Uber CEO Dara Khosrowshahi

0:00You came into Uber. Uber was at$4.5 billion in losses. Now you've turned it around. Last year it was around$10 billion in free cash flow. Almost$10 billion, yeah. The fact that we are a complete platform allows us to grow faster than our competitors and be more profitable. We started as the Go app, just rides. Now Uber Eats is just as big as our rides business. Now we're going to travel. So it's pretty cool to announce this partnership with Expedia. We got almost 50 million members. So Uber One members, when you book hotels on Uber, you get 10 % back on all of your bookings. And on 10 ,000 hotels, you get at least 20 % off.

0:33So this is going to work. What's behind the door? Could be behind the door.

0:48Dara, welcome to Sorcery. Thank you very much. Excited to be here. Well, thank you for having me here. Where are we today? What's going on? So we are at the PAC, which is a performing arts center, downtown New York. And this is our Go Get Showcase, Go Get 26. It's our consumer-facing showcase where we show all of the cool stuff that we're building for you this year. I heard that King is in town just for this. We tried to get him over here, but it didn't work. Well, that's amazing. So today you announced a lot of really exciting things. One of the most exciting things we'll come back to at the end, but you had a full circle moment with Expedia.

1:28So what's going on? We sure did. So we announced actually now making hotels available on Uber. You know, we started as the Go app, just rides, obviously. And then we went with Get Uber Eats. And it was just an idea. Now Uber Eats is just as big as our rides business. Now we're going to travel. So go, get, travel, and we're partnering with Expedia, which is a company that I ran before I came to Uber for 12 years. So it's pretty cool to announce this partnership with Expedia. Amazing. Okay, so let's walk through and you can kind of showcase a couple of the things that are going on. Definitely. So this is one of the features that I'm the most excited about.

2:09It's called Shop for Me on Uber. And this was actually inspired by our users who were hacking into essentially our system. We have a service called Uber Courier. Have you ever tried it? Yeah. Yeah. So Courier is basically it's a driver or it can be riding a bicycle, et cetera, a scooter who will take an item for you from one point to the other. If you forgot your computer or home or something, they can bring it to work for you. Anything that you want. What we saw is people who are using Courier were hacking the system and they were like asking couriers to go to a store for them and buy something for them and i'll pay you i'll pay you back and when we saw this we're like wow we can actually productize this you know we're constantly looking at the searches etc what consumer behavior is and what we can productize so this is actually a feature that essentially our users were hacking into and it is shop for me on uber uh we send a courier out to a store and you know we've always wanted for you to be able to go anywhere get anything on uber but the truth is you can get anything that's available on uber uber eats right now literally you can ask a courier to shop for you go to any store um you can upload a picture of what you're looking for on the net what it's going to cost and the shopper goes there and we only get our best shoppers we match you with the best shoppers yeah and they will get approval like is this what you wanted?

3:34Here's what it's going to cost. You can make it all happen. Payment is taken care of and the experience is absolutely seamless. It's amazing. So we actually really needed that a couple of weeks ago. My videographer, Brett, sorry, Brett. I'm so sorry. We got to Hill and Valley Conference right before it started. He ripped his pants. Oh my God. Wow. So he just ordered pants on Uber and it came over. Yeah. I love it. I hope they fit. I think they did. Yeah, he was looking good, right? Yeah. My wife is going to be very happy about this. It may cost me a little something, but it's a very, very cool feature.

4:10Amazing. Okay, so let's keep on going. All right, let's keep walking. You know, over here, you want some coffee or are you properly caffeinated? So this is part of it, right? Like you can now get coffees in your Ubers? Yeah, so this is very, very cool. And what we try to do is create these surprise and delight moments for consumers. At the same time, another part of the strategy is we have users who use Uber rides and we have people who use Uber Eats and we're trying to create cross-platform interaction. Consumers who use both services usually spend three times more. They're much more loyal to the platform.

