The economics of direct bookings: when owning your booking channel actually pays for itself

Is direct booking worth it for short-term rental operators?

For most operators, yes, but later than they think. Direct booking pays for itself once you have roughly 15 to 25 units and can already move 20% of demand to your own site, mostly from repeat and referred guests. Below that, OTA commission costs less than the marketing spend needed to replace it.

That is the short answer. The rest of this piece is the arithmetic behind it, with every assumption exposed so you can substitute your own numbers.

What OTAs actually charge: the published rates

Start with what is verifiable. Not every platform publishes a rate, and the ones that do publish more than one.

Three things follow from that table, and they matter more than the headline percentages.

One: the Booking.com rate you have read on a blog is not a rate. Booking.com does not publish one. Your number is in your accommodation agreement and in your Extranet under Finance, Reservation statements. Go and read it before you model anything. If it is higher than your contract says, you are enrolled in Preferred Partner or Visibility Booster.

Two: the commission base is wider than the nightly rate. Vrbo's 5% applies to cleaning and pet fees. Booking.com charges commission on VAT or GST in most countries. Operators who model commission on room revenue alone understate it.

Three: you pay commission on bookings that never happened. Booking.com charges commission on non-refundable bookings "regardless of whether the guest stayed at your property or not", and on overbookings, because the room was made available on the platform.

What the OTAs earn in aggregate

Platform-level filings give a second, harder anchor. These are consolidated numbers across all products, not accommodation commission rates, so treat them as a ceiling on plausible blended take rather than as your rate.

• Booking Holdings reported total gross bookings of $186,107 million and total revenues of $26,917 million for 2025, stating total revenues at 14.5% of total gross bookings in 2025, up from 14.3% in 2024. Source: Booking Holdings Inc. Form 10-K for the fiscal year ended December 31, 2025, accessed 2026-08-22.
• Airbnb reported revenue of $12,241 million on Gross Booking Value of $91,273 million for 2025, with 533 million Nights and Seats Booked. That is a take rate of 13.4% of GBV. Source: Airbnb, Inc. Form 10-K for the fiscal year ended December 31, 2025, accessed 2026-08-22.

Two independent public companies, two different business models, both landing between 13% and 15% of gross bookings. That is the size of the tax.

What the OTAs spend to earn it, and why that number decides your answer

This is the part most direct-booking arguments skip, and it is the part that actually determines whether owning your channel works.

From the same filings:

• Booking Holdings' marketing expenses were 4.4% of total gross bookings in both 2025 and 2024, and 30.4% of total revenues in 2025. (Booking Holdings 10-K, FY2025, accessed 2026-08-22.)
• Airbnb spent $2,588 million on sales and marketing in 2025 against $91,273 million of GBV, which is 2.8% of GBV. (Airbnb 10-K, FY2025, accessed 2026-08-22.)

Read that carefully. The two largest demand-generation machines in travel accommodation, with global brand recognition, decades of SEO equity, and bidding algorithms tuned on billions of sessions, spend between 2.8% and 4.4% of gross bookings to generate demand.

You will not beat that on cold traffic. A single-market operator buying brand-agnostic search terms against Booking.com and Airbnb is bidding into an auction those two are optimised to win. If your plan for direct bookings is "we will run ads and take the 15%", your customer acquisition cost will land well above 4.4% of GBV, and you will have traded a commission you understood for a marketing budget you do not.

So the real comparison is not 15% versus 0%. It is 15% versus the sum of payment processing, acquisition cost, and platform amortisation. And the only place that sum is reliably small is demand you already own: repeat guests, referrals, and people searching for your brand by name.

That reframes the whole question. Direct booking is not a demand-generation strategy. It is a demand-retention strategy. The OTAs are genuinely good at the first job. They charge you for it forever, including on the guests they introduced you to three years ago.

Direct versus OTA, across the dimensions that actually matter

The last two rows are the honest ones. OTAs have zero fixed cost and near-zero time to value. That is not a small advantage. It is why they are the correct answer for most people starting out, and why "OTAs are the enemy" is a marketing line rather than an analysis.

The model: work out your own break-even

Here is a full worked calculation. Every input is an assumption you should replace.

Assumptions

On the 15% blended take. This is an assumption, not a published rate. It sits between Vrbo's published 8% all-in and Airbnb's published 15.5% host-only fee, and near Booking Holdings' consolidated 14.5% of gross bookings. Your real number depends on your channel mix and your Booking.com agreement. Pull it from your statements. If your mix is Vrbo-heavy your saving is much smaller and the whole case weakens.

