ADR, occupancy, and RevPAR are the three core revenue metrics in short-term rentals, and they are constantly confused for one another. The short version: ADR is your average price per booked night, occupancy is the share of available nights you filled, and RevPAR is the two multiplied together, your revenue per available night. ADR measures price. Occupancy measures fill. RevPAR measures what you actually earned per night you had to sell.
The reason this matters is that ADR and occupancy pull against each other. Raise rate and occupancy tends to fall. Cut rate and occupancy tends to rise. You can move either one in isolation and feel like you are winning while you are actually losing money. RevPAR exists precisely because the first two can each lie, and only their combination tells the truth.
"ADR can rise while you earn less. Occupancy can rise while you earn less. RevPAR is the one that cannot fool you."
How the three relate, in one line
RevPAR equals ADR multiplied by occupancy. A unit with a $250 ADR running 60% occupied has a RevPAR of $150. A unit with a $200 ADR running 80% occupied has a RevPAR of $160. The second property charges less per night and earns more per available night. If you optimized for ADR, you would pick the first and lose. If you optimized for occupancy, you would pick the second for the wrong reason and might push it to 95% by discounting, and lose a different way. RevPAR is the only one of the three that ranks them correctly.
What each metric is good for, and what it hides
None of these is useless. Each answers a real question. The error is asking a metric the question it cannot answer.
ADR is good for pricing diagnosis.
It tells you the average price the nights that sold went for, which is useful for judging your rate against the comp set. It hides the empty nights entirely, so it can rise while revenue falls. Never read it alone.
Occupancy is good for demand and structure diagnosis.
It tells you how much of the calendar is filling, which surfaces structural problems like a bad minimum stay or a distribution gap. It hides price completely, so a full calendar can mean strong demand or chronic underpricing. Never read it alone either.
RevPAR is good for the verdict.
It combines rate and fill into the number that actually pays the owner, and it counts every available night, so it cannot be gamed by sacrificing one input for the other. This is the metric you manage to, and the one you report.
The two ways operators optimize the wrong metric
Almost every revenue mistake we find on a new book is one of two errors, and both come from optimizing one of the first two metrics instead of RevPAR.
The first is chasing ADR. The operator raises rates to protect a high average price, occupancy quietly bleeds, and the calendar empties faster than the rate climbs. ADR on the dashboard looks great. Revenue is down. Because the headline price is healthy, nobody catches it until the owner statement disappoints.
The second, and far more common, is chasing occupancy. A full calendar feels like the scoreboard, so the operator discounts to keep it full, often pushing units past 90% occupied. Every one of those nights earns less than it could have. The book runs busy and underpaid, and because the calendar is full, it reads as success. Both errors are invisible to anyone watching only the metric being optimized. Both are obvious the moment you look at RevPAR.
"Optimizing ADR or occupancy alone is how you lose money while a dashboard tells you that you are winning."
Why RevPAR has to be same-store to be honest
There is one more trap, and it is on the reporting side rather than the pricing side. RevPAR can be inflated without any real improvement by changing the mix of units in the calculation. Add a batch of high-performing units, churn out a few weak ones, and portfolio RevPAR climbs even if no individual property got better. That is a mix-shift mirage, not a result.
The fix is to measure RevPAR same-store, comparing only the units that were active in both periods. Same-store strips out the effect of adding and dropping properties so that a year-over-year RevPAR gain reflects actual revenue work, not a reshuffled portfolio. Every performance number we report is same-store for exactly this reason. When we say Geneva Lakes Vacations lifted Adj. RevPAR from $88 to $128 year over year, a 46% gain on the KeyData same-store methodology, that is the same cohort of units both years. The gain is real because the comparison is honest.
Which one should you actually manage to?
Manage to RevPAR. Use ADR and occupancy as the two diagnostic dials that tell you why RevPAR moved and which lever to pull next, but never let either become the goal. The goal is revenue per available night, measured same-store, benchmarked against a real comp set. A pricing tool will optimize the nightly rate inside whatever structure you give it, and the good ones do that well. Deciding the structure, reading the three metrics together, and keeping the comparison honest is the revenue management layer on top, and it is where the durable gains live.
If you want to see your own ADR, occupancy, and RevPAR read together against a real comp set, same-store, we run a free revenue audit for operators at 20 or more units, with no commitment. We also back the engagement with the Pacer Promise: cancel in the first six months and we return 50% of fees paid.
Adapted from Pacer's editorial archive, May 2026.