4:43So we want to introduce the different services in a delightful way. This one is pretty cool. So when you order an Uber reserve, we also have Uber Elite for the next level. Yeah. You can actually order a coffee to be in your Uber when you're going to work. It's one of those really cool interactions. And again, maybe you didn't use Uber for delivery, et cetera. It's a little way for us to upsell you the other services that we offer, but a pretty cool way to do it. That's amazing. Wow. Yeah. So hopefully you can try next time. I will. Hopefully you won't get coffee on your pants, though, because then, well, maybe it will be good for us, actually.

5:22Every time I wear a white T-shirt or anything, immediately. Coffee, matcha, it's just all over me. That's why I go with dark, you know, sweaters. They're indestructible. So this one is feedback that we got from our Uber One members. Should we sit in the chairs? Yeah, I mean, if you like to. I don't know if they work. You get the front. Let's see here. Do these chairs work? No, they don't work. Look at that. All right, how's this? What can I get you? You like coffee on your way to your Uber. So this is Uber. Oh, my God. Okay, you got some wings. Here you go. Thank you. I'm going to take those with you.

5:53Yeah. So this is for Uber One members. Are you an Uber One member? Of course. All right. Excellent. So are Uber One members one of the complaints that they had? And as you can imagine, people who use Uber tend to travel a lot. And because we're available in 72-plus countries now, they tend to be global travelers. Like, use your Uber when you land. One of the additional benefits that they were asking for is I want to get Uber One credits and benefits when I'm traveling, not just when I'm at home. So, of course, we listen to our most valuable users, our members, and this is going to be available for international travel as well.

6:33How much of the customer requests just end up getting front run to become new launches? Oh, a huge number. We're constantly watching what customers say. And the cool thing is now with AI, we can have AI agents essentially look at searches, et cetera. What are the most common requests? Customer service, same thing. Will we get calls? All of that can be summarized and can be put into what could be a potential roadmap that we look to build. And then do you test them internally if it's like actually commercially viable? You know, so we just test them. We put them out into the market. I mean, one of our philosophies is we just want to build stuff.

7:11and if you're not failing like this is i think a cool feature is going to be a win but we want to move faster we want to take risks some of these uh products are going to work most of them we hope but some of them aren't going to work and that's okay because you know one of the hallmarks of uber is we're builders we want to build fast most of our growth is going to come uh through building new great products versus going out and buying stuff uh and it's part of the building process that we got. What are some of the weirdest things that have popped up? Some of the weirdest things that have popped up for us are actually like all of the requests that we find.

7:49Even people saying, can you go and get some cash from me at the bank? You know, like the trust that people put with Uber money. Curies. Yeah. We're going to think about that. All right. So over here, let's see what we've got. Oh yeah. So this is travel mode, right? So this is what I was talking about in terms of travelers using Uber a lot. And I don't know about you, but room service at hotels has suffered over the past 10 years. This is true. You know, you never know if it's going to be available. You often have to wait for a long time. It's often really expensive. So we now know when you're in a hotel and you open up Uber Eats, kind of switch to travel mode, and we offer essentially room service, but it's from all the restaurants around the hotel, local places, et cetera, deliver to your hotel.

8:43And because we recognize you're in a hotel, we make all the adjustments, you know, meet in the lobby, or if we know the hotel can deliver to the door, deliver to the door, et cetera, so that we can make your in-room experience and room service kind of an Uber Eats room service, which is very, very cool. I was just visiting Vegas and the room service there is like$500 for a pot of coffee. It's insane. Yeah. And you don't have any money to gamble. There you go. You can use your breeds for that. Yeah. Amazing. Okay. Oh, here we go. Yeah. Sweet or salty. What do you think? I like salty. Yeah. Here you go.

9:21Thank you. I'm going to go for sweet. I want to see what sweet has to offer. There you go. Thank you very much. Really appreciate it. What do we got in here? I don't know. Some bonbons. What is a bonbons? Oh my god, it's little Swedish candies.

9:40This is salty?

9:44We can't have this because we're going to be chewing crazy. Yeah, you're right. I love this. Thank you. I love my gifts. Sorcery is brought to you by Brex, the financial stack trusted by more than 30 ,000 companies, including one in three venture-backed startups in the U.S. Nearly 40 % of startups bail because they run out of cash. Rex is literally built to help founders avoid that. Unlike traditional banks that let your money sit idle, chipping away at it with fees, Rex's designs help you spend smarter and move faster. Their all-in-one solution combines checking, treasury, and FDIC protection into one powerful account.

