Contents9
A £220 booking is not £220. Depending on where it came from, you'll keep somewhere between £155 and £216 of it — and the gap between those two numbers, repeated across a year, is bigger than most of the things you're currently worrying about.
This is about how to work out what a booking from each channel — OTA commission, discounts and fees included — is actually worth to you, and what to do about it. It is not an argument for leaving the OTAs. It's an argument for knowing the price of each one.
If you haven't read Occupancy, ADR and RevPAR, start there — this post assumes you know what RevPAR is. This is the fourth lever from How to increase RevPAR at a small hotel, taken apart properly.
Three things eat a booking, and they stack
Most owners can tell you their Booking.com commission rate. Far fewer can tell you their effective take rate — what the channel costs once everything is counted. Three separate things are at work:
1. Commission. Booking.com states a global average commission of 15%, varying by property type and location. Expedia's headline rate sits in a similar band. This is the number you know.
2. Visibility programmes. These add to the commission. Booking.com's Preferred Partner programme is opt-in and raises your rate in exchange for better placement (check your extranet — the figure is shown against your property). Expedia's Accelerator works on a pay-per-stay basis where you set the compensation rate yourself; TravelAds is pay-per-click on top. None of these are free, and all of them are easy to switch on and forget.
3. Rate discounts. This is the one that hides. Booking.com's Genius programme gives the guest a discount — 10% at Level 1, 15% at Level 2 — and you still pay commission on the discounted rate. It doesn't show up as a cost anywhere. It shows up as a smaller number in the rate column.
That last point is why the arithmetic surprises people. A 17% commission and a 10% Genius discount is not a 27% cost, and it isn't 17% either. Work it through on a two-night stay at £110:
Guest pays | Commission | You keep | Effective take | |
|---|---|---|---|---|
Booking.com, 15%, no Genius | £220.00 | £33.00 | £187.00 | 15.0% |
Booking.com, 17%, no Genius | £220.00 | £37.40 | £182.60 | 17.0% |
Booking.com, 17% + Genius L1 (10%) | £198.00 | £33.66 | £164.34 | 25.3% |
Booking.com, 17% + Genius L2 (15%) | £187.00 | £31.79 | £155.21 | 29.4% |
Expedia, 18% all-in | £220.00 | £39.60 | £180.40 | 18.0% |
Direct, UK card (1.5% + 20p) | £220.00 | £3.50 | £216.50 | 1.6% |
The commission line on the Genius rows is lower than the row above it. That's the trap. Your invoice from Booking.com goes down and your bank balance goes down with it.
One honest note on the direct row: £3.50 is the card fee on a standard UK card. A non-EEA card costs 3.15% + 20p — £7.13 on the same booking — and you'll take plenty of those. It's still not close.
The number to track: net RevPAR
Gross RevPAR tells you what guests paid. Net RevPAR tells you what the building actually earned.
Net RevPAR = (room revenue − distribution costs) ÷ available room nights
Distribution costs means commission, plus the rate you gave away to get the booking, plus payment fees, plus any advertising you ran to win it directly. Nothing else — not staff, not laundry, not the mortgage. Just the cost of the booking arriving.
Two houses can post identical gross RevPAR and be £13,000 a year apart. The only thing separating them is where the bookings came from.
A worked month
Twelve rooms, thirty nights, 360 available room nights. Same house as the last post. Rate £110, occupancy 75%, so 270 room nights sold. The mix:
Channel | Nights | Guests paid | Cost | You keep | Net ADR | Take |
|---|---|---|---|---|---|---|
Booking.com (no Genius, 17%) | 75 | £8,250.00 | £1,402.50 | £6,847.50 | £91.30 | 17.0% |
Booking.com (Genius 10%, 17%) | 75 | £7,425.00 | £1,262.25 | £6,162.75 | £82.17 | 17.0% |
Expedia (18%) | 30 | £3,300.00 | £594.00 | £2,706.00 | £90.20 | 18.0% |
Direct | 90 | £9,900.00 | £337.50 | £9,562.50 | £106.25 | 3.4% |
Total | 270 | £28,875.00 | £3,596.25 | £25,278.75 | £93.62 | 12.5% |
Gross RevPAR: £80.21
Net RevPAR: £70.22
The gap: £9.99 per available room night, every night, occupied or not.
That's £3,596 a month. £43,155 a year, on a twelve-room house. It is almost certainly the largest single line item in your business after wages, and it is the one nobody puts on a spreadsheet.
The direct cost line includes card fees and £12 of Google Ads on a third of direct bookings, because pretending direct bookings arrive free is how people end up with an unusable model. Even carrying that, direct costs 3.4% against 17–18%.
Note the net ADR column. Your gross ADR is £106.94 across the month. Your net ADR is £93.62. If you've been reporting the first number to yourself, you've been reporting a number no bank account has ever contained.
What shifting the mix is actually worth
Take twenty nights from Booking.com and win them direct instead. Same rooms sold, same rate, same occupancy. Nothing else changes.
Baseline | 20 nights shifted | |
|---|---|---|
Room nights sold | 270 | 270 |
Occupancy | 75.0% | 75.0% |
Gross RevPAR | £80.21 | £80.51 |
Distribution cost | £3,596.25 | £3,315.95 |
Net RevPAR | £70.22 | £71.30 |
+£1.08 net RevPAR. +£390 for the month. +£4,684 a year.
