Airbnb and Uber are moving into adjacent travel categories with the deliberate confidence of platforms that already own the customer relationship. Airbnb now offers experiences, restaurant reservations, and flight searches. Uber has car rentals, trains, and hotel bookings inside the same app that summons a sedan. The logic is clean: own the $1.9 trillion global travel booking market by becoming the single interface between intention and transaction. The problem is that logic has failed twice in the past fifteen years.
The bundling thesis runs like this. Both companies have distribution that legacy travel operators cannot replicate. Airbnb averages 448 million nights booked annually. Uber completes 9.4 billion trips per year. Those are attention moats. Adding a flight search or restaurant widget is low-cost surface area. The customer is already there. The platform takes a percentage. The model scales without new customer acquisition spend. Expedia and Booking Holdings spent a combined $11 billion on performance marketing last year. Airbnb and Uber spent $3.2 billion between them. Bundling is how platforms with distribution leverage it into margin.
The historical record argues otherwise. Expedia launched in 1996 as a bundled product. Flights, hotels, cars, activities—all in one transaction. By 2019, Expedia's bundled attach rate had fallen to 12 percent of bookings. Customers preferred to shop components separately. Booking Holdings tried the same approach with Booking.com and Priceline. The result was modular browsing but separate purchases. The behavior pattern is consistent. Travel is not retail. The decision architecture is different. Shoppers bundle clothing purchases because the cognitive load is low. Travelers do not bundle flights and hotels because the trade-offs are high-stakes and non-linear. A cheaper flight often means a worse hotel, and vice versa. Customers optimize each component separately, even when friction increases.
The margin structure creates a second problem. Airbnb earns 14 to 16 percent take rates on lodging. Uber earns 28 to 31 percent on rides in dense metro markets. Flight bookings generate 1 to 3 percent commissions. Restaurant reservations generate zero unless the platform charges diners directly, which Resy and OpenTable have proven is a narrow and contested market. Activities and experiences carry 15 to 25 percent margins but represent less than 4 percent of total travel spend. Bundling only works if the high-margin product subsidizes customer acquisition for the low-margin product. Airbnb and Uber are proposing the opposite. They want lodging and ride-hailing to subsidize categories with worse unit economics. That works if bundling increases frequency. There is no evidence it does. Expedia's bundled customers book 1.9 times per year. Unbundled customers book 1.7 times per year. The difference does not cover the operational cost of managing four product lines instead of one.
What allocators and hospitality strategists should watch is whether either platform can demonstrate frequency lift in the next eighteen months. Airbnb's Q1 2026 disclosures will show whether users who book an experience also book more nights. Uber's transportation-plus-hotel cohort data will show whether cross-category bundling drives incremental trips or cannibalizes existing margin. If frequency does not move, both companies will face a decision. Keep the bundled surface area and accept margin compression, or quietly deprioritize non-core categories. Booking Holdings made that choice in 2021 and shut down Booking Experiences after five years. The market did not notice.
The real risk is not that bundling fails but that it distracts from the product problem both platforms already have. Gen Z travelers are not opening Airbnb to browse. They are opening TikTok, finding a location, then searching availability. Uber's frequency in non-metro markets has been flat since 2023. Distribution is not the constraint. Discovery is. Building a bundled travel app solves a problem customers do not have. Building a product that understands why someone wants to go somewhere solves the problem both platforms are quietly losing ground on.