10:24You can send and receive money globally at lightning speeds, Get 20 times the standard FDIC coverage through their partner banks and even high yield from day one. With same day and even same hour liquidity, access your funds anytime. Companies like Scale AI, DoorDash, Service Titan, HIMSS, Anthropic, Flexport, Robinhood, and Plaid trust and use Brex. Start today at brex.com slash sorcery. That's B-R-E-X dot com slash sorcery. Turing is training the next generation of AI with tasks that require real expertise and real world judgment. That's why companies like NVIDIA, Anthropic, Salesforce, and Gemini partner with Turing.

11:08Turing builds realistic reinforcement learning environments and data systems based on real operational traces. The kind of infrastructure Frontier Labs need to train superintelligence. Visit Turing.com slash S-O-U-R-C-E-R-Y. All right. And then last and certainly not least, this is what we talked about. Yes. This is hotels on Uber that we're going to see. Huge. Huge. And what we're super excited about this one is for Uber One members, right? So first, all of the information that you already have with Uber, credit card, your identity, all of that stuff we already have. We're working with Expedia, my old company.

11:50How'd that happen? Yeah, it's amazing. I did seriously. I had to recuse myself from the process because I'm still on the Expedia board. Oh, okay. So from a governance standpoint, it's like, hey, Dara, you can't be. So then how do you set that up? I asked the team, basically. You know, travel has always been. You just send them over and push over a note on a table? No, no. We did it the right way. But we've always thought about travel. It's a huge use case on Uber, and it was natural for us to add hotels as another product. In Europe, for example, we already have trains available in the UK and France and Spain.

12:24So originally it's like, well, it's wire up everything that moves. But then how do we actually build around the experience of movement and traveling as well? Hotels was a natural for us. The focus for me has been deepening the relationship with Uber One members. You know, we got almost 50 million members. It's a unique offering, which is you get discounts on rides, you get discounts on delivery as well. What else can we offer these members to keep our membership growth really, really high and keep retention really high as well? And savings on hotels was something that came up. And so Uber One members, when you book hotels on Uber, you get 10 % back on all of your bookings, regardless of which hotels you book on.

13:06And on 10 ,000 hotels, kind of a rolling list, you get at least 20 % off. So there's an incentive to book on Uber and we think it's going to be a delightful experience. I think you mentioned there's 700 ,000 available hotels. And we're going to add Verbos. So both for hotels and vacation rentals as well. You know, if you're vacationing with a family, you need that Uber XXL. We'll get you to Uber as well. So when you book an Uber, do you book Uber Black SUV? What do you book? I usually book Uber black and then SUV if I'm going to the airport. What else? What are your tips? So I don't know if it's a tip.

13:48Okay. I'm a little spoiled. I order Uber Reserve all the time. So even if it's like a date night with my wife, instead of just like booking the Uber and you're not sure, etc., I order Uber Reserve. And, you know, we know exactly when we have to go out there. The car is waiting. It's a pretty awesome luxury to have. When are we going to have Uber planes? So we are having Joby. We have a relationship with Joby Aviation. I don't know if you saw them. They were flying these. I saw that. It was so cool. Like 10 minutes to JFK. It's unbelievable. So we actually have a relationship with Joby. And we're hoping now to get Joby on Uber as well once those flights actually start carrying passengers.

14:29So it's, you know, any way you want to go someplace, save some time, maybe save a little bit of money when you're traveling, we want to be there for you. Yeah. I was talking to Brooke about this on the way over. Airplanes don't have great Wi-Fi. Maybe Uber Wi-Fi. It's something we're going to have to work. Just an idea. Free Wi-Fi if you're an Uber One member. I think it can happen. Oh, I think we got to test this out. All right. You go for it. You want me to? Well, where do you want to go? You want to go to Paris? Tokyo? I would love to go to Tokyo. Tokyo, right. Hearing international taste. All right.