On the 1.5% marketing cost. This assumes you are converting demand you already have: repeat guests, referrals, and people typing your brand into a search engine. It is deliberately low because that is the only version of the direct channel that works. If you plan to buy cold traffic, replace 1.5% with something above 4.4%, and the saving per direct booking collapses to roughly 8% or less. Model it honestly.

The formula

Break-even inventory (units)
  = annual fixed cost
  ÷ (net saving rate × GBV per unit per year × direct share)

With the assumptions above:

Break-even units = €18,333 ÷ (0.11 × €48,180 × direct share)
                 = €18,333 ÷ (€5,300 × direct share)

Break-even by direct share

Read that table as a warning, not an invitation. The right-hand column is doing the work. Direct share is not something you choose. It is a function of how much demand you already own.

What it looks like in cash, at a 20% direct share

The five-unit operator is looking at €36,135 of annual commission and correctly deciding not to spend €40,000 building something to recover €5,300 a year of it. That is not timidity. That is arithmetic.

The 110-unit operator is looking at €794,970 of annual commission. At that scale the build pays back in under five months and every subsequent year of direct share is close to pure margin.

Why the repeat rate changes the answer more than inventory size does

Notice that direct share is the term the break-even is most sensitive to. Double your inventory and you halve the break-even. Double your direct share and you also halve it, and doubling direct share is usually cheaper.

Direct share has a floor you get almost free:

Free direct share ≈ repeat guest rate × share of repeat guests you can convert direct

If 8% of your guests return and you convert two-thirds of them directly through email and a booking link, you start at roughly 5% direct share. At that level you need over 65 units before the channel breaks even.

If 30% of your guests return and you convert three-quarters of them directly, you start at roughly 22% before you have spent anything on acquisition. That drops the break-even to around 15 units.

This is why two operators with identical inventory can get opposite answers. A city-apartment operator serving one-off business travellers has almost no free direct share. A coastal villa operator whose guests come back with the same friends every August has a great deal of it.

Measure your repeat rate before you build anything. It is the single most predictive input, and most operators have never calculated it. Take unique guest emails or phone numbers over the last three years, count how many appear more than once, divide by total unique guests. If the answer is under 10%, the case for building is weak regardless of your unit count.

When direct booking is not worth it

Being plain about this is the point. Direct booking is the wrong investment if:

• You have fewer than about 10 units and no meaningful repeat business. The fixed cost is larger than the recoverable commission. Spend the money on photography, pricing, and getting your OTA ranking up.
• Your repeat rate is under 10%. You would be buying the demand twice: once from the OTA that introduced the guest, once from an ad platform to bring them back.
• You are Vrbo-heavy. At an 8% all-in cost, the saving after payment processing and marketing is small enough that the build may never clear.
• Your occupancy is under 45%. You have a demand problem, not a distribution problem. Direct booking does not create demand.
• You do not have someone who will own the channel. A direct site with no one maintaining rates, content, and guest email is a depreciating asset with a hosting bill.
• Your operation runs on spreadsheets. Fix the operational layer first. A direct booking that produces a double-booking costs more than the commission you saved.

There is also a second-order effect worth naming honestly: the billboard effect, where an OTA listing generates direct bookings by acting as a shop window. It is real and it has been measured, in hotels. Cornell's Chris Anderson ran the original experiment in 2009, listing and delisting four hotels from Expedia in alternate weeks, and found that being listed lifted the hotels' own non-Expedia reservations by 7.5% to 26% (Anderson, The Billboard Effect, Cornell Hospitality Report 9(16), 2009). His 2017 follow-up, using a comScore panel, found 65% of consumers who booked directly with a hotel had visited an OTA first, down from 75% in 2011, and put the effect at 5% to 35% depending on method (Anderson and Han, The Billboard Effect: Still Alive and Well, Cornell Hospitality Report, 2017). Both studies are hotel data, and both predate the current short-term rental market, so do not plug either range into your model. The practical lesson is narrower: the OTAs are part of how your direct guests find you, which is one more reason the workable strategy is to keep the listings and move the repeat and referred guests, not to delist and push.

How to get direct bookings without buying demand twice

Assuming the arithmetic clears, the sequence that works is the cheap-demand-first sequence:

1. Capture guest contact details on every stay. Pre-arrival messages, check-in instructions, and the departure note are all consented touchpoints. This is the asset. Nothing else matters until it exists.
2. Own your brand search. People who have stayed with you will type your name. Make sure the top result is your booking page, not your Booking.com listing.
3. Make the direct price defensibly better. Not a rate-parity violation. A better package: late checkout, transfer included, the fee you do not charge. Guests compare total cost at checkout, not nightly rate on a listing card.
4. Email the past-guest list before the season opens. Highest-converting, lowest-cost direct demand there is.
5. Build the site your inventory deserves. A booking flow that feels less trustworthy than an OTA sends your best guests back to the OTA. The checkout is where the money is won or lost, not the homepage. Whether that means your PMS's own website builder or a custom build is a separate decision, and it turns on inventory size; we have written that comparison up for Hostaway operators in Hostaway's website builder vs a custom booking site.
6. Only then consider paid acquisition, and hold it to a CAC target below your net saving rate. If a direct booking costs more than 11% of its value to acquire, you have rebuilt an OTA with worse economics.