I want to be straight about the size of that: it's about £4,700 a year for the ongoing work of winning twenty nights a month direct. Worth having, not life-changing. If someone is selling you a direct-booking strategy on the promise of transformation, this is the honest shape of it at twelve rooms.
The bigger prize is for houses that are further out of balance. Run the same 270 nights at 200 on Booking.com, 40 on Expedia and 30 direct — a very common shape — and net RevPAR is £67.06. Rebalancing that house to the baseline mix above is worth £3.16 net RevPAR, £1,136 a month, £13,635 a year. Identical rate. Identical occupancy. Identical amount of cleaning.
The counter-example: don't buy the nights with rate
The obvious temptation is to fill those twenty extra nights with a discount instead. Cut the OTA rate 10% across the board, pick up twenty more nights, finish at 80.6% occupancy.
Baseline | Discount 10%, +20 nights | |
|---|---|---|
Room nights sold | 270 | 290 |
Occupancy | 75.0% | 80.6% |
Gross RevPAR | £80.21 | £79.90 |
Net RevPAR | £70.22 | £69.97 |
Occupancy up 5.6 points, twenty more rooms to service, and you finish 24p per available room night worse off. The discount applies to everyone, including the guests who were going to book at £110 anyway. This is the same trap as the one in the last post, and it catches people every season.
Cancellations: the bookings you count twice
Cloudbeds' 2026 State of Independent Hotels report, drawn from 90 million bookings, puts OTA cancellations at 21.8% against 10.6% for direct.
Run that forward: to end up with 100 stayed bookings you need to take 128 OTA bookings, or 112 direct ones. Every cancelled booking is a room that sat on your calendar as sold, wasn't sold to anyone else, and came back to you with less notice than you'd like.
The same report has average lead times at 40 days, 47 in EMEA — so an OTA cancellation typically lands with weeks to resell, not days. It's a real cost, but it's a manageable one. Don't let anyone tell you it's catastrophic; do put it in the model.
Your OTA commission is your marketing budget
Here's the framing that makes the decision easy. If you're paying £37.40 in commission on a £220 booking, you can spend up to £33.90 winning that same booking direct and be exactly level.
If the booking would have come via | You may spend up to this, per booking, and break even |
|---|---|
Booking.com at 15% | £29.50 |
Booking.com at 17% | £33.90 |
Booking.com at 17% + Genius 10% | £52.16 |
Expedia at 18% | £36.10 |
Against those numbers, most of what actually wins direct bookings is cheap. A booking engine that doesn't lose people at step three. A phone number above the fold. A rate that's genuinely no worse direct. £12 of brand-name search so the guest who's already decided doesn't get intercepted on the way to you.
You have thirty-odd pounds a booking to play with. Spending nothing is a choice too, and it's the expensive one.
What direct booking doesn't fix
The OTAs are not just a booking channel, they're demand generation, and delisting is almost always a bad idea for an independent. Guests who've never heard of you find you there. Some of them come back direct next time, which is the point.
Four moves that are worth making, roughly in order of return per hour spent:
Audit what you've actually opted into. Preferred Partner, Genius, Accelerator, TravelAds — open the extranet and read your real effective rate. A good number of owners are in programmes they switched on for a slow month in 2023.
Decide about Genius deliberately. It's a real visibility lever and for some houses it earns its keep. Just price it as what it is: a 25%+ effective take rate, not a 17% one.
Make direct at least as easy as the OTA. Most independents lose direct bookings on friction, not price. If your booking form asks for more than the OTA does, you're paying commission for your own web design.
Match the rate, compete on everything else. Rate parity clauses vary and you should check yours, but the flexible bit is usually what you add — late checkout, the good room, parking. Costs you little, doesn't touch the rate.
Then track net RevPAR monthly next to gross. Once you've seen the two lines apart, you won't go back to reading one of them.
FAQ
What's a good direct booking share for a small UK hotel? There isn't a target that fits every house. Cloudbeds' 2026 report puts OTAs at 63.4% of bookings for independents globally, "rising to nearly 80 percent in some markets" — so if you're at 30–40% direct you're doing better than most. A rural B&B with repeat guests can reach 60%+; a city-centre house selling to first-time visitors will sit lower, and that's not a failure.
Should I leave Booking.com? Almost certainly not. The commission buys demand you can't generate yourself, particularly from guests who've never heard of your town, let alone your house. The goal is knowing what it costs and not paying more than you meant to.
Does the Genius discount really cost me more than the commission? On a 10% Genius rate with 17% commission, the discount costs you £22 on a £220 booking and the commission costs £33.66. So no — but together they take 25.3%, and the discount is the half you can't see on an invoice.
Can I charge less on my own site than on Booking.com? Check your contract; rate parity terms differ and have changed in several markets. What's usually safe and works nearly as well is adding value rather than cutting rate — breakfast, an upgrade, late checkout, a bottle of something.
How do I actually measure net RevPAR? Take last month's room revenue, subtract every commission invoice, every card processing fee, and any advertising that ran to win bookings. Divide by rooms × nights in the month. If your PMS reports revenue by channel you can do it in ten minutes; if it doesn't, that's worth fixing — see hidden fees in hotel software.
What about the cost of the software itself? Fair question, and it belongs in the picture. A PMS with a channel manager and booking engine included runs £29–£79 a month at this size — £348 to £948 a year against £43,155 of distribution cost in the example above. If it moves your mix by even a few nights a month it's paid for itself several times over. What a PMS actually costs in the UK has the full breakdown.
