Written by Stanislav Sadovnikov, Founder, Magnum Estate · Last updated 2 November 2026
Direct Bookings 2026: the short answer
18% of travellers who begin their search on an OTA go on to book directly with the hotel — up 3.3 percentage points. They do it for control and better service, not for a discount. For a Bali property paying commission on every platform booking, that 18% is the most valuable number in the distribution data, and most operators do nothing deliberate to earn it.
Direct booking is not a marketing campaign. It is an operational discipline.
What the behaviour actually is
The pattern is research on the platform, then a deliberate switch. The guest has already compared, already decided roughly where they want to stay, and is now looking for a reason to transact with you rather than with the intermediary.
That is a narrow, winnable moment. It is also a different problem from being discovered in the first place — with 26% of searches now starting on an OTA, discovery and conversion have separated.
What the data says guests want at that moment: control and better service. Not primarily price. The traveller switching to direct is looking for flexibility, a human they can reach, and confidence that the booking is correct.
Two related findings sharpen this:
- 65% of travellers agree hotels should adjust rates during busy periods — dynamic pricing is broadly accepted when the logic is transparent. Direct channels can explain; OTA listings cannot.
- Australian travellers lead direct booking rates globally and show the strongest willingness to pay for breakfast add-ons and arrival flexibility. With Australia at 23.51% of Bali’s August arrivals, that is the single most important alignment in this dataset for a Bali operator.
What a direct booking is worth
More than the saved commission, and the difference matters when deciding how much to invest in winning one.
| Component | Direct | OTA |
|---|---|---|
| Commission | None | Deducted from room revenue |
| Guest data | Yours | The platform’s |
| Communication before arrival | Direct | Mediated |
| Upsell and ancillary opportunity | Full | Limited |
| Repeat booking | Accrues to you | Accrues to the channel |
| Cancellation terms | Yours to set | Platform-influenced |
The ancillary point is the under-counted one. Pre-arrival contact is where airport transfers, spa bookings, dining reservations and room upgrades are sold. Those are total-revenue items, and they are far harder to capture when the platform owns the conversation.
What actually moves the number
Rate parity honesty. Most platforms require rate parity. Where you have latitude, it is usually in value rather than price — a included breakfast, a guaranteed early check-in, a flexible cancellation. Australians in particular pay for exactly these.
A booking engine that works on a phone. The switch from OTA to direct happens on mobile, in under a minute, and a slow or confusing booking flow loses it. This is the single most common failure and the cheapest to fix.
A reachable human. “Control and better service” means a message answered quickly by someone who can actually change a booking. For a small property this is a genuine advantage over a platform’s support queue.
Reviews and recommendation. With word-of-mouth doubled to 14% of research starting points, a guest arriving on a peer recommendation often comes straight to the property. That is a direct booking won earlier in the funnel.
Repeat guest discipline. The cheapest direct booking is a returning guest. Repeat guest rate belongs in the monthly report.
The honest limits
You cannot win the 26% who only ever use the platform. A material share of travellers want the platform’s protection, loyalty programme and single point of contact. Treating them as a conversion failure wastes effort.
Direct is not free. It carries website, booking engine, payment processing, marketing and staff time. The comparison is net of those, not against commission alone.
Discounting direct below the platform rate usually breaches parity and erodes the rate for everyone.
Scale changes the maths. A large resort can justify a revenue and marketing function. A four-villa operation cannot, and for it the right answer may be a strong platform presence plus excellent repeat-guest handling.
Where the 18% actually comes from
Worth being precise about who switches, because the tactics differ.
Guests with a complication. A late arrival, a connecting room, a dietary requirement, an anniversary. The platform cannot easily handle any of these, so the guest looks for a human. This is the largest and most winnable group, and it is won by being reachable.
Returning guests. They already know the property and use the OTA only to check price. A returning guest routed through a platform is a pure commission loss, and the fix is a direct relationship established during the previous stay.
Guests who want flexibility. Cancellation terms are a genuine differentiator where parity rules constrain rate. A more generous direct cancellation policy is a real offer, and it costs only in the cases where it is used.
Price-checkers who find parity. They verify the direct rate is not worse, then book direct for the service. This group is why rate parity discipline matters: a direct rate higher than the platform rate converts them straight back.
The group that will not switch, and should not be chased: travellers who want the platform’s protection, loyalty currency and single point of contact. In a market where 26% of all searches now start on an OTA, that is a large and permanent share of demand. Treating it as leakage wastes effort that the first two groups would reward.
What it means for an owner
Ask for direct booking share as a percentage of room nights, tracked monthly. It is the clearest single indicator of distribution health.
Ask what the direct booking actually costs once the engine, payments and staff time are counted. A 20% commission saved for 15% of cost is a thinner win than it sounds.
Check the booking flow yourself, on a phone, as a guest. Most owners never have. It takes two minutes and frequently explains the number.
Weight the strategy to your source markets. An Australian-heavy property has the most to gain here. An Indonesian-domestic-heavy one, with 38% starting on OTAs, has less.
Treat ancillary revenue as the real prize. The commission saving is immediate and modest. The transfer, spa and dining revenue unlocked by owning the pre-arrival conversation compounds.
At Magnum Estate the booking relationship stays with the team that operates the residence, which is what makes pre-arrival upselling possible at all. We would rather report a lower occupancy with a better revenue mix than the reverse.
One structural note before the metrics. Direct booking share is not a target to maximise at any cost. A property that drives direct share to 80% by underpricing, or by withdrawing from platforms that supply genuine incremental demand, has optimised a ratio rather than a business. The objective is net revenue after acquisition cost, and there is a level of platform dependence that is simply correct for a given property’s size, location and source mix. Knowing where that level sits requires measuring both channels honestly, which is the whole argument of this article.
Limitations and sources
- The 18% OTA-to-direct switch (+3.3pp), the 65% dynamic pricing acceptance, the Australian direct-booking and add-on findings, and the 14% word-of-mouth share: SiteMinder, Changing Traveller Report 2026, almost 12,000 travellers across 14 countries. Published by a hotel distribution technology company — a relevant commercial interest. We have not read the full report.
- These are global figures, not Bali-specific. Application to Bali is our inference from source-market composition.
- Australia at 23.51% of August 2026 direct foreign arrivals: BPS Statistics Bali Province, August 2026 release.
- We have not sourced typical OTA commission rates, booking engine costs or direct-booking cost ratios for Bali and have avoided quoting figures we cannot support.
- The cost components of direct booking are our own operating experience, not a published study.
- Not investment advice.
