Online travel agency commissions typically run from 15% to 25% of the room rate, and on a $150 booking that is $22 to $37 that never reaches the rooms revenue line the operator manages. Distribution cost is one of the largest controllable expense items below the gross operating profit line, which is why channel mix deserves the same scrutiny as labor.
This article examines how OTA economics work in practice. It publishes information, not financial or commercial advice; operators should model their own channel costs against their own books.
What does an OTA booking actually cost the hotel?
The commission is the visible part. The less visible parts add up: rate-parity obligations constrain how the hotel prices its own website, content teams spend hours keeping photos, descriptions and room types current on each platform, and cancellations inside free-cancellation windows return zero revenue while the commission structure already shaped rate decisions. Industry analyses, including work published by Cornell's Center for Hospitality Research, have long noted that the full cost of a distribution channel exceeds its stated commission rate once these items are counted.
For a 120-room property filling 40% of room nights through OTAs at a 18% effective commission, the annual channel bill runs well into six figures before any marketing line is touched. That figure scales directly with dependence: the higher the OTA share, the larger the invoice.
Why do operators accept these rates?
Because OTAs deliver reach a single hotel cannot buy: global visibility, review content, and placement in front of travelers who search by destination rather than by brand. For independent properties without a loyalty program of scale, OTAs are often the largest single source of new-to-file guests. The commission, in effect, purchases customer acquisition that would otherwise require a marketing budget many independents do not have.
The trade-off is that the guest relationship belongs to the platform. Contact details, stay history and preferences sit in the OTA's database, not the hotel's CRM, unless the operator deliberately captures them on property.
Related stories: Duty of care in hotels: what operators owe guests, staff and the liability ledger · Personalization vs privacy: what data hotels collect from guests, and where the line sits.
How do operators measure true channel cost?
A workable method is to cost each channel per confirmed room night, not per commission invoice. The per-channel stack includes:
- Commission or net-rate margin paid on the booking.
- Payment processing fees, wherever the transaction settles.
- Staff time on content management, availability loading and rate loading per platform.
- Cancellation and no-show behavior by channel, which differs measurably between book-direct guests and OTA guests.
- The cost of acquiring that guest a second time if the hotel fails to capture them for direct rebooking.
Operators who build this table usually find direct bookings are not free either — website, metasearch bids and marketing carry real cost — but the margin difference per night is typically large enough to fund meaningful loyalty and rebooking programs.
What levers shift the mix?
Operators rarely move the mix with a single tactic. The combinations that show up repeatedly in operator practice are: capturing guest emails at check-in and marketing the next stay directly; offering returning book-direct guests a recognizable benefit that does not violate rate parity, such as flexible cancellation, room preference or F&B credit; using OTAs deliberately for need periods rather than as default fill; and auditing OTA content quarterly so conversion on those channels is not quietly degrading.
The goal is not zero OTAs. It is paying commission on the nights that need it, and not on the nights that would have booked direct anyway.
Channel strategy is a margin decision made one booking at a time. Operators who can state their true cost per channel, per night, are the ones positioned to negotiate — with platforms and with their own marketing budgets.
