How to Tell If Last Weekend Was Priced Right
A five-check post-mortem for a past date: when you sold out, what the last rooms went for, the shape of your pickup curve, and what your comp set did.
You can tell whether a past date was priced right from five things you already have. How early you sold out, what the final rooms sold for, the shape of your pickup curve, what your comp set charged, and how many bookings you turned away. Sell out days early and you were underpriced. Finish with unsold rooms while sitting above the market and you were overpriced. Everything else is reading the middle ground, and there is a scorecard for that below.
Why this question never gets answered
Operators ask it constantly and get "it depends" in return. The honest answer needs data most small hotels never wrote down — what the rooms sold for in sequence, and what the market was doing at the same time.
The good news is that the data exists in your property management system and in a booking site. You just have to look at it in a specific order, on a schedule.
Do this on Monday for the weekend that just ended. It takes twenty minutes.
Check 1: when did you sell out?
The single most useful tell. If you sold out days ahead of arrival, you were underpriced — full stop.
A sold-out Saturday that filled on Tuesday means you had four more days of demand arriving with nowhere to go. Those guests booked your competitor, at whatever they were charging.
Selling out on the day, or finishing at 95–98%, is what a correctly priced peak night looks like. Selling out early is the market telling you your last rooms were cheap.
Check 2: what did the last rooms go for?
Pull the booking list for the date in the order the reservations were made. Look at the final five or ten.
On a night that fills, those rooms should be your highest-priced of the date — they are the scarce ones. If they went out at the same rate as the rooms you sold three weeks earlier, your pricing never responded to the date filling up.
That is not a demand problem. That is a rate ladder that stayed flat.
Check 3: read the shape of the pickup curve
Pickup is the net change in rooms on the books for a date between two snapshots. Pace is how fast that build compares to a benchmark — normally the same date last year. You need both: pickup gives you the movement, pace gives it meaning.
Here is a worked example. A 40-room hotel, one Saturday:
| Days out | Rooms on books | Pickup since last check | Same Saturday last year |
|---|---|---|---|
| 28 | 12 | — | 9 |
| 21 | 19 | +7 | 14 |
| 14 | 28 | +9 | 20 |
| 7 | 36 | +8 | 29 |
| 4 (Tuesday) | 40 — sold out | +4 | 33 |
| 0 | 40 | 0 | 38 |
This date ran ahead of last year at every single checkpoint and then hit the ceiling on the Tuesday. Ahead at 28 days out is your earliest warning; ahead at 14 days with a nine-room week is your instruction to move the rate.
The curve was shouting from three weeks out. Nobody was reading it.
Check 4: what did your comp set do that night?
Your comp set is the six to ten nearby hotels you genuinely lose bookings to. For the date in question you want two facts: their rates, and whether any of them sold out.
If the market sold out and you sold out first at a lower rate, you were the cheapest way into a full town. If the market had availability all weekend and you did not, the same conclusion applies more strongly.
If you were the only one with rooms left and you sat above everyone, that is the overpriced case — and it is much rarer than operators fear.
Check 5: the turn-aways nobody records
A denial is a booking request you could not accept because you were sold out or restricted. A regret is an inquiry that walked away over price.
These are the only direct evidence of demand sitting above your rate, and almost no independent hotel captures them. Front desk logs a tally on a sheet: date requested, rooms, what they were quoted, why they did not book. Two weeks of that changes how you price peak nights.
If you had denials on a night you sold out early, the case is closed.
The scorecard
Copy this table and fill it in Monday morning.
| Check | Where to find it | Underpriced if | Overpriced if |
|---|---|---|---|
| Sell-out timing | PMS booking log | Sold out 2+ days before arrival | Rooms unsold at arrival |
| Last rooms sold | Reservations in booking order | Final rooms at your lowest tier | Final rooms unsold at top tier |
| Pace vs last year | On-the-books snapshots | Ahead at every checkpoint, rate unchanged | Behind at every checkpoint, rate unchanged |
| Comp set rate | Booking site for that date | You were below the market median | You were above the 90th percentile |
| Comp set sell-outs | Same check | Market sold out too | You alone had rooms left |
| Denials / regrets | Front desk tally | Any denials at all | Multiple price regrets |
| Final ADR vs last year | PMS | Occupancy up, ADR flat | ADR up, occupancy sharply down |
Now score the worked example. Sold out four days early. Last rooms at $140 against a final ADR of $148 — the cheapest rooms went last. Ahead of last year at every checkpoint. Comp set median $175 with two of six sold out.
Six checks, six pointing the same way. That Saturday was underpriced, even though it beat last year on both occupancy and rate.
That last part is the trap. Beating last year is not the same as pricing right.
What the evidence says
Two research points frame this review.
On the direction of error: Cornell analyzed 67,008 hotel observations from 2001 to 2007 (Enz, Canina & Lomanno, Cornell Hospitality Report Vol. 9 No. 10, 2009) and found hotels priced 20–30% below their comp set ran 15.2% higher occupancy and 12.2% lower RevPAR. The same report noted that around 54% of hotels were priced below their comp set in both 2001 and 2004. Underpricing is the common failure, which is why most post-mortems land there. The authors note their data shows correlation, not causation.
On how steeply you should have moved: HVS found that rate response steepens sharply once occupancy pushes past roughly 75–80%. That is compression. It is the reason a flat rate ladder loses the most money on exactly the nights that matter most.
On when the curve should be complete: Cloudbeds' 2026 State of Independent Hotels (90 million bookings, 180 countries, 2025 data) put the independent booking window at 40 days. SiteMinder's Hotel Booking Trends panel for 2025 put the global window at 32.15 days. The two panels disagree, so treat either as a rough shape and build the real curve from your own history.
The mistakes that ruin the review
Reviewing only the nights that went badly. Sold-out nights are where the largest errors hide.
Comparing to a flat target instead of a comparable date. Sixty percent at 21 days out means nothing until you know what that date looked like last year.
Reading pickup without pace. Pickup is a net number. Cloudbeds' panel put OTA cancellation at 21.8% against 10.6% direct — a quiet week of pickup can be a busy week of bookings and cancellations cancelling out.
Not writing down the rate you charged. Your PMS stores what sold. Store what you were asking on each snapshot date too, or the post-mortem is guesswork.
Doing it once. One weekend is an anecdote. Eight weekends is a pattern you can price against.
The bottom line
Sell-out timing, the price of your last rooms, your pace against last year, your comp set's rate and sell-outs, and your turn-aways. Five checks, twenty minutes, one recurring Monday slot. Score them the same way each week and you will stop guessing whether the weekend went well — and start knowing which lever was wrong.
The check that eats the most time is reconstructing what your comp set charged on a date that has already passed. Our competitor analysis module keeps a day-by-day record of your rate against each competitor's best rate, with sold-out flags and the age of every scrape stamped on the table. The Monday review then starts with the market history already there.