Booking Window

Defining the sample

For each reservation, subtract its creation date from its arrival date using a consistent calendar basis. Booking and arrival on the same date produce zero days. To find the mean, add these intervals and divide by the number of reservations with valid dates.

Define the cohort first: for example, reservations arriving in June. That is different from reservations created in June. State whether cancellations are included and how date changes are handled. With no valid reservations, the mean is undefined; a zero should not be interpreted as evidence that everyone booked on arrival day.

A hypothetical example

Three reservations arriving in the same period were made 7, 21 and 62 days ahead. Their average booking window is 30 days, while their median is 21. The difference illustrates why the distribution matters: one early reservation pulls the average above the middle observation.

What the result tells you

Lead time gives context to future bookings. An empty month may deserve a different interpretation when typical guests book two weeks ahead rather than six months ahead. The measure helps you examine when demand usually arrives, but it cannot predict future sales on its own.

Look at comparable seasons, unit types or booking sources. One overall average can hide very different patterns between short breaks and holidays planned further in advance. A small sample also deserves more caution than a consistently observed pattern.

When it helps planning

Booking window helps you decide when to review demand for an upcoming period. Compare the same distance from arrival and similar booking sources. Read it with pace and pickup to distinguish booking habits from recent changes in demand.

It is not length of stay

Booking window does not measure stay length or how far into the future the property has opened its calendar for sale. An upcoming period’s average can also change as later bookings arrive. When comparing with a previous year, consider how far before arrival each snapshot was taken.