Booking.com
An online accommodation booking platform.
Booking.com shows up as a performance marketing and experimentation heavyweight. It is cited as a high-ROI arbitrage business that can keep building on performance marketing, and it reports 80 to 90 percent experiment failure rates. Lesson: expect most tests to fail and plan for it. It also pays affiliates like WeWard only on sales, a low-risk channel.
- 1Pitch brands directlyApproached brands and affiliate platforms
- 2Promote the brandWeWard offers promotion for a commission
- 3Tracking records salesExisting tracking technology already in place
- 4Brand pays commissionPaid only on sales, little risk
From Yves Benchimol on WeWard
What was said about Booking.com
Booking.com expects 80 to 90 percent of its experiments to fail. Booking.com reports experiment failure rates of 80 to 90 percent, in line with Google Ads and close to Ronny Kohavi's own teams: about 66% at Microsoft, about 85% at Bing and 92% in Airbnb search relevance.
Kohavi's point is that the more optimized a product is, the harder it is to win, so a high failure rate is normal and should be planned for. He counts failure per experiment, not per idea, since one idea may take several iterations or bug fixes before it launches. Quote: "be ready to fail eighty percent of the time".
Booking.com is one of few businesses that can build on performance marketing. Brian Chesky cites Booking.com as a high-ROI arbitrage business that can build permanently on performance marketing, unlike most companies, where it creates no lasting advantage.
Booking.com also buys performance channels beyond search ads. It is an affiliate partner of the walking app WeWard, paying only a commission on sales through existing tracking technology, alongside Nike, Adidas, Expedia and Amazon.
Twenty years in, Booking.com still has not fully solved matching. Preply's Simon Mizzi points to Booking.com working on matching for about 20 years without fully solving it, as a sign that tutor matching will stay unsolved for years.
He adds that marketplaces like Booking.com optimize the platform, not what happens after the booking.
How WeWard hit 20M users without paid ads — Yves Benchimol, WeWard
One of WeWard's affiliate partners, paying only a commission on sales.
Booking.com is one of few businesses that can build on performance marketing. Brian Chesky cites Booking.com as a high-ROI arbitrage business that can build permanently on performance marketing, unlike most companies, where it creates no lasting advantage.
Booking.com also buys performance channels beyond search ads. It is an affiliate partner of the walking app WeWard, paying only a commission on sales through existing tracking technology, alongside Nike, Adidas, Expedia and Amazon.
How Preply Built the World’s Largest Language Learning Marketplace
Cited as having worked on matching for about 20 years without fully solving it, a sign that Preply's tutor matching will stay unsolved for years. Marketplaces like it optimize the platform, not what happens after the booking.
Twenty years in, Booking.com still has not fully solved matching. Preply's Simon Mizzi points to Booking.com working on matching for about 20 years without fully solving it, as a sign that tutor matching will stay unsolved for years.
He adds that marketplaces like Booking.com optimize the platform, not what happens after the booking.
Brian Chesky’s new playbook
Cited as one of the few high-ROI arbitrage businesses that can build permanently on performance marketing, unlike most companies where it builds no lasting advantage.
Booking.com is one of few businesses that can build on performance marketing. Brian Chesky cites Booking.com as a high-ROI arbitrage business that can build permanently on performance marketing, unlike most companies, where it creates no lasting advantage.
Booking.com also buys performance channels beyond search ads. It is an affiliate partner of the walking app WeWard, paying only a commission on sales through existing tracking technology, alongside Nike, Adidas, Expedia and Amazon.
The ultimate guide to A/B testing | Ronny Kohavi (Airbnb, Microsoft, Amazon)
It reports experiment failure rates of 80 to 90 percent, similar to Kohavi's teams.
Booking.com expects 80 to 90 percent of its experiments to fail. Booking.com reports experiment failure rates of 80 to 90 percent, in line with Google Ads and close to Ronny Kohavi's own teams: about 66% at Microsoft, about 85% at Bing and 92% in Airbnb search relevance.
Kohavi's point is that the more optimized a product is, the harder it is to win, so a high failure rate is normal and should be planned for. He counts failure per experiment, not per idea, since one idea may take several iterations or bug fixes before it launches. Quote: "be ready to fail eighty percent of the time".
More in Travel
TravelAirbnb
A marketplace where hosts list unique homes and travelers book stays in 220 countries.
46 sources · 87 tips
TravelUber
A ride-hailing app that connects riders with nearby drivers.
27 sources · 65 tips
TravelLyft
A ride-hailing app connecting riders with drivers, positioned as a semi-managed marketplace.
7 sources · 14 tips
TravelWaymo
A fully driverless ride-hailing service that operates autonomous cars in select US cities.
8 sources · 12 tips
TravelSouthwest Airlines
A low-cost US airline that flies only Boeing 737s and focuses on short-haul routes.
4 sources · 22 tips
TravelAllTrails
A trail discovery and navigation app for hiking, trail running, mountain biking and backpacking.
4 sources · 20 tips
TravelBYHOURS
A booking platform for regular hotel rooms rented by the hour for short stays.
1 source · 12 tips
TravelNomad Table
A solo travel app where users post activities and nearby travelers get notified and can join.
1 source · 16 tips