Steps one through four cost close to nothing and produce most of the free direct share. Do them before you commission anything.

What this looks like in practice

Two of the operators we have built for sit on either side of the line described above, which is useful, because it shows what the arithmetic looks like when it is real.

Makarska Exklusiv runs 110 curated luxury villas and apartments across eight towns on Croatia's Makarska Riviera, has served over 100,000 guests, and carries a 4.95 average guest rating. Before the rebuild, high-intent international guests were leaking to Booking.com and Airbnb, so the brand was paying commission on demand its own reputation had already earned. At 110 units, the model above puts annual commission near €795,000 on the stated assumptions, which is why the build was never a close call. The platform shipped in 12 weeks. The founding team's assessment afterwards: "The direct bookings speak for themselves. The platform genuinely elevated our guest experience, and guests arrive already knowing the calibre of stay we deliver." (Source: STYLABS case study; the live site is makarska-exklusiv.com.)

Two things in that sentence deserve emphasis over the commission maths. A 4.95 rating across 100,000-plus guests is a very high free-direct-share input. And an operator handling airport transfers and 24/7 guest support has upsells that no OTA checkout can sell for them. Both make the direct channel worth more than the commission line suggests.

Hireavilla is the other kind of case. It runs curated villas across North Goa, Alibaug, Mangalore, Bali and Dubai. Before the rebuild, inventory lived across spreadsheets and channel managers and bookings ran on phone calls and manual confirmations. There, the direct-booking site was not the first problem. The operational layer was. We built the marketplace and an ERP that gives the team one view of inventory, bookings and guest communication across all destinations, because a direct channel sitting on manual availability produces double-bookings, and a double-booking costs more than the commission it saved.

That is the general lesson and it is worth more than either case study: the operational readiness test comes before the arithmetic test. If your availability is not reliable in real time, your break-even calculation is optimistic by whatever your double-booking rate costs you.

Both operators, and two more, run on Hostaway with a custom site and dashboard on top. What that layer consists of is documented at /work/hostaway.

Frequently asked questions

What is the Booking.com commission rate? Booking.com does not publish a single rate. Its partner help states the exact percentage depends on your country, property type, and the accommodation agreement you signed. Find yours in your accommodation agreement, or in the Extranet under Finance, then Reservation statements. Preferred Partner and Visibility Booster add a further percentage on top.

How much does Airbnb charge hosts? Airbnb publishes two structures. Under the split fee, most hosts pay 3% and guests pay a service fee of 14.1% to 16.5% of the booking subtotal. Under the single or host-only fee, most hosts pay 15.5%, with the remainder typically paying 14% to 16%. Software-connected hosts and hotels generally use the single fee.

Is direct booking cheaper than Airbnb for the operator? Only after fixed costs are covered. A direct booking still costs roughly 2% to 3% in payment processing plus your marketing spend plus amortised build and running cost. On typical assumptions the net saving is around 11 percentage points of booking value, which only beats commission once your inventory and direct share are large enough.

How many units do I need before building a direct-booking site? On the assumptions in this article, roughly 17 units at a 20% direct share, or about 35 units if only 10% of bookings go direct. Below 10 units with weak repeat business, the fixed cost exceeds the commission you can realistically recover. Substitute your own ADR, occupancy and commission rate.

Does a direct-booking site mean leaving the OTAs? No, and it usually should not. OTAs generate discovery you cannot replace cheaply. Booking Holdings spent 4.4% of gross bookings on marketing in 2025. The workable model is to keep OTAs for new-guest acquisition and move repeat and referred guests to your own channel, where you keep the margin.

What is the fastest way to increase direct bookings? Collect guest contact details on every stay, then email past guests before your season opens. That demand is already yours and costs almost nothing to convert. After that, make sure your brand name search returns your booking page rather than your OTA listing, and give direct guests a better total package rather than a lower nightly rate.

How do I calculate what OTA commission is costing me per year? Multiply your average daily rate by 365, then by your occupancy rate, then by your unit count, then by your blended commission rate. At €220 ADR, 60% occupancy and 15% commission, that is about €7,227 per unit per year. Use commission from your actual statements, not a published headline rate.