Every Monday morning, at a certain kind of independent hotel, the revenue manager opens the CRS, exports a channel report into a spreadsheet, and manually reconciles it against the PMS. The columns are always the same: Booking.com, Expedia, Agoda, HRS, direct, GDS. She adds a column for commission. She multiplies. She stares.
Then she walks into the owner’s office with a number.
That number, in 2026, is getting worse. Not better.
The war is not stalling. It is reversing.
The industry has spent a decade on direct-booking campaigns. “Book Direct.” Loyalty discounts. Best-rate guarantees. Rate parity pressure. Branded landing pages. The strategic press has covered every new tactic.
None of it is working, and the data is finally blunt enough to say so out loud.
In 2024, online travel agencies captured 61% of bookings at independent hotels, up from 60% the year before, according to Phocuswright data reported by PhocusWire. North America OTA share hit 52.7% in 2025, up 3.3 percentage points year over year. These are not rounding errors. They are a decade-long trend that has continued through every direct-booking campaign the industry has run.
Independents are not holding the line. They are losing ground, slowly, every quarter.
The tax nobody itemizes
Call it what it is: a commission tax, a lead-generation tax, a distribution tax. A tax paid not to a government but to a distribution platform, in exchange for guests the platform often did not originate and in many cases had never heard of until the guest typed the hotel’s name into a search bar.
The rates, as documented by Cloudbeds and Preno in their 2026 distribution guides:
- Booking.com: 15% baseline, with most properties paying between 15% and 25% after Preferred Partner, Genius, and market modifiers.
- Expedia: 15–30% for independents, 10–15% for major brands. The spread is the leverage: big chains negotiate; independents sign the standard agreement.
- Smaller OTAs (HRS, Agoda, others): variable, frequently higher.
The cleanest all-in benchmark comes from Kalibri Labs, which analyzed multi-year transactional data across thousands of hotels in its Demystifying Distribution and AHLA Demystifying the Digital Marketplace research:
Hotels pay an average of 13 percent to 17 percent of a night’s room rate to acquire guests through OTAs, compared to 3 percent to 8 percent through brand.com.
Kalibri Labs, Book Direct: The Numbers Tell the Story
Per occupied room, Kalibri puts the property-direct cost of acquisition at $2.61 (phone and walk-in), brand.com at $9.21, and OTA at $22.43. That is the shape of the tax. An eight-times markup, every time, on every qualifying stay.
A worked example (the one the revenue manager runs in her head)
One hundred rooms. 75% annual occupancy. $150 ADR. 40% of room-nights sold through OTAs, which is roughly the independent average.
- 100 rooms × 365 nights × 75% = 27,375 occupied room-nights
- × $150 ADR = $4,106,250 gross room revenue
- × 40% OTA mix = $1,642,500 OTA revenue
- × 20% blended commission = ~$328,500 / year in OTA commissions
That is the size of the check, at a modest property, at an ordinary commission rate. Call it one and a half front-desk salaries. Call it the downpayment on a renovation. Call it the difference between a break-even year and a healthy one.
Now look at the column next to it: the 60% of room-nights sold direct. At Kalibri’s brand.com benchmark of 3–8% acquisition cost, that same revenue costs the property roughly $123,000 to acquire, versus the $328,500 we just wrote to Booking.com and Expedia for their smaller slice. The OTA portion of the business costs roughly 2.7 times more to acquire per dollar of revenue than the direct portion. That is the gap the industry has been trying to close for a decade, and it is widening.
It isn’t all OTA loss. The billboard effect is still real.
No serious piece on distribution should pretend the OTAs are pure overhead. They are not.
The billboard effect, originally identified by Cornell’s Chris Anderson in his 2011 paper The Billboard Effect, analyzed 1,720 InterContinental Hotels reservations between 2008 and 2010 and found that each OTA reservation produced between 3 and 9 additional direct bookings. Anderson’s 2017 update, Still Alive and Well, confirmed the effect persists, attenuated, but real, with roughly 65% of direct bookers having visited an OTA at some point in the search funnel.
So the honest answer is not “leave the OTAs.” The honest answer is: the OTAs are a marketing channel with a very expensive close rate, and the hotel’s job is to convert more of the traffic they generate into direct bookings at the moment of commitment.
The 18% opportunity
This is the part the industry has not yet made enough of.
According to SiteMinder’s 2025 Hotel Booking Trends report, which analyzed 125 million reservations across 44,500 hotel customers and $50 billion in bookings:
18% of travellers who start their search on an OTA ultimately book directly with hotels, up 3.3 percentage points.
SiteMinder, Hotel Booking Trends 2025
Translation: roughly one in every five OTA searchers is, in fact, persuadable. They start on an OTA. They land on the hotel’s website or call the front desk to confirm something. And they close on the direct channel, if the hotel catches them.
SiteMinder’s same dataset shows direct bookings averaging $519 per booking versus $320 on OTAs, a 62% revenue-per-booking premium on the direct channel, over and above the commission savings.
The economics are not subtle. The question is purely operational: can the hotel catch that direct-bound 18% when they reach out?
In 2026, at most independent hotels, the honest answer is: no. The call goes to voicemail. The website chat widget is a glorified contact form. The email response comes 18 hours later. The direct-bound guest reverts to the OTA, and the property pays 20% to close the booking it was five minutes away from closing for free.
Net Distribution Cost: the metric the dashboard is missing
Every GM knows RevPAR. Every GM tracks ADR, occupancy, and GOPPAR.
Very few GMs track Net Distribution Cost (NDC): the blended all-in cost to acquire a guest, expressed as a percentage of room revenue, after commissions, loyalty program costs, payment processing, and channel-marketing spend.
Kalibri’s data suggests the industry benchmark for NDC is somewhere between 15% and 25% of room revenue at most independents. At a well-run property with strong direct share, NDC can fall into the 9–12% range. At a commission-dependent property, it balloons past 22%.
We will argue, in a future post, that NDC belongs on the GM’s morning dashboard, next to RevPAR. The reason is simple: RevPAR measures how much you earned. NDC measures how much of that you got to keep.
A hotel that drives RevPAR up 6% while NDC climbs from 18% to 21% has made the owner poorer. The industry has been celebrating those quarters for a decade.
The exit isn’t more marketing. It’s better conversion at the moment of contact.
The conventional answer to the OTA tax is a bigger direct-booking budget. More SEM, more retargeting, more loyalty discounts. Those tactics chase the top of the funnel.
The opportunity the data is pointing to is lower in the funnel: the 18% of OTA searchers who are already on their way to a direct booking, who call or visit the site with intent to close, and who are lost because the hotel is not equipped to meet them.
That loss has a specific shape. It is the call the front desk missed because the receptionist was checking in a family of four. It is the website form that took 16 hours to answer. It is the chat that defaulted to “email us at hello@.” It is the guest who wanted to book direct and could not figure out how, who reverted to Booking.com, and whose conversion became a 20% commission.
Closing that gap is not a marketing problem. It is an infrastructure problem. You need a system that answers every call, responds to every inbound inquiry in real time, recognizes the returning guest before the phone stops ringing, and makes the direct-booking close the path of least resistance instead of the path of most friction.
That is the product FlowStay builds.
Not because the OTAs are bad. They are not bad. They are expensive, they are effective at lead generation, and for a long time they were the only scalable way for an independent hotel to reach guests at the moment of intent.
That window has closed. The technology to catch the 18% directly, to convert them before the OTA does, is here, and it is independently affordable for the first time.
Every week an independent hotel delays is another week of 20% commissions on bookings that could have closed direct. The Monday morning spreadsheet knows.
It has been telling the revenue manager for years.