All right, we're going to go ahead and get started. I'm Doug Anmuth, JPMorgan's Internet Analyst. We're very pleased to have with us Uber's CEO, Dara Khosrowshahi. Uber's mission is to create opportunity through movement. Uber is a global leader in two secular growth industries: ridesharing and food delivery. Last year, the company generated $163 billion in gross bookings, along with nearly $7 billion of free cash flow. Dara has been CEO of Uber since 2017. He was previously CEO of Expedia and, prior to that, CFO of IAC Travels. So welcome, Dara.
Thank you for having me. Great to be here.
Absolutely. All right, let's start high level. It's been more than a year since Uber provided its midterm targets at the 2024 Investor Day. What's changed most in the business since then, and how do you feel the company is progressing toward those goals?
I think the good news is that, you know, anytime you give multi-year projections, you know, there's some danger of things changing or things going off course one way or the other. We're right on track. Just as a reminder for folks out there, we talked about mid to high teens growth as it relates to gross bookings. Obviously, margins leveraging quite nicely in terms of EBITDA growing mid-30%- 40%. Substantial free cash flow generation as a percentage of EBITDA, you know, more than 90% of EBITDA we thought would be moving into free cash flow. We are on track or ahead of essentially all of those targets. If you look at our business, we think that the growth levers continue to be huge. We're relatively underpenetrated.
You know, typically in most markets, less than 5% actually of the adult population in markets use us on a monthly basis. The growth formula that we see is increasing audience, increasing frequency, and then hopefully price being neutral in terms of gross bookings growth. We do not want to pass on price unless we have to, or we are just passing on costs to the consumer. If you look at last quarter, for example, audience grew 14%, very healthy and consistent with last quarter. Frequency grew 3%. We just keep improving the quality of service, so to speak. Higher penetration of membership, which drives frequency structurally. On the pricing side, there has been some inflation in, for example, mobility pricing over the past couple of years in the U.S.
We're trying to keep that inflation in check as much as possible, something that you saw in the last quarter. A couple of priorities for us that we're working on right now. Number one is driving affordability. Again, you saw last quarter in terms of mobility pricing essentially being flat on a year-on-year basis. Our membership program, for example, is a very significant affordability lever. We're giving our members a discount and trading that for much, much higher frequency and better retention. Membership, we now have over 30 million members, and delivery membership penetration is at 60%. Best of breed for us today is 70%. We think we've got a good amount of penetration left as it relates to membership. On our Eats business, for example, we're driving a higher and higher percentage of merchant-funded offers. These are offers that essentially are funded by our merchants.
Merchants, you know, might be buy one, get one free. It might be spend $15, get another $10, et cetera. These offers essentially drive merchant business up. Because merchants get to pay in their goods, it's, you know, a really nice benefit for the consumer and limited cost to the merchant. That is one big angle that we're going after. The second big angle that we're going after is increasing our penetration in less dense markets. This is a really important initiative for the company that started with food, going to the suburbs, going to secondary and tertiary markets, and actually now is extended into mobility as well. For example, now 20% of mobility GBs are coming from these less dense markets. The growth rate in these less dense markets is significantly higher than the growth rate in the core.
We think the potential in these less dense markets is very, very significant. The third big growth driver that we're pushing is continuing to scale our growth bets. These are bets like Reserve, Two Wheelers and Three Wheelers, Uber for Business, Hailables, taxis generally. On the food side, it's our direct business, which is Delivery as a Service or grocery and retail. All of them are scaling in terms of size. The margin for those growth bets is lower than the base margin that we have for, let's say, our base business, online food delivery business, X business. The margins continue to increase in that business. Of course, autonomous, which is, we think, a huge time extender, safer way eventually of transportation. Still have some work to do in terms of costs and scaling that business.
We think that the promise that we're seeing and the alliances and relationships that we're building, they are absolutely best of breed. I'd say so far, so good. I think in this case, not much changing is good news for us because these were pretty ambitious targets. We think that in any environment, strong economy, medium economy, weak economy, tariffs, no tariffs, I think we're a pretty safe bet.
Let's hit on that point just on macro a little bit. You talked about on the one Q call that Uber was recession resistant. How do you think about the degree of resilience of the business? And do you manage any differently in a potentially tougher environment?
