Usually yes at 8 to 25 rooms, for a duller reason than the commission everybody quotes: some of your guests already know your name and are being routed through Booking.com anyway, and a checkout on your own domain simply stops that. It is the wrong move in four cases, which are that nobody searches for your property by name, that your rooms are still sold from more than one calendar, that your own price is already the higher of the two, or that the business is too small to carry a first year of software and setup. We put that last line at $350,000 a year of bookings and turn down anything under it. The arithmetic below takes about ten minutes with your own occupancy and rate, and the free half of the work is worth doing whichever answer you get.
The four times the answer is no
Nobody is searching for your name. A booking page can only convert demand that is already there, and it will not generate a single new enquiry on its own. If the only way a stranger finds you is by scrolling a platform's search results, a checkout on your own domain has nothing to catch, and you will end up paying for clicks to fill it. Put the year into the things that make people look you up: the Google listing, the photographs, the local press, the guests who already came.
Your rooms are sold from two calendars. Somebody keeps a spreadsheet, or the front desk holds a room that the platform never hears about. Sell from a third place as well and sooner or later the same room goes to two people, and the one you have to call is usually the guest who booked with you direct. Get everything onto one calendar first. What a channel manager does, and how to check the one you probably already own is the half hour that fixes this.
Your own price is higher than the platform price. Price one open week both ways on the same afternoon, with every tax and fee counted, then look at the two totals. If yours is the higher one, every visitor you send to your own site is being shown the worse deal, and a better website will only help you lose those bookings faster.
The money is too small. Below roughly $350,000 a year of bookings, the commission a direct channel can realistically win back does not cover twelve months of software, setup and your own attention, which is why we set our cutoff there and turn work down under it. Our about page carries the same figure.
You can fix the calendars, the price gap and the name recognition, and none of it takes longer than a quarter of work. Revenue climbs over time without you buying anything, so the fourth reason tends to solve itself.
The sum that settles it
Three numbers go into this calculation.
Your real commission rate. Look it up rather than taking a figure from an article. Booking.com says plainly that the exact percentage "depends on your country, property type, and the accommodation agreement you signed", and tells partners to find theirs in their accommodation agreement or under Reservation statements in the Extranet's Finance tab. Airbnb publishes its own: under the single fee most hosts pay 15.5%, the rest typically 14% to 16%, and it applies to traditional hospitality listings and to anyone whose listings run through property management software.
The share you could actually move. Most people skip this one, put their whole platform revenue into the sum and come away disappointed a year later. What belongs here is only the slice that would have reached you anyway had you made it possible: repeat guests, people who type your name into Google, the ones who call, the wedding party whose organizer emails you. Guests who found you by browsing a map on a platform will keep booking on that platform.
What the direct path costs to run. Software from nothing to about $150 a month depending on what you already own, plus 2.9% and 30 cents a card payment on Stripe's published US rate, plus whatever the build or the builder costs, plus the hours. One thing to net off before you decide the card fee is new money: if a platform pays you by virtual card, that payment already runs through your own card terminal and you are already paying to process it.
No commission calculation contains the hours, and they are substantial. A platform booking arrives with the platform standing between you and the guest. They field the pre arrival questions, they apply their own cancellation terms, and when a card is disputed they are the ones talking to the bank. On a direct booking every one of those jobs is yours. On a direct booking somebody at your property handles the parking question at nine on a Sunday, the refund after a canceled flight and the paperwork when a charge is disputed months later. That work is usually worth doing, because it is also where repeat guests come from. It is not free, and it lands on whoever is already the busiest person at the property.
Take the booking value you think you can move, multiply it by your commission rate less about three points for the card, and hold the answer against a full year of that third number.
Here it is on a property where the answer is no. Nine rooms at $135 a night, full 55% of the year, sells 1,806 room nights and turns over about $243,810. Say three quarters of that arrives through platforms, so about $182,858, and say you could genuinely move 8% of it, which is around $14,629 of booking value. At a 15% commission that saves about $2,194, and the card fees on it take back roughly $430. You are left with about $1,764 for the year, against software at $468 to $1,800 and a build on top. The two roughly cancel out, so at that size the project earns nothing in its first twelve months and should wait.
Now a property where the answer is yes. Twenty two units at $205, full 70% of the year, sells 5,621 nights and turns over about $1,152,305. Two thirds through platforms is about $749,000. Move 15% of that, around $112,350 of booking value, and a 15% commission saves about $16,852 while the cards take back about $3,318. The return is about $13,534 in the first year and roughly the same every year after, while the build is paid for once. At that size it covers its own cost with months to spare.
Run it with your own occupancy and your own rate before you believe either example, and be mean with the share you think you can move. If you want each step of the direct plan priced separately rather than as one decision, every step and what it costs sets them out in order of expense.
How to find the share you can move, without guessing
Open your Google Business Profile and look at how many of last month's searches were for your property by name rather than for a category like hotels near the station. Add the guests who have stayed with you more than once. Add the bookings that already arrive by phone and email, the ones somebody writes into the calendar by hand. That total is roughly the demand that is already yours and is currently being routed through a platform that charges you for the privilege.
