sometimes show how many seats remain in the current bucket, and that number is more useful than any forecast.

The booking class sits on the ticket#

Every bucket carries a letter, and that letter decides more than the price. It governs changeability, mileage accrual and the chance of an upgrade.

A cheap price in a poor booking class is not always the better purchase. What sits behind it is explained in what fare rules say. For status collectors the letter matters most of all, because accrual follows the class and not the price, and two equally priced tickets can differ sharply.

Disruption shows the effect again. When a flight is cancelled, replacement seats also follow buckets, and a cheap fare then ranks lower, as what to do when a flight is cancelled describes.

The three factors that matter#

Demand on that route#

Busy routes empty their lower buckets early, and booking late there reliably costs more. Weakly loaded routes behave the opposite way. Lower buckets stay open, and last minute offers do occur.

Judging your own route takes minutes. How many airlines fly it, how often per day, and is it a business or a holiday connection? Those three answers say more about the right moment than any rule of thumb.

Points collectors should also check the accrual rule. How that relates to the value of a redemption is in what a point is worth.

Your flexibility#

A day earlier or later often moves the price more than three weeks of waiting. The same applies to the time of day and to nearby airports. Flexibility is the most valuable factor in this calculation, because it works immediately and without any forecast.

A simple test helps: how many alternatives exist on the same day? With one flight a day the price is inelastic; with ten, demand spreads out.

Without flexibility the order reverses. Then booking happens early, because the risk runs one way only.

What waiting would cost#

Before any waiting comes one question: what happens if the price rises? A traveller who has to make the trip carries that risk alone.

With a fixed trip, waiting is a bet against your own necessity. With an optional trip, waiting is cheap, because a high price simply ends the plan. The distinction replaces every rule of thumb and anchors the decision in your own circumstances, where statistics about other people’s bookings cannot reach.

For families the position sharpens. Four seats in one bucket are rarer than one, and the price follows the bucket that holds everybody. Holiday periods are the extreme case: demand is known in advance and airlines plan accordingly, so the lower buckets are often gone months ahead.

For business travel the weighting frequently reverses. Changeability matters more than price, because a moved meeting costs more than any fare premium.

Worked example: fixed trip or flexible trip#

Two travellers look at the same flight. One has to be there on a date, the other will only go if it is cheap.

One question has two answers here. The flight does not decide; the commitment to the trip does.

For the fixed trip the cancellation option is worth most. A booking that can be cancelled free within a window locks the price and keeps the decision open. Watching a fixed trip instead of booking it is the most common mistake of all, because it collects data about a price you will pay anyway.

The flexible traveller works the other way round. Price alerts earn their keep, the date and even the destination can move, and a rise in price is information, not a threat.

Unsure which case applies? Looking at the same flight a week later shows what a late purchase costs on that route.

What does not move the price#

Several assumptions persist and change nothing.

Browser history. Prices come from the reservation system. Differences arise from currency, point of sale and channel, never from repeated searches.

Private browsing. It changes the display, not the allocation. The same flight in the same currency shows the same bucket.

The hour of the search. Prices do move during the day, but not on a schedule anybody can exploit.

The number of searches. Searching ten times shows the same bucket ten times. Search volume is irrelevant to the allocation.

What does change the price is the sales channel. Portals, agencies and the airline’s own site draw on the same buckets but sell different conditions, as booking direct or through a portal shows.

When it is not worth waiting#

Waiting for a better price is not worth it in these situations:

• Fixed dates. A traveller who must go carries the risk alone and therefore buys early.
• Holiday periods. Demand is certain, and the lower buckets empty early.
• Few connections. On routes with one flight a day there is no alternative day.
• Travelling as a group. Several seats in one bucket are rarer than one.
• Onward arrangements already booked. With the hotel and the car in place, the flexibility has been spent and the flight is no longer optional.

With separate tickets another point applies. A cheap price across two bookings carries a risk the price does not show, as what applies with separate tickets describes.

Award bookings follow different rules entirely. There the release of award space decides, and the cash price logic does not apply.

Waiting also has a hidden cost beyond the fare. Every day spent watching is a day the preferred seat, the preferred time and the preferred connection stay unbooked, and those run out before the price moves.

What a search should look like#

Three steps are enough and take about fifteen minutes together.

First set the range that works. Two days either side of the preferred date widen the choice without changing the trip.

Then sort the connections by connection time and fare conditions as well as by price. A cheap price on a tight connection is a different product.

Finally decide and stop. After the second round the comparison adds no further information, and an hour of searching for a small difference is rarely a good trade when that hour recurs with every booking.

A short cross check helps. Searching the same flight in another currency or from another country shows the same seat at a different total, and the difference comes from taxes and distribution, not from demand.

Regular travellers on the same route build a rough expectation of their own. After three or four bookings the usual range is visible, and that replaces every general rule.

The best time to book is a judgement, not a date. Buckets set the price, demand on the route sets the pace, and your own flexibility sets the room to manoeuvre. With a fixed trip, book early and use a cancellation window. With a flexible trip, wait and be ready to drop it. Once it is settled whether the trip is committed, the booking moment is settled too, and the search shrinks to picking the right connection.

The most common mistakes#

• Waiting for the cheap weekday. Moving the travel day changes the price; moving the booking day does not.
• Assuming prices fall steadily. Buckets jump, they do not drift.
• Comparing price alone. The booking class and its conditions are part of what is being bought.
• Waiting with a fixed trip. An early booking with a cancellation window removes the risk without removing the choice.
• Booking late for holidays. In demand peaks the lower buckets are gone long before most people start looking.
• Comparing endlessly. After two rounds the extra searching costs time and finds nothing new.

Related topics#

• What fare rules actually say
• Booking direct or through a portal
• What applies with separate tickets
• How fare classes work
• What to do when a flight is cancelled
• What points are actually worth