Which channels produced booked, billed stays?

A thirty to ninety night stay bills a few thousand to ten thousand dollars. But the person who booked it is a relocation manager, a project coordinator or an insurance adjuster, and if the stay goes well they will book again — often several times a year for years.

Cost per booking values the first transaction and ignores the relationship. Because the match is on the booker rather than the stay, every later booking from the same contact credits the campaign that found them, which is usually the difference between a channel looking marginal and looking like the best one you have.

Corporate relocation, project and crew housing, and insurance or displacement placements all book furnished apartments by the month, and all of them arrive through overlapping search terms. Their length, rate and repeat behavior are not remotely alike.

Mapping a booking-type column as Product / service shows each line of business as its own line, with how much of it traces to a channel, so a long insurance placement no longer hides inside the same figure as short project stays, and each line of business can be funded on its own evidence.

Many operators do not own the apartments. They lease them, furnish them, and carry the rent, the utilities and the furniture rental whether a guest is in residence or not. That changes what a marketing channel is worth in a way that owned-asset businesses do not experience: an empty unit is not merely an opportunity forgone, it is a bill arriving on the first of the month. Occupancy pressure therefore varies by market and by week, and so does the value of a booking.

The consequence is that cost per booking, calculated across a portfolio, is a blunt instrument. A booking into a market where three units are sitting empty is worth considerably more than the same booking into a market that was going to be full anyway, and the media account cannot distinguish them. Operators feel this acutely and have almost no way of acting on it. Operators know precisely which markets are carrying vacancy in any given week, because they are paying for it. What they cannot do is point the advertising at that fact.

A city or market column mapped as Location brings part of the two into line. It shows each market's billed revenue and how much of it traces to a channel, so a market carrying vacancy whose revenue sits mostly in Direct / Unknown is visibly one the advertising is not reaching. It does not rank channels within each market, but it stops the portfolio being averaged into something that looks acceptable everywhere, and it is the level at which the business is actually managed. It also makes the awkward markets visible rather than absorbed, which is often where the conversation about renewing a lease on unsold inventory actually begins.

Bookings arrive through intermediaries as often as directly. A relocation management company, a serviced apartment agent or a broker places the guest, and the operator bills the intermediary at a net rate while the client company pays the gross. The margin is real and it varies considerably between partners, and none of it is visible in the number a booking platform or an advertising account reports. The same guest, in the same apartment, generates two quite different amounts of retained revenue depending on who arranged the stay. Nothing about the booking itself reveals which.

If the export carries gross rates on some rows and net on others, the channel ranking is measuring the commercial arrangement rather than the demand. A source producing agency bookings will appear to generate more revenue per stay than one producing direct corporate bookings, when the operator may well be keeping less from each of them. This is a quiet and quite common way for a media budget to be misdirected for a year. The error survives because both numbers are real and both appear in the accounts; they simply answer different questions, and the marketing report has no way of knowing which one it was handed.

Export the amount actually received, consistently, and map a booking-source column as Product / service to mark the intermediary. Direct and agency bookings then show as separate lines, each with its revenue after commission and how much of it traces to a channel, and the channel ranking itself is computed on what the operator keeps. That comparison is difficult to assemble by hand and tends to change how the direct-booking effort is funded. Direct demand is expensive to build and cheap to serve, and no operator can decide how hard to chase it without knowing what the alternative actually costs them per stay.

A stay quoted at ninety nights ends at sixty-two because the project finished early, or runs to a hundred and forty because the house purchase fell through. Insurance placements extend until the building work is signed off and nobody knows when that is. Early departures carry notice periods, extensions are charged at a different rate, and the final invoice rarely matches the reservation in either direction. The reservation is a forecast made by somebody with imperfect information about a project, a house purchase or a building repair that has not happened yet.

Reporting on booked nights therefore measures an intention that was superseded. It also systematically flatters whichever sources produce the longest quoted stays, regardless of whether those stays happened, and it hides the sources producing short bookings that routinely extend — which in this business is where a good deal of the profitability sits. A source producing short bookings that routinely run on can look like the weakest performer in the account for as long as the reporting stops at the reservation. It is a quiet way to defund the most profitable demand you have.

Because the match runs on the booker and every billed row credits them, the report is built on invoiced nights rather than reserved ones. An extension is additional revenue against the same acquisition, not a new booking. An early departure simply produces less. Nothing is projected forward and no average length of stay is assumed on the operator's behalf, which is what makes the figure reconcile to the ledger. That reconciliation is what lets the report be read next to the operating statement rather than instead of it.