Pricing a shoulder season without gutting your rate

The instinct when a calendar thins out is to cut the nightly rate everywhere and wait. It feels responsive.

It is also the change most likely to cost money, because it discounts the nights that were going to book anyway. Pull ninety days of booking data and split it by day of week before touching anything.

In most markets the shoulder-season gap is not a month-wide problem. It is Monday, Tuesday and Wednesday, and those three nights are being priced as though they compete with Saturday.

Look at two numbers per weekday: the share of nights booked, and the average lead time. A night with low occupancy and a short lead time is price-sensitive and will respond to a discount.

A night with low occupancy and a long lead time has a demand problem instead, and cutting the rate simply lowers what you earn from the guests who were already coming. That distinction matters because the two look identical on a calendar.

Both show gaps. Only one of them is a pricing question.

Hosts who skip this step end up discounting demand they already had, which is why a flat cut so often produces more bookings and less revenue at the same time. Before discounting at all, check what the minimum-stay setting is excluding.

A two-night minimum on a Tuesday removes every single-night business traveller from the pool. In many markets that is a larger effect on occupancy than any plausible rate change, and unlike a discount it costs nothing per night.

A reasonable first experiment is to drop to a one-night minimum from Monday to Wednesday, hold the two-night minimum from Thursday to Sunday, and leave the nightly rate alone entirely. Give it three weeks before reading the result.

Booking curves lag, and a shorter window is reading noise rather than signal. If the gap persists after that, discount the specific weeknights rather than the week.

A twelve per cent reduction applied to three nights protects the weekend rate, which is where most of the revenue sits, and concentrates the incentive where demand is actually soft. Measure the outcome as revenue per available night rather than occupancy.

Occupancy is easy to move and easy to be fooled by. A fully booked calendar at the wrong rate is a worse outcome than a seventy per cent calendar at the right one, and only the revenue figure tells you which one you have.

One caveat worth taking seriously: several cities tie short-term rental registration to a minimum-stay threshold, and dropping below it can move a listing into a different permit class. Confirm the local rule before changing that setting, and note the date you checked, because these rules change more often than the advice about them does.

Two failure modes are worth naming because they look like success. The first is a calendar that fills the week after a cut, which feels like proof the price was wrong when it usually means the discount was simply larger than it needed to be.

The second is a run of five-star reviews arriving alongside falling revenue, which happens when a listing is priced below what its guests would happily have paid. Neither is visible from occupancy alone, and neither shows up inside a single month.

The only reliable read is a rolling comparison of revenue per available night against the same window a year earlier, adjusted for any change in the number of nights you actually made available. That last adjustment matters more than it sounds: blocking a week for maintenance and then comparing raw monthly totals will make a good month look like a bad one.

None of this requires software. A spreadsheet with four columns — date, day of week, booked, rate — covers every decision above, and building it by hand once is a better education in your own market than any dashboard.

Buy the pricing tool later, when you know what question you are paying it to answer.