15:01So this is Tokyo. You know, this is. So this is actually the search. Let's see. It's doing it for us. It's doing the demo. This is the AI. I'm used to doing this myself in the product demos. So you can put in the dates as well. You know, any kids. And then you get a bunch of choices. And you earn$223 in your credits. Choose your room. And, you know, see all the details. So it's all of the stuff that you're used to. and it's incredibly easy. We already have all your payment details, and it's just like, you know, ordering Uber. It's as easy as that. We can go to the room for here. And a lot of little to the room.

15:47All right, let's see. Oh. You want to pick it? Go check it out. So this is going to work. What's behind the door? What's behind the door? Could be behind the door. Check it out. Let's see. Oh my God, it's Tokyo. Bye, guys. I'm just kidding. That's really cool. It's almost like you're there. I know. It's like a next transportation portal. Yeah, there we go. Uber Teleport. Maybe one day we'll get that going. That's hilarious. I'll put this right back. There you go. So what were some of the biggest lessons? You spent nearly a decade at Expedia. What were the biggest lessons from there? So for me, it's supply, supply, supply.

16:31Okay. So at Expedia, obviously we're a consumer company, but what's most important is the inventory of hotels that you've got, the pricing of the hotels that you've got available. And that's certainly something that we see at Uber. I mean, again, you think that you want to add consumers, you want to add audience to Uber, but what it's about is getting as many cars as possible, signing up as many grocers and restaurants as possible. And when you build out the best supply, then consumers show up because we make it easy and we make it worth your time and worth your dollar. I think you're pretty humble.

17:05You came into Uber. Uber was at$4.5 billion in losses. Now you've turned it around. Last year it was around$10 billion in free cash flow. Almost$10 billion, yeah. So did you have to rewrite your brain? How did you come in and do this amazing turnaround? What were the key? Well, thank you. I appreciate it. But, you know, one thing I would tell you is that a lot of people talk about Uber. You know, when I came there, obviously there was a lot of controversy in terms of the culture, etc. But the team had built a good business in terms of demand and the brand and our presence globally. So while things were not easy, I kind of was standing on the shoulders of giants.

17:45And this was an incredible brand. And for me, I got to bring in some of the new team and then kind of continue some of the old team together and build on the foundations that we already had. The fact is that Uber was incredibly popular. We really focused on supply. We focused on safety as well, because both our drivers and our riders want to feel safe. And then once we got to Uber Eats and started building that cross-platform relationship, you know, we compete with lots of people. but most of our competitors are monolined. They either have rides or they have eats. The fact that we are a complete platform allows us to grow faster than our competitors and be more profitable than our competitors as well.

18:27Do you think eats will surpass rides at some point? God, you're asking me to like choose between my favorite children. I will say I order about three to four salads a day and maybe two rides. So it's up for you. It looks like it's. I mean, the two are essentially the same size. I would say that if Eats was just about food, I'd say rides would be bigger, but actually Eats now, it's grocery and we're wiring up like every other local business. I wouldn't be surprised if Eats is bigger than rides 10 years from now. Yeah. So what did you learn from Barry Diller? I learned so much from Barry Diller, but you know, one of the things that I learned from him is that I'm comfortable going against the grain.

19:10You know, Barry was the ultimate counter puncher he kind of took on ab you know abc nbc cbs uh you didn't grow up with them but a lot of people grew up with them and he he built fox which was a new uh a new product i've always been comfortable going as against the grain and it's one of the things that encouraged me to join uber from the outside things looked really really difficult uh but we got in there And at first it was, you know, climbing uphill. But the desire to go against the grain was something that I learned from him. I listened to a couple interviews that you did and seeking truth and getting to answer seems to be one of the leaders in how you get to products faster, innovations faster.

19:53Where did that come from? That came from Barry as well. So he actually met me when I was an analyst at an investment bank. He like, he wanted to see the person building the model. And, you know, what I find is that as you move up in organizations, actually, while you get a broader view of the company, a lot of times you actually don't understand what's really going on in the company. Right. And so you do have to be that truth seeker. And it requires you, like, not to manage a way that you might be taught. You know, I go and I sit down with engineers and product members, two, three, four levels down.