Yeah, I mean, you have to manage different in a potentially tougher environment. I think there are multiple angles as it relates to the resiliency that you see in our business. First, I would say that the categories that we're in, food, transportation, grocery, historically, if you look at historical cycles, the spend levels in those categories during down cycles tends to be much less variable than other categories. You know, if you're not doing well or you're going through a recession, you may put off the European trip, you may put off the home improvement, but you're still going to treat yourself to that nice, you know, Friday night, take the folks out for dinner or order, you know, a nice Uber Eats dinner. One is like either we got lucky or good, you can debate, but we're in a bunch of really good categories.
Second for us as it relates to geographic is we have incredible geographic diversity. We operate in 70 countries. Over 50% of our bookings come outside of the U.S. We're not really subject to tariffs because we're the ultimate local business. You know, the money that comes from a Boston consumer goes to a Boston restaurant, or the vast majority goes to a Boston driver. We're not really subject to all the tariff talk going on. Obviously, we hope for the best. That's kind of another differentiator for us. Third for us is that our business model and our expenses are about 75% variable. This is a very, very low fixed cost business. I think coming out of COVID, for example, which was the biggest shock to the system, I think a quarter after the kind of peak of COVID, our mobility business was profitable.
I think that's something that no one would have predicted. Last, and certainly not least, is that our revenue and our costs tend to be correlated in economic cycles. You can imagine in a weak economy, to the extent that consumer spend gets weaker, employment gets weaker as well, which would mean that the cost of our sourcing drivers, couriers would generally come down. In stronger markets, both of them kind of shift up and down. In weaker markets, they also shift up and down. You look at these categories, the categories that we operate in, our geographic diversification not being subject to tariffs, our variable cost structure, and the fact that our revenue and costs are correlated, makes for a pretty good combination.
I think the team at Uber has, you know, gone through a lot. People said we could never be profitable. I think we've disproven that. And then some, we've gone through COVID, navigated really, really well. I think when and if there's a significant economic cycle, I'm confident that the team can execute around that.
Okay, great. Let's shift gears, talk about delivery. The top line has remained very strong over the past several quarters. You've continued to gain category position in most markets. How should we think about the delivery growth algorithm, and how does it differ across core and then with grocery and retail?
Yeah, absolutely. We've been very, very happy with our performance in delivery. We're actually seeing order transaction growth rates accelerating. The last quarter transaction growth rates were about 15%, which is very, very strong. Our frequency continues to be at or near all-time highs, obviously assisted by the membership program as well. If you look at the top line, we've been growing our top line in terms of gross bookings, 17%-18% pretty consistently over the past seven quarters. This has been a very, very consistently growing business. We believe that we have grown our category position in 10 out of our top 10 markets. We're now category position one in eight of our top 10 markets as well. All of these metrics, both competitively and top line, are working out well.
I think we've been able to demonstrate, obviously, very, very strong leverage in terms of profitability growth. This comes from just better densification and more efficiency as it relates to the cost of our logistics network. It relates to a relatively fixed cost base. The business is growing in the top line, but the fixed cost base of the business is quite fixed. You add on top of that our advertising business, which is growing at very significant rates, well over $1 billion in terms of revenue, growing over 50%. This is a very, very high margin business. We're confident about the top line. Structurally, we're set up really, really well in terms of the bottom line. As it relates to the growth formula, honestly, the growth formula is pretty simple.
Number one is we want to keep improving the core service itself, which means average time of delivery, error rates, selection, the speed of search, the quality of search as we build out larger models. All of these metrics are moving the right way. You can't actually point to a causation between, let's say, less error rates and higher frequency, but there's definitely a correlation. We know that it's the right thing to do, and the quality of the service continues to improve. That's number one. Number two is affordability and price. The big levers there are membership programs, 60% of gross bookings coming from members that get locked into our ecosystem, so to speak. Obviously, they want to be locked into the ecosystem because they're getting a great deal. Then merchant-funded offers. These are the offers that I talked about.
A higher and higher percentage of our gross bookings is coming from merchant-funded offers. Generally, foot traffic to restaurants has looked like it's weakened a little bit. There's no weakness at all seen in terms of delivery. Merchants are leaning into the channel that is working out really well. The third big push for us is grocery and retail. The grocery and retail category is larger than the online food delivery category. This is about we are increasing selection there as far as the number of merchants that we have. We introduced Dollar General, Home Depot. There are some really exciting announcements coming up. As we improve the selection for grocery and retail, we're seeing a higher, higher percentage of online food delivery consumers trying out grocery and retail.
