A prior piece on this site posed Feynman’s test to the luxury hotel industry: explain OTA dependence in plain language. Not the name. The mechanism. What actually happens, step by step, at the level of data and behavior and compounding disadvantage.
Most could not.
This is the answer.
What the Hotel Receives. What the Platform Keeps.
The transaction looks simple. A traveler searches. A hotel appears. A booking is made.
Feynman would stop you there. Simple is not the same as understood. Describe what actually transfers.
When a traveler searches on Booking.com or Expedia, the platform records the behavioral demand data: what dates were searched, what price triggered hesitation, what price triggered conversion, which competing hotels were considered and rejected, what device was used, what time of day the decision was made, what traveler segment the user belongs to based on accumulated behavioral history. Every click, every scroll, every abandoned search is a data point that updates the platform’s model of that traveler and of the market.
The hotel receives reservation data: a name, stay dates, room type, rate plan, and contact details that vary by channel and policy.
These are not the same thing. Reservation data tells the hotel who booked and when. Behavioral demand data tells the platform what that traveler wanted, what they rejected, what price moved them, and what alternatives they would have accepted. The OTA retains the demand data. The hotel receives the reservation.
The OTA learns from the search behavior that produced the booking. The hotel learns from the reservation that resulted.
That asymmetry is the mechanism. OTA dependence is not just paying for bookings. It is paying for bookings inside a system where the platform keeps the behavioral learning that makes future demand easier to control. Everything else compounds from that asymmetry.
How the Ranking Engine Works
No outside party has access to the proprietary ranking formula. The operating logic, however, is documented in publicly disclosed OTA materials, and it is sufficient to explain how the system shapes hotel visibility.
OTAs create real distribution value: reach, booking volume, and access to travelers a single property cannot efficiently acquire on its own. The issue is not whether OTAs should exist in the channel mix. The issue is whether the hotel has enough owned demand to negotiate from strength rather than dependency. Understanding how the ranking system works is what makes that distinction actionable.
Expedia explicitly identifies compensation as one placement signal, stating that marketplace placement is shaped by search relevancy, offer strength, guest experience, availability, price competitiveness, and what Expedia is paid, including commissions and booking compensation. Booking.com separately describes preferred visibility programs that can increase prominence for eligible partners, with visibility affected by conversion performance, content quality, cancellation behavior, pricing, and participation in those programs. The systems are not identical, but the pattern is consistent: commercial participation can affect visibility inside OTA marketplaces.
The ranking system is not a neutral list. It is a prediction engine optimizing for platform outcomes alongside traveler outcomes.
Here is the operating sequence.
The system evaluates each property against each search query and predicts the probability that showing this hotel to this traveler will result in a completed booking. The inputs include price competitiveness, review score, content quality, historical conversion rate for the property, traveler segment, device type, length of stay, and season.
Because results are personalized, there is no single ranking. A traveler who typically books four-star properties sees a different list than one who searches on price. A hotel cannot optimize for a position that does not exist consistently across users.
Commercial participation layers on top. Hotels in preferred programs, at higher commission tiers, or using sponsored placement tools occupy different visibility surfaces than those that do not participate. The effect is not that commission alone determines ranking. The effect is that paid visibility programs create additional exposure surfaces inside a system where visibility is already shaped by commercial, behavioral, and performance signals.
Cancellation behavior is tracked and can affect visibility. Platforms have an economic interest in stable, completed bookings. Properties whose reservations cancel at elevated rates represent less predictable revenue. The precise weighting of this signal is not publicly disclosed, but both major platforms document cancellation behavior as a factor in property performance assessment.
The system learns continuously. Here is the sequence that follows from that:
Search behavior creates demand data. Demand data improves platform prediction. Better prediction improves traveler routing. Better routing increases platform conversion. Higher conversion strengthens ranking confidence for properties that perform. Hotels whose direct demand is weak pay for visibility inside the same system to remain competitive.
That is the reinforcement loop. It is not a trap designed against hotels. It is the natural operating structure of a marketplace that benefits from supplier participation and improves its own models with every transaction it processes. The structural consequence for hotels that rely on it heavily is that the path back to visibility, when it declines, runs through increased platform participation.
The Visible Cost and the Structural Cost
The industry tracks commission rates. Commission rates are not the cost of OTA dependence. They are the entry fee into a system with structural costs that compound invisibly.
