Over the past year, Uber has quietly been pushing ahead with two businesses that most people think of. Of course, there are ride-hailing and delivery services, but if you’ve been using the app long enough, you’ll get access to things like hotel bookings with Expedia, a “Shop for Me” concierge feature, and boat rentals in Europe.
There’s a lot going on internally, so to speak. Consider a six-month-old business unit called AV Labs that is developing debit cards for drivers, a data-labeling side hustle for fellow earners looking to make more money, and a fleet of sensor-equipped vehicles designed to collect more driving data than ever before, separate from Uber’s regular driver network. Uber is positioning the initiative as a way to strengthen its relationships with self-driving car partners, some of which it owns stakes in, but it certainly looks like a hedge. Uber competes directly with some of the same partners, most notably Waymo, and owning the data layer gives Uber both a degree of influence and selectivity.
Whether Uber will become a full-fledged “everything app” akin to Asian superapps like Grab remains an open question. But in this conversation, Uber Chief Product Officer Sachin Kansal shares with TechCrunch the company’s financial services ambitions, its increasingly complex relationship with Waymo, its new AV Labs data operations, and how AI is starting to show up in ways that riders and drivers will actually notice.
This interview has been edited for length and clarity.
TC: You announced hotels, boat rentals, and other shopping features earlier this year. How was that list made? What didn’t make the cut?
SK: Every year, our team obviously builds a lot of stuff, but we decided some of it was worth sharing with the world on the biggest stage. This year, the theme we were drawn to was travel. We know that travel is a very common use case for Uber users, with 1.5 billion rides on the Uber platform actually taking place outside of the user’s city of residence each year. The headline announcement this time was actually a partnership with Expedia to bring hotels to Uber. But traveling is so much more than that. You will need a vehicle to get from the airport to your hotel, and you will also need food. We’ve heard from many users that they’ve stopped using room service and are only using the Uber Eats app. The goal of “shop for me” was to allow people to shop from local stores, even those whose entire catalogs are not available on Uber Eats. In my opinion, traveling is actually the third part of the stool. It used to be rides, then food was added, and now travel is added.
Is Uber moving toward offering its own financial services, like Asia’s “app for everything”?
For us, financial services span multiple different actors, including not only consumers but also drivers, couriers, and merchants. We currently offer multiple products primarily focused on drivers and couriers. There’s something called the Uber Pro Card, which drivers can use as a debit card to transfer all their earnings. We are currently beginning to experiment with some of these products for sellers in certain regions of the world. As far as consumers are concerned, we’ll see if it makes sense for us in the long run. There is now a currency that consumers can use, and we call it Uber Credit. This ties in with our membership program. For example, for hotels, members receive 10% cash back on $1,000 transactions, with $100 of that returned as a credit that can then be used on rides and meals.
Will Uber ever offer its own buy now, pay later product?
We don’t know because we want to make sure the experts do what experts do. You can do that at checkout because we’ve already announced partnerships with other companies in the industry that already offer that service. When it comes to our general product strategy, we don’t try to be everything to everyone.
For European boat rentals, tapping a tab takes users to the partner’s own booking flow rather than checking out within Uber. Will that takeover model be the model for the future?
Sure, sometimes you need to rely on your partner, especially when you’re doing something new. That’s because two-way integration just takes time, and sometimes it’s good to try it out before integrating deeply. In Expedia’s case, we decided that deep integration made sense, so we partnered with Expedia to build the entire UI ourselves. However, in some cases it may make sense to hand over the remaining experience to experts in the field. And you can always integrate them deeply if you get significant traction.
Your Uber One membership product currently has 51 million members, accounting for approximately half of all bookings. Is there any data that shows cross-selling actually works, i.e., delivery users initiate more rides later on?
In terms of delivery, 2-3 orders can make up for the monthly fee you’re paying. As members become more familiar with the program, the frequency of the program increases in areas of work they are already using. And that’s leading to increased use of other aspects of business as well. Mobility-only people are starting to use delivery, and delivery-only people are starting to use mobility.
