Uber bought a hotel tab the year it stopped mattering
The super app push reinforces the booking channel at the moment the deciding has moved upstream
On April 29, at its annual GO-GET event, Uber announced an in-app hotel booking tab powered by Expedia. The launch covers around 700,000 properties for users in the United States, with Vrbo to follow later in the year. Uber One members get rotating 20% discounts on a list of more than 10,000 hotels and 10% back in Uber credits. The reciprocal leg of the deal puts Uber rides inside the Expedia app from June. Dara Khosrowshahi, who ran Expedia for twelve years before joining Uber, framed the move as part of Uber becoming "an app for everything."
Most coverage settled on a single frame: Uber is late to the super app. That framing accepts the premise. The harder question is whether the super app is the right architecture for travel at all in 2026. Probably not — and not because Uber is late. The reason is that the layer Uber is reinforcing, the booking channel, has been demoted by the layer above it.
The deciding has moved upstream
Hotel discovery has separated from hotel booking. The comparison work — choosing where to go, narrowing a shortlist, setting a rate expectation — has moved upstream. It happens on social platforms, in AI search, inside chatbots, across creator content, and increasingly through agentic tools that gather and synthesize. By the time a traveler reaches a booking channel, the property is largely chosen and the rate is already framed.
A super app strategy is, by construction, a booking-channel strategy. It assumes the user opens an app, browses inventory, and transacts inside that app. Uber's hotels tab inherits all three assumptions, in the layer where strategic value is eroding fastest.
Super apps need empty verticals
The super app pattern works in growth-stage markets where the vertical the super app expands into is uncontested. WeChat captured payments before banking was fully digitized. Gojek and Grab captured food delivery before food delivery had consolidated. Each new vertical landed in a category no large player had yet locked in.
US consumer travel booking is the opposite condition. Booking and Expedia have spent two decades building inventory, loyalty, and metasearch presence. Marriott, Hilton, and Hyatt have built direct booking economics, app-based loyalty, and member-rate exclusivity. Google Hotels sits across the comparison surface. Airbnb owns its category. There is no vertical left empty for a super app to fill. A hotels tab can still be added in this kind of market — it cannot take the vertical. It becomes another funnel into existing inventory infrastructure, which is exactly what Uber's tab is, since the inventory is Expedia's.
The deal favors Expedia
The more interesting party in the partnership is Expedia, not Uber. Uber gets a hotels tab that uses someone else's inventory, a take rate on bookings, and an Uber One retention lever. Expedia gets a wholesale traffic source from a 150-million-user app at zero acquisition cost. The reciprocal leg — Uber rides inside the Expedia app from June — confirms the direction. Expedia is integrating mobility into its own checkout. Uber is renting Expedia's inventory.
Expedia controls the inventory, the rate, the merchant-of-record economics, and the customer relationship at the booking moment. Uber controls a tab. If hotel volumes become meaningful, Expedia keeps the strategic position; if they remain modest, Expedia paid almost nothing for the experiment. The structure of the partnership favors Expedia because Expedia owns the inventory layer. That fact would hold under any negotiation.
Booking.com is making the opposite bet
Booking.com has spent the past year visibly repositioning around AI, with material reinvestment in AI products and partnerships designed to keep it present in conversational and agentic surfaces. The premise is straightforward: the deciding part of the funnel is moving upstream of the booking channel, and the booking channel can no longer assume the comparison work will keep coming to it.
Uber's hotels tab is the opposite bet — that the deciding will keep happening inside apps users open intentionally, and that bundling rides, food, and hotels in one transaction surface is enough to capture meaningful share of travel intent. Two of the most travel-fluent commercial operators in the consumer internet are pointing in different directions on the same question. That, more than the super app framing, is the part of this announcement worth investor and chain-level attention.
The architecture is the signal
Hotels on Uber will record real bookings. Bundling will move some volume in the early months. None of that resolves the strategic question, which is which architecture travel discovery consolidates around. Super app architecture treats the app as the destination — the user opens the app, the app contains the categories, the categories share a checkout. Agentic architecture treats the app as one of several entry points and the model as the comparison surface — the user describes the trip, the model assembles options across categories, the booking lands wherever the model hands off.
Both will exist. The question is which one absorbs more of the deciding. Uber has chosen to invest in the surface most preserved by the present and most exposed by the near future. The booking volume will eventually answer whether that was the right call. The press cycle has already answered the wrong question.
by Markus Busch, Editor/Publisher Hospitality.today
Read also: The hotel discovery layer: booking channels have stopped doing the work
Enjoying this analysis? Hospitality.today delivers daily insights on hotel distribution, AI trends, and travel commerce — straight to your inbox. Subscribe for free at Hospitality.today →