DwellSignal
Operating the asset

Minimum-night rules and the next 30 days

On 29,000 entire homes, a one-night minimum and a three-night minimum show the same share of the coming month already taken. Longer rules are too thin in the data to call. A null result, with the sample.

27%of the next 30 nights already taken, at a one-night and a three-night minimum alike

Hosts argue about minimum stays the way landlords argue about pet policies: a longer minimum means fewer turnovers and fewer bad guests, a shorter one means more nights sold. We can test one half of that from the calendar. This page reports what we found, which is not much, and says how thin the evidence gets past three nights.

The measure

For each entire-home listing we hold a forward calendar reading. The figure below is the share of the next 30 nights already unavailable at our latest reading, bookings and host blocks alike. It is a measured figure, not a model, and it is a snapshot of demand pressure rather than realized occupancy. The listing's minimum-night rule is read from the same dataset. Only listings carrying both values are counted, which is 29,751 entire homes.

The result

012,4191 night: 12,4191 night12,4192 nights: 12,3322 nights12,3323 to 4 nights: 4,4033 to 4 nights4,4035 to 6 nights: 1665 to 6 nights1667 to 13 nights: 437 to 13 nights4314 to 29 nights: 5814 to 29 nights5830 nights or more: 33030 nights or more330
Listings by minimum stay. The table below carries every column.
Minimum stay Listings Middle listing Lower quartile Upper quartile
1 night 12,419 27% 10% 50%
2 nights 12,332 27% 10% 50%
3 to 4 nights 4,403 23% 3% 47%
5 to 6 nights 166 30% 7% 53%
7 to 13 nights 43 23% 7% 47%
14 to 29 nights 58 50% 10% 76%
30 nights or more 330 27% 7% 50%

The first three rows carry almost the whole sample, and they say the same thing: a one-night minimum, a two-night minimum and a three-to-four-night minimum have the coming month filled to the same degree, with the longer rule a few points behind at the middle and noticeably behind at the lower quartile. The listings with a three-night rule that are doing badly are doing worse than the one-night listings that are doing badly.

The rows below that are too small to read. Fifty-eight listings with a two-week minimum showing 50% is a curiosity, not a finding, and we would not print it on a place page.

What this does and does not say

It does not say the minimum-night rule has no effect on revenue. A three-night rule that fills the same share of nights does so with fewer bookings, fewer cleanings and fewer guest interactions, which is the argument for it. The calendar cannot see cleaning costs.

It does say that, in this dataset, a short minimum does not buy you visibly more forward demand at the median. If you are choosing a rule for a listing in a market where the typical entire home shows a quarter of the coming month taken, the rule is not the lever. Rate and the market are.

Why we published a null result

Because the alternative is a page that tells you the rule matters, backed by nothing. Every number above has its sample beside it, and the rows we do not trust say so. When the calendar readings for the two-week and monthly rules reach a few thousand listings, we will re-run this and date the update.

Every number here is produced by the same rules the product uses and is withheld where the sample cannot carry it. The definitions live in the methodology.
Newsletter

Movers, rule changes, and the numbers behind them

One email a week. Withheld figures included.