Your listed nights
Your costs
Booked can be 1/0, booked/empty, reserved/available, yes/no. Leave out nights you blocked. Only include nights that were comparable (same listing, similar season and day type) when you can: prices raised on weekends and holidays make those nights look price-proof. Rows are read in the order given (oldest first) so the back-test can score your most recent nights. At least 40 nights with 8 booked and 8 empty, at four or more different prices.
Upload or paste your nights, then click Analyse.
Result
Booking chance by price
What a different price would do
Around your usual price, from the fitted curve. Expected profit is per listed night, after the platform fee and turnover cost; an empty night earns nothing. Rows marked "outside" are beyond the prices you have listed: read them as a guess, not a result.
| Change | Nightly price | Booking chance | Profit per booked night | Expected profit per listed night |
|---|
Back-test
How it is calculated
Model: P(booked) = 1 ÷ (1 + e^−(a + b × x)), where x is ln(listed nightly price). Fitted by Newton-Raphson on the log-likelihood with a tiny ridge so perfectly separated data cannot push the slope to infinity; the 95% interval uses the inverse Hessian. Expected profit per listed night: P(booked) × (price × (1 − platform fee) − turnover cost ÷ average stay − cost per guest night); an empty night earns nothing. The best price is searched on a fine grid only between the lowest and highest price in your data. The 80% ranges and the chance the best price beats your usual one come from 300 seeded bootstrap resamples (booked and empty nights resampled separately), so a result is reproducible. Back-test: the first 75% of rows fit the model, the last 25% are scored by log-loss against simply assuming your overall occupancy.
Not modelled: seasonality, day of week, events, lead time, minimum stays, reviews, competitors, cleaning fees paid by the guest, taxes, and the value of a longer stay. Prices are not random, so this is correlation, not an experiment. Unverified: checked against simulated nights with known curves in unit tests, not against a published dataset or a commercial pricing tool.
Common questions
What is a good nightly rate for my short-term rental?
The rate that earns the most over all your listed nights, not the one that books fastest or the highest one that still books. A higher price earns more per booked night but books less often; this tool fits how your own booking chance falls as your listed price rises, subtracts the platform fee and turnover cost you enter, and finds the price with the highest expected profit per listed night, searching only inside the prices you have actually listed.
How is this different from dynamic pricing tools like PriceLabs, Wheelhouse or Beyond?
Those services price from market data across many listings and cost a monthly fee or a share of revenue. This tool does one narrower thing for free: it reads only your own history and tells you how price-sensitive your own bookings have been and what that implies for one price. It does not know your market, events or competitors, so use it as a check on a pricing service or a spreadsheet, not as a replacement for market data.
How much booking history do I need?
At least 40 listed nights with a result, at least 8 booked and 8 empty, and prices that vary: four or more distinct price levels with a typical spread of about two points or more. If you always list the same price, your history cannot say what would happen at another, so the tool refuses instead of guessing. A full season or more, pooled across listings you price the same way, works best.
What goes in the cost of turnover and the nightly cost?
Turnover cost is what each stay costs you to reset the unit that the guest does not pay for: cleaning not covered by a cleaning fee, laundry, restocking. It is spread across the average stay length. Nightly cost is what a guest night costs to host, such as utilities and consumables. The platform fee is the percentage the platform takes from the host; enter yours. Change them and watch the best price move.
Why might the result be wrong?
Prices are not set at random. If you raised prices on weekends, holidays and event dates, those nights book more often for reasons that have nothing to do with price, and the curve can slope the wrong way; the tool flags that. It also cannot see seasonality, lead time, minimum stays, reviews, your listing's quality or competitors. Treat the answer as a hypothesis for a small, reversible price change on comparable nights, not a number to bet a season on.
Does it know local rules, taxes or platform pricing policies?
No. Occupancy taxes, licensing limits, minimum-stay rules and platform pricing features are outside what the tool models. It does not recommend anything based on them, and the price it suggests is a statistical result from your data only.
Is my booking history uploaded anywhere?
No. The file is read and analysed in this browser tab. Nothing is sent to a server, there are no accounts or cookies, and the tool never stores your bookings. It works offline once the page has loaded.