No. Across 20 western Michigan vacation rentals we manage, the number of reservations a property took in a year was a weak predictor of what it earned. The number that actually tracked annual revenue was booked nights - and those two things are not the same, because a reservation can be two nights or twelve.
This matters because "get more bookings" is the advice owners hear most, and it is the advice that most reliably adds work without adding income.
The figures below come from our own trailing-twelve-month booking data, pulled from OwnerRez on July 1, 2026, for the twelve months ending June 30, 2026. It is a real operating portfolio, not a market survey.
How many reservations does a Michigan vacation rental take in a year?
In our comparable set, reservation counts ran from 20 to 82 in a single year, with a median of 48. Booked nights over the same twelve months ran from 65 to 254, median 168. Average stay length ran from 2.5 nights to 8.5 nights, median 3.2.
The comparable set is 20 doors. We excluded properties without a full twelve months of history, bundle and linked-calendar listings, the two doors running a monthly or mid-term model, the one door outside Michigan, properties that close for the winter by design, doors under long owner-directed blocks, and one property being wound down. Those exclusions are the difference between a number you can reason about and a number that just averages apples into cabins.
Why doesn't reservation count track revenue?
Because a reservation is a unit of work, not a unit of income. Two properties can take the same number of reservations and end the year in completely different places.
Foreman Lake Getaway in Idlewild - two bedrooms, two baths, sleeps six - took 82 reservations, the highest count in the set, and turned them into 225 booked nights. It finished sixth of 20 by annual revenue. Solid, but not the top.
Another door in the set took 27 reservations and turned them into 230 booked nights - about a third of Foreman's reservation count, and almost exactly the same nights on the calendar. It finished second.
And Dragon House in White Cloud - three bedrooms, two baths, sleeps eight - took 26 reservations, within one of that second-place door, but those turned into 114 nights rather than 230. It finished seventeenth.
Same reservation count, half the nights, fifteen places apart. The reservation count told you nothing. The nights told you everything.
What does the math actually say?
Run the correlation and it is unambiguous. Across the 20 doors, annual gross correlated with reservation count at 0.17 - weak. It correlated with booked nights at 0.55 - moderate, and the strongest simple relationship in the data.
We tested whether either figure depended on one unusual property by recomputing with each door dropped in turn. The reservation-count correlation stayed weak in every version (0.07 to 0.36). The booked-nights correlation stayed moderate in every version (0.48 to 0.65). That stability is the reason we are willing to publish it.
Split the set in half by reservation count and the same story appears in plain arithmetic. The ten busiest doors took 2.07 times the reservations of the ten quietest. For that, they got 1.72 times the booked nights and about 1.20 times the revenue. Doubling your reservation count bought roughly twenty percent more income.
Do longer stays earn more?
We cannot honestly tell you yes. Average stay length correlated with annual gross at 0.47 in the full set - which looks meaningful until you drop one door and watch it fall to 0.08. A relationship that fragile is not a finding, and we would rather say so than dress it up.
What we can say is narrower and more useful: longer stays did not cost these properties booked nights. The doors with the longest average stays did not end the year with an emptier calendar than the doors churning through weekends. If a longer minimum stay were quietly costing you occupancy, this data would show it, and it does not.
What does each extra reservation cost you?
Every reservation is a full turnover - a clean, a linen change, a restock, an inspection, a message thread, and a check-in. On a same-day turn, that is a six-hour window between a 10 AM checkout and a 4 PM arrival, which is what a short-term rental cleaning fee actually pays for. It also sits on top of the 18 to 28 field hours a year an average door consumes before anything breaks.
An owner going from 40 reservations to 80 has doubled the turnovers, doubled the arrival risk, and doubled the number of times something can go wrong at 4 PM on a Friday - for, in our portfolio, about twenty percent more revenue.
What should an owner actually look at?
Stop counting reservations. Count booked nights, and then ask what it cost you to get them.
Two caveats we would want stated if we were reading this. Twenty doors is directional, not proof. And correlation is not cause - the doors with long stays also tend to be larger or more remote, and we have not separated those effects. What survives both caveats is the practical point: reservation count is a workload number wearing a revenue number's clothes.
If you want to know which of those two numbers your property is actually optimizing for, that is the first thing we look at on a property management call. It is also why we published what a western Michigan lakefront cabin earns in nights and bookings rather than in a single headline figure.
This article describes what we observe managing short-term rentals in western Michigan. It is not financial advice, and every property is different.