20:30We jam, no decks, really talk about what's going on. And I'm very, very honest with them in terms of what our challenges are. Like, I don't talk corporate speak to them. Everyone has a pretty good BS meter wherever you are. And if a senior member of the team is BSing his team, then they're going to BS you right back. So part of it is not depending on the hierarchy to get you their information, going and getting their information directly. And the second is to be a truth-sayer and then be a truth-seeker as well.

21:26You type in an idea like AI-powered supply chain companies with positive free cash flow, or defense tech companies growing revenue over 25 % year over year. Publix AI then dispatches a swarm of agents that scan every single US stock, evaluates them and instantly builds a custom index around your thesis. What really stands out is how clearly it explains why each stock is included. And before you invest, you can even backtest your idea against the S &P 500 so you're making decisions with real context, not just guessing. And beyond generated assets, Public lets you invest in stocks, bonds, options, crypto, all in one place.

22:02They'll even give you an uncapped 1 % match when you transfer your investments over from another platform. If you want to build a portfolio that actually reflects your thesis, visit public.com slash sorcery. Paid for by Public Investing. Full disclosures in the description. Enterprise AI runs on Merge, the AI infra platform for integrations, agent tooling, and model orchestration. so your teams ship product, not plumbing. Mistral, Dropbox, and Drada already trust Merge and production. Start building at merge.dev. Founders scale faster on Deal. Set up payroll for any country in minutes, hire anyone anywhere, get visas handled fast, and get back to building.

22:41Visit deal.com slash sorcery. That's D-E-E-L dot com slash sorcery. You mentioned how AI is helping inform product decisions. How else is AI impacting the business and operations, whether it is teams, communications, or the next era of Uber? So it's like changing how we build in every single way. One of the cool things about Uber is that we've always been comfortable with a probabilistic world. You know, most companies are kind of they want certainty, et cetera. But while you can get certainty in your digital life, you know, with an app, you build the interactions, et cetera, in real life, things can go wrong.

23:26You know, you can order that Uber and we told you it's a four minute ETA. The driver cancels. We have to actually get you another driver, et cetera. So there's all kinds of uncertainty that happens in the real world. And we have been built on top of that. So everything that we built has been algorithmic in nature. your pricing, your matching, your East feed, et cetera. Now, it went from very simple algorithms to deep learning algorithms, and now, obviously, to much larger foundation models. So for persons, like, culturally, companies are often uncomfortable with uncertainty. We live in uncertainty, because the real world can punch you in the face sometimes, right?

24:13All kinds of things can go wrong. And so approaching with this probabilistic mindset of what outcomes can be is really important in terms of driving our culture. We've had a leg up because we've been building algorithmically, but our developers are all using Claude and actually Codex is pretty cool as well. And in developing, we're seeing the number of diffs per developer go up. We've seen the number of lines per diff go up as well. So there's a lot of input going in and there's a lot of productivity coming out from our engineers. All the tools, if you're doing a code review, et cetera, we have agents doing code reviews, but we also want to make sure that a person looks at the code reviews.

24:57A lot of the debugging that goes on, understanding where all the services are. People used to kind of go hunt and click within the system. And as you can imagine, our system is very, very complex. Now agents can do that for you. So the way that we build has completely changed. We're working to change our customer service as well to have agents, digital agents first help our human agents. And then our human agents start to train AI models as well. And it's interesting there, you know, I think a lot of the approaches to building AIs have been kind of a surface approach. Let's optimize 10 or 20 % of the process.

25:36We're trying to go ground up at first principles. So originally, when we were trying to train our AI agents, we were like training them on the policies that we have. If you're an Uber One member and your delivery is 10 minutes late, what do we do over the ETA, et cetera? Those policies are in place because they're the best guess for us as to how to treat you as a customer, how to get to the right outcomes. Now with our agents, we can train them on, what's my goal? My goal is to make Molly, who's an Uber One member, a happy customer. And so the agents themselves can now reason around the outcome that you're going after.