14 June 2026
Direct bookings vs OTA in 2026: the numbers you need to plan by
The direct-booking debate has been running for fifteen years and every year the same argument gets recycled: OTAs charge 18% and hotels should be able to keep that money for themselves. It is true. It is also incomplete. When we pulled the numbers from 640 properties running on Guest Talk over the last twelve months, five patterns showed up that we think are worth planning around.
1. The 15-25% direct-share range is where most operators live.
Median direct-booking share across our base is 21% for boutique hotels and 26% for hostels. The distribution is tight: 68% of properties fall between 15% and 33%. The top decile (properties above 45%) is dominated by hostels with strong brand identity and repeat guests, and boutique hotels in destinations where the hotel itself is the reason to visit (spa retreats, historic conversions, design hotels featured in travel media).
What this tells you: unless you have a differentiated brand, expect to land in the 20-30% range. Above 40% requires either brand pull or aggressive metasearch spend.
2. Google Hotel Ads free links are the single biggest lever.
Properties that enabled Google Hotel Ads free links saw a median 6-percentage-point lift in direct-booking share within 60 days, with no marginal cost. It is free traffic, high intent, next to the OTA price. If you are not on it, get on it — the feed setup is a two-hour job for any channel manager worth paying for.
Paid metasearch bidding (Trivago, Kayak, TripAdvisor) adds another 2-4 points at CAC that typically sits below the OTA commission you would otherwise pay. But only if you meter the bids carefully. Runaway metasearch spend on low-margin room-nights is a way to lose money quickly.
3. The booking widget matters more than the marketing spend.
Conversion rate on your booking widget is the multiplier on every dollar you spend on direct-booking marketing. Our data shows a 3x range between the worst-performing widgets (0.8% conversion on well-qualified traffic) and the best (2.4%). The difference is not the traffic; it is the widget. A single-page checkout, three-field guest form, transparent pricing including all taxes on the first screen, and no forced account creation together explain most of the delta.
4. Abandoned-cart recovery pays for itself in the first month.
Roughly 60% of hotel booking checkouts are abandoned. A two-step recovery flow (plain reminder at 30 minutes, discount tail at 24 hours) recovers 3-6% of abandoned sessions in our data. On a property doing EUR 100,000 gross monthly with 60% of that from direct traffic, that is EUR 1,800 to EUR 3,600 recovered per month. The typical setup cost is one afternoon.
5. WhatsApp arrival messages lift the direct-booking review score.
Properties that send a pre-arrival WhatsApp message (door code, parking info, arrival window) between 12 and 6 hours before check-in see a 0.2-0.4 point lift in their direct-booking review score, versus a control group of properties that only send email. Review score is a direct input into your direct-booking conversion (via TripAdvisor and Google reviews), so the compound effect is meaningful.
What does not seem to work.
Two tactics that we hear pitched a lot but do not show up in the data: loyalty programmes at properties under 50 rooms (the guest base is too small to build meaningful repeat), and social-media-driven direct booking (organic Instagram traffic converts at half the rate of paid metasearch on the same properties). Skip these until you have exhausted the compound effect of the five above.