A small total means not enough people know the place exists, and the money belongs in marketing before it goes anywhere near software. Once the total reaches a decent fraction of your year, that demand is already yours and you are paying a platform to hand it back to you.
If the plan is to be cheaper on your own site, read the contract first
This is where most of the advice written for American owners is quietly wrong, because it was written about a market with different rules.
Under Booking.com's own explanation of parity a US or Canadian property is on wide parity, which it defines as giving Booking.com the same or better rates, conditions and availability than you make available on any channel of your own, online or offline. That reaches your own website, your app, your phone line and your front desk. The one way out is a closed user group rate, and it has real conditions. Members opt in. They sign in with a password. They have a profile. They have already booked with you once as a member. And the rate is never shown publicly. A discount that any passing visitor sees in your booking flow is not outside parity in the US or Canada, whatever a European case study told you. What your parity clause actually covers goes through it properly.
Be careful with the workaround everyone suggests. Emailing a promo code to past guests does not make a closed user group rate on its own, because the price still sits in a public booking flow for anyone holding the code. What qualifies is a rate a guest has to sign in to see, attached to a profile, after they have joined and booked with you once as a member. Plenty of small booking engines cannot do that at all, which is worth asking about before you plan around it. Where yours cannot, the lever you still have is everything that is not the price: the late checkout, the parking space, the room you actually choose for them, the cancellation terms you are allowed to set for yourself.
At a 15% commission you keep about 85% of the price, and after the card fee there is room for a direct discount of about 12 to 13% before you are worse off than you were on the platform. Discount by 10% to win the booking and you keep two or three points of margin on it.
Whatever price you do settle on, show it correctly. Since May 2025 the FTC's rule on unfair or deceptive fees has covered short term lodging, your own website included: whatever price you show has to show the total, with every mandatory fee in it, more prominently than anything else on the screen. Taxes can sit outside that figure but have to appear before the guest pays.
What made us stop treating platform demand as ours
I co-own MINT @Naschmarkt, 18 serviced apartments in five apartment types in Vienna. In late June our Booking.com arrivals fell to about 1.4 to 1.9 bookings a day and stayed there, flat for five weeks, about 75% off the mid June peak. Nothing was broken. We checked the connection first, the way everybody does, and the calendar was syncing perfectly.
What had happened was that the dates people were searching for were already sold out with us, and a listing with little left to sell gets pushed down the results. The platform quietly stopped showing us to the people who were still looking, and nobody at the property was told, because there is nobody there to tell you. We had been treating that flow of guests as ours when we were only ever renting it.
The strongest reason to build your own channel is the risk you carry while every guest arrives through a company you cannot telephone, and that risk is worth more than the 15%. One caveat before you copy anything we do on price: we are in Austria, where Booking.com dropped parity for a property's own website, so we can advertise a direct rate that a US or Canadian property cannot. Your contract is stricter than ours.
What to do in the meantime, all of it free
Here is what to do in the meantime, in the order I would do it.
- Fill in the Google Business Profile properly and switch on Google's free booking links. Google charges nothing for those links and running paid hotel ads does not change where they rank. You need a Business Profile and your live rates reaching Google Hotel Center, which in practice means asking your property system whether it sends them.
- Turn on the booking engine you may already be paying for. Open your property system before you shop for anything. Little Hotelier puts a booking engine in its cheapest plan, Cloudbeds adds no commission of its own to bookings through its engine, Sirvoy charges no commission on any plan, and Beds24 gives you a booking page you can put your own web address on. Plenty of owners are paying for one of these and have never switched it on. If yours does not have one, freetobook gives away its booking engine, diary and property system with no commission at all, checked on September 2026. Run a plain checkout somebody else maintains for a year and you will find out what your direct demand actually is. What fifteen booking engines cost if you find you own nothing.
- Price the same week on both channels and close the gap. Do it today. It costs an afternoon and it is the single thing most likely to be quietly costing you bookings.
- Write to the guests who already stayed. Send one message with a link in it and a reason to use the link. The email that is not a newsletter has the shape of it.
- Answer the phone with the calendar open. Make sure whoever picks up at seven on a Saturday evening can see availability and hold a room there and then, at no cost to you.
- Hand every departing guest a card with your own booking link on it. Print your web address, a phone number and one reason to use them. It costs pennies and it reaches the only people who were ever going to book direct anyway.
- Start recording where every booking came from. In a year you will want to answer this question with your own numbers, and nobody can reconstruct it afterwards.
Those seven jobs deliver most of what a direct channel does, and after a year of them you will have the number the sum above needs.
Who we do not take on
We turn down properties booking under roughly $350,000 a year, because the commission saved at that size will not pay back a build inside its first year and we would rather say so than take the money. If you are above the line, or close enough to argue about it, send your occupancy, your rate and your commission percentage and we will run the sum with you and tell you honestly which side of it you fall on. Book a call.