Start with what is visible. A room at $400 per night, a three-night stay, a 20 percent commission: the hotel pays $240 on a $1,200 transaction. Most operators know this number. It is real and significant. It is also only the part the industry already knows how to count.
The structural costs are three.
The first is data loss. The hotel received no behavioral demand data from that booking. It does not know what alternatives the guest considered, what price sensitivity governed the decision, or what the platform learned about that traveler’s preferences. That intelligence is what the platform uses to serve that traveler more precisely on the next search. The platform’s model improves from the full search path. The hotel’s model improves only from the narrower reservation and stay data it receives. The asymmetry widens with every transaction routed through the intermediary.
The second is lifetime value transfer. If the guest’s next search also begins inside the OTA environment, the economics repeat. Four future stays at the same terms would generate $240 in commission on each $1,200 transaction, totaling $960 in repeat commissions from a single guest relationship, before any paid placement or preferred participation costs are included. The hotel did not lose one booking. It lost the relationship that governed five, and paid to introduce the guest to the intermediary that now holds it.
The third is the visibility tax. The base commission is not the effective rate. Hotels participating in preferred visibility programs pay 3 to 5 percent above the base. Sponsored placement adds more. Loyalty program participation adds more. In some participation structures the effective rate rises materially above the quoted commission once the full program mix is calculated. The industry reports commission rates. It rarely reports effective rates.
Three costs. One is visible. Two are structural. The industry accounts for one.
Once visibility is shaped by a platform that retains the behavioral learning from every transaction, the economics extend well beyond the commission. The commission is the charge attached to a larger transfer: search intelligence, future demand positioning, and the guest relationship itself.
Why Luxury Hotels Are More Exposed
The mechanism applies across all hotels using OTA distribution. The exposure is greater at the luxury end, and the reason begins with what the luxury commercial model actually requires.
Luxury hotels are not merely selling rooms. They are monetizing a relationship built on knowing the guest, anticipating preferences, and delivering recognition across visits. That relationship depends on owning the demand intelligence that precedes the stay. OTA intermediation captures precisely that layer. The behavioral data that would allow a luxury property to understand who this guest is, what they value, and how to earn the next stay without an intermediary. That data stays with the platform.
The financial exposure follows from that structural problem. A 20 percent commission on a $150 room night costs $30. On a $500 room night it costs $100. The data loss, the lifetime value transfer, and the visibility tax all scale with transaction value. The compounding is proportionally larger at luxury ADR levels.
The personalization disadvantage is also harder for an independent property to close. A platform that has processed millions of luxury traveler searches, modeling preferences, price sensitivity, and behavioral patterns across properties and geographies, can serve a more precisely tailored result than any single property can construct from its own guest records. Every booking routed through the OTA improves the platform’s model of that traveler. Every booking widens the gap between what the platform knows and what the hotel knows.
The mechanism is the same across hotel categories. The cost of operating inside it, financially and relationally, is higher at the luxury end.
Now You Know the Mechanism
Feynman’s test was simple: if you cannot explain what happens step by step, you do not understand the thing. You only know its name.
The name is OTA dependence.
The mechanism is this: platforms capture behavioral demand data at the point of every intermediated transaction. They use that data to build and refine predictive models that determine visibility, personalization, and traveler routing. Hotels receive reservation data. The asymmetry between what the platform retains and what the hotel receives compounds with every booking. The reinforcement loop that follows from declining visibility runs through increased participation in the platform’s own programs. The lifetime value of guest relationships transfers progressively to the intermediary. The commission is the visible fee. The structural costs are harder to measure, easier to ignore, and often more strategically damaging.
The industry has had the label for thirty years. The platform filled a vacuum that existed because hotels did not own their demand. Understanding the mechanism is not an argument against OTA distribution. It is the prerequisite for knowing what a shift from rented demand to owned demand actually requires, and why that shift has to begin before the traveler reaches the platform, not after.
Where the Response Has to Begin
Americas Great Resorts works upstream of this system, before the traveler enters the OTA’s search, ranking, and personalization environment. That is where the shift from rented demand to owned demand begins.
Once the mechanism is understood, the strategic implication is unavoidable. The response cannot start inside the OTA system. It has to start before the traveler enters it.
