Uber's Product Chief Reveals Strategy Behind Hotels, Robotaxis, and the Push Toward a Travel Super App

Uber is no longer just the app you open for a ride or a meal. Over the past year, the company has quietly broadened its footprint to include hotel bookings through a partnership with Expedia, a "shop for me" concierge service, and even boat rentals in Europe. Behind the scenes, it is also building out financial tools for drivers, a data-labeling business, and a six-month-old unit called AV Labs that is outfitting vehicles with sensors to amass driving data.

Chief Product Officer Sachin Kansal recently spoke with TechCrunch about these initiatives, the company's evolving relationship with autonomous vehicle partners like Waymo, and how artificial intelligence is beginning to surface in tangible ways for both riders and drivers.

Travel as the Third Pillar

Kansal explained that travel emerged as the natural theme for this year's product announcements. According to Uber's own data, 1.5 billion trips annually occur outside a user's home city, making travel an already common use case. The headline launch was hotel bookings, powered by Expedia, but Kansal framed travel as a broader ecosystem that encompasses airport rides, food delivery, and local shopping.

The "shop for me" feature was designed to let users purchase items from local stores even if those stores lack a full catalog on Uber Eats. Kansal described travel as the third leg of Uber's stool, joining rides and delivery as a core category.

Integration depth varies by partner. With Expedia, Uber built the entire user interface in-house. For newer categories like European boat rentals, Uber hands users off to a partner's booking flow. Kansal said this approach lets Uber test demand before committing to deeper technical integration.

Financial Services and the Membership Flywheel

On financial services, Uber is taking a measured approach rather than attempting to replicate Asian super-apps. The company currently offers the Uber Pro Card, a debit card that drivers and couriers can use to receive their earnings. Merchant-focused financial products are being tested in select markets, while consumer-facing services remain limited to Uber credits tied to the company's membership program.

Uber One, the company's membership product, has reached 51 million members and now accounts for roughly half of all bookings. Kansal said delivery members typically break even on their monthly fee within two to three orders. Over time, membership increases frequency within a user's primary category and encourages cross-usage — mobility-only users start ordering delivery, and vice versa.

When asked about a potential buy-now-pay-later product, Kansal was noncommittal, emphasizing that Uber prefers to partner with established providers rather than build everything itself. "We're not trying to be everything to everyone," he said.

AV Labs and the Waymo Relationship

Perhaps the most strategically significant initiative is AV Labs, a business unit launched roughly six months ago. The plan is to equip hundreds of vehicles with sensors, deployed through fleet partners, to collect millions of miles of driving data. Uber says this data helps autonomous vehicle partners tackle long-tail edge cases that go beyond typical performance benchmarks.

Beyond raw data, Kansal pointed to Uber's operational expertise as a key asset. The platform's 10 million earners possess institutional knowledge about pickups and drop-offs. Uber also handles 25 million lost items annually, a logistical challenge that autonomous operators will need to solve.

The relationship with Waymo illustrates the complexity of Uber's AV strategy. Uber recently wound down a Waymo pilot in Phoenix that began with about a dozen vehicles, while simultaneously scaling to hundreds of Waymo cars in Austin and Atlanta. Kansal acknowledged that Waymo is both a partner and, in some cities, a direct competitor. He emphasized that Uber is not trying to build its own L4 autonomous system. Instead, the company is focused on creating infrastructure that supports multiple AV providers within a hybrid network of human drivers and autonomous vehicles.

AI Features and the Data Labeling Business

Artificial intelligence is already appearing in features that users and drivers can directly experience. Earners have access to an AI assistant that advises on where to find more demand. On Uber Eats, a grocery cart assistant lets users quickly build a cart from spoken or typed requests like "milk, eggs, bread." Riders can now use voice commands to request trips, including specifying group size and luggage.

Uber has also established a commercial data-labeling business. Earners can transcribe audio and label data for generative AI companies, and Uber sells this service to those clients. Kansal was emphatic that no conversations are recorded during active rides. The labeling work happens only when earners are off-trip, and they are compensated for it.

AV Labs data, by contrast, is a separate stream, and Uber is still determining how to share it with partners. Kansal described that model as being in its early stages.

Looking ahead, Kansal envisioning a future where an AI agent could plan and book an entire trip, though he declined to provide a timeline. He stressed that Uber wants to avoid shipping agents that do not function reliably. As for his own priorities, Kansal said he devotes 70 to 80 percent of his time to strengthening existing and upcoming products, leaving roughly 20 percent for new ideas. He also regularly drives and delivers himself to experience the platform from the other side.

Uber's strategy is clearly expanding in multiple directions at once — travel, autonomous data, financial services, and AI — all while navigating relationships with partners who are also rivals. If you found this breakdown helpful, share it with your network and let us know where you think Uber's product roadmap is headed next.

Source: TechCrunch AI