Shipping is one of the most difficult businesses to make a profit in the tech industry. Is Uber Eats still relying on ride-hailing to stay healthy?
In the early years of Uber Eats, we were not yet profitable, but over the past several quarters, Uber Eats has become a separate and profitable business for us and has generated significant profits.
In an article I wrote this spring, I argued that Uber was in unexpected direct competition with Airbnb, which currently offers airport transportation through its partners. Does it look like that? Who are you most interested in?
There is no shortage of competitors. Lyft in the US, Didi and 99 in Latin America, Bolt and Ola worldwide, and DoorDash and Deliver Hero for delivery. But I spend very little time thinking about it. The majority of my time, my late nights, is spent providing our users with all the value we can provide.
You recently concluded your Waymo pilot in Phoenix and expanded elsewhere. How can you maintain a consistent experience when working with the same suppliers and competing with them in some cities?
Phoenix is the first city we’re launching with Waymo, where we have about a dozen cars, but our larger launches are in Austin and Atlanta, where we have hundreds of cars. When we recently saw the Phoenix pilot, we mutually decided that it didn’t make sense to continue. Waymo is a great partner for us, but it’s also a competitor in many cities. We are not in the race to become an L4 autonomy provider. Our focus is on laying race tracks so that you can coordinate with multiple players. We believe in hybrid networks, where human drivers and self-driving cars exist in the same city. This is because it allows us to balance demand and supply.
When it comes to AV Labs, what can Uber offer autonomous partners that it doesn’t already have?
Through our fleet partners, we plan to equip hundreds of cars with sensors through which we will collect millions of miles worth of driving data. This is very useful for long tail problems. You need to look at all edge cases, not just the P95, P99 level. Beyond the data itself, there’s a lot of know-how from 10 million earners about how transportation works. We process 25 million lost items every year, but how do you respond operationally in a world of autonomy? That’s the operational expertise we can provide.
Is Uber selling driver and rider data to Gen AI companies?
Divide this into two parts. For Gen AI companies, we can label data using our revenue streams or through voice collection. Certainly, we have a commercial relationship with them and we sell it to them. This is a new piece of business and we are very bullish on this. AV Labs is independent and is still figuring out a model for sharing data with partners. It’s a little early.
Are drivers recording conversations with riders to create this data?
No, no, no, just to be clear, your conversations will not be recorded during your ride. When I’m not traveling, I’m not driving or delivering, I’m just talking or listening to audio and transcribing it. By the way, they get paid to do it.
Where has AI actually appeared in a way that riders and drivers have noticed?
If you are earning on our platform, we have an earner assistant. The number one question on their mind is: How can I make more money? And they’ll say, “Look, it’s actually pretty bright in the South Bay, but you might want to go eight miles away where there’s a lot of demand.” On the Eats side, there’s a grocery cart assistant that lets you say, “I want milk, eggs, and bread,” and it instantly creates a cart for you. Vehicles also allow you to request a ride using your voice. You can say, “I’m looking for a ride to the airport. I have 6 pieces of luggage and 6 people.”
So, does this mean that a completely agent-based Uber, a service where you can plan and book your entire trip, is just around the corner?
I can’t put a date on it, I can’t say exactly what the feature set will be, but I think AI will be a big enabler of that. You can tell your agent exactly what you want, leaving the complexity to the platform. Easier said than done — we want to make sure we’re not just checking a box by shipping an agent that might not perform very well.
As a CPO, with so many ideas on the table, how do you personally prioritize?
I would say that I spend 70% to 80% of my time making sure that our existing products, or the products that we’re launching, are as solid as possible. Every new idea is like a shiny object. If you have 100 ideas, 5 of them will be good. And those five require a lot of nurturing and belief. So maybe 20% of your time is spent on new ideas. By the way, this includes going out and delivering products in my car just to see the product directly from the other side.
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
