Measurement
How to evaluate ancillary revenue per occupied stay
A measurement framework for guest services: which metric to use, how to define the denominator, what to record from day one, and how to avoid flattering yourself with the numbers.
- Written by
- Florencia Mazzoni
- Published
- min read
- 3 min read
Most properties measure ancillary revenue, if at all, as a total for the season. That number is almost useless: it moves with occupancy, with season length and with a single large booking, and it cannot be compared to anything — including itself last year.
The metric worth using
Ancillary revenue per occupied stay: total margin from guest services divided by the number of stays that actually took place in the period. Not per night, and not per guest.
Per stay is the right denominator because services are bought by a booking party, once, around a set of shared decisions. A group of six booking one boat day is one decision, not six. Measuring per night makes a two-week stay look weaker than a long weekend for no meaningful reason.
Use margin, not gross booking value. A property that celebrates gross ancillary volume is celebrating money that mostly belongs to suppliers.
The four numbers to record from day one
- 01Occupied stays in the period — the denominator, defined once and never changed mid-year
- 02Number of stays where at least one service was sold — the attach rate
- 03Total margin from services, separated by category
- 04Requests received that could not be fulfilled, and why
The fourth is the one nobody keeps and the one that changes decisions. It is the only number that tells you what to build next, and it is invisible in any financial report.
Reading attach rate correctly
Attach rate — the share of stays that bought anything at all — is more diagnostic than revenue, because it separates two very different problems.
| Pattern | What it indicates | Where to look |
|---|---|---|
| Low attach, high value per sale | Presentation is reaching too few guests | The contact process: timing, channel, whether it happens at all |
| High attach, low value per sale | Only the easy categories are being sold | Missing suppliers in higher-value categories |
| Low attach, low value | There is no commercial process | Everything upstream of the sale |
| High attach, high value | The model is working | Consistency across properties and shoulder season |
Segment before you conclude
A blended average across an entire portfolio conceals more than it reveals. At minimum, separate by:
- Guest type — families, couples and groups buy entirely different things
- Length of stay — a four-night stay has fewer decisions than a twelve-night one
- Season — shoulder season behaves differently, and averaging it in hides both
- Property, where a portfolio is involved — one villa can carry the whole average
Segmentation frequently reverses the conclusion. A portfolio average that looks acceptable often turns out to be two properties performing and six doing nothing.
Three ways to mislead yourself
- 01Counting gross value instead of margin, which inflates everything by the supplier’s share
- 02Including a single exceptional booking in an average without saying so
- 03Changing the denominator between periods — from stays to nights, or from occupied to available — which makes every comparison meaningless
What is a good number?
There is no credible industry benchmark, and anyone offering one should be asked for its source. Ancillary performance varies enormously by destination, guest profile, property type and what is even legally available locally. A figure quoted without those variables is decoration.
The only benchmark that means anything is your own operation last season, measured the same way. Which is why the recording matters more than the target: you cannot improve a number you have never calculated, and you cannot calculate it retroactively.
Start with the four numbers. One season of honest measurement is worth more than any external comparison.
Every sale we make is recorded by category.
Which means the measurement framework above is a by-product of the work, not an extra project.
Request an assessmentRelated reading
- How villas can generate additional revenue after a reservation is confirmedThe window between confirmation and arrival is where most villa ancillary revenue is won or lost. What that window contains, why it closes early, and how to work it without hiring.
- How a performance-based guest revenue model worksWhat "performance-based" means mechanically: where the margin comes from, how attribution is defined, what the property must still verify, and where the model creates friction.
- The hidden operational cost of unmanaged guest requestsAd hoc guest requests look free because nobody bills for them. The cost shows up as fragmented attention, uncontrolled liability and decisions made without evidence.