It was about 18% of our audience on Eats tried out grocery and retail. It's up to 30% in certain markets. We think we have a good kind of running room there. One interesting factor that we're seeing with grocery and retail is it's actually, you know, when we first got into grocery and retail, I kind of thought about it as an upsell channel. You know, let's take people who are buying food. Let's get them, you know, shopping convenience, and then we'll get them kind of doing their weekly shop and driving basket sizes higher. In the retail category, actually now there are retail occasions that we can really merchandise around that are attracting either big-time audience of our established audience to the site or attracting new audience. For example, you know, Valentine's Day is an example.
Mother's Day was our best week ever as it related to grocery and retail. That business is getting to be a higher percentage of our overall. It's accelerated for two quarters running. It is variable contribution, profitable, not EBITDA profitable yet. I know it'll be too soon to drive EBITDA profitability. We see a small advertising business there that is now ready to scale with our Instacart partnership in the U.S. in particular. You know, a ton of promise there very, very early as it relates to the growth. We're quite pleased. We think the growth runway for our Eats business is quite significant.
Okay, great. Let's talk mobility. You've consistently grown the business at a 20% rate and expanded margins. There is some noise in the first half, just as pricing is lower than expected. Can you talk about the drivers here and then how you expect that to play out through 2025?
Yeah, we think actually, you know, pricing being lower, it's, I know it's a little bit strange as it relates to talking to companies philosophically, but I think one of the things that separates technology companies from maybe some other companies is that we actively try to drive pricing lower. It's a really important philosophy for us at Uber because it just drives, you know, kind of increases our overall total addressable market. The biggest factor as it related to pricing on mobility that has is starting to shift is the cost of insurance. As you know, insurance costs, the commercial insurance costs across many, many industries, but especially the transportation industry has increased significantly over the past couple of years, double digits. Same thing has happened to us, and we passed on those costs to our consumers. Unfortunately, we've had to.
We're seeing those insurance headwinds start to ease. One is the market general is starting to ease. CPI, market CPI for insurance for the first time in a while increased single digits as opposed to double digits. We are taking some steps to improve our own insurance results. One is we continue to kind of choose safer paths, avoiding left turns, avoiding areas where there's a high probability of some accident. These are kind of product improvements that we're making. Another product improvement that we're quite excited about is that safer drivers. We're now scoring drivers based on their driving behavior. These are kind of third-party safety scores. What we see is as we score drivers, they start driving better. You know, it's very, very specific as to the amount of harsh braking, the amount of harsh turnings, et cetera, speed up, slow down.
We are now attaching the safety score and generally driver quality to driver earnings. Higher quality, safer drivers will get paid more per match, so to speak, and will get more matches than the opposite. That is also driving driver behavior towards safety. Last but not least is the regulatory environment. States are very, very aware of insurance costs skyrocketing. We have seen some very encouraging regulatory trends in a number of states, Georgia, for example, a couple of others coming. We hope for those regulatory trends to continue. All of that has translated into our internal insurance costs and the headwinds that we had easing. We have essentially passed on that easing to consumers. If you look at our mobility growth, transaction growth was very similar to what it was in the past. Audience growth, frequency growth have been quite consistent.
We just were able to ease the pricing growth that we've had in the past. I think what you should expect going forward, at least the next quarter, is transaction growth to be broadly similar. Really the question in gross bookings growth for mobility is, you know, do we need to pass on pricing or do we not need to pass on pricing? You know, I'm hoping for the latter.
Okay, great. Let's talk about autonomous. AV, the biggest discussion and really debate when it comes to the Uber story. We're also now about a year into an elevated level of focus and rollouts. What do you think has changed most in the AV narrative over the past 12 months?
I think what's changed is that a bunch of speculation is starting to be kind of shaped by reality and the execution of the business. I think you've seen us now be very, very active as it relates to our partnerships with autonomous providers. We've always said that as the largest kind of mobility platform in the world, one of the largest delivery platforms in the world, and then one of the largest logistics freight platforms in the world, those are the three biggest use cases as it relates to autonomy. Every player who is developing autonomous technology, we think should be working with us across use cases, so to speak. We really have a view into autonomous and technology and the development of the technology that no one else in the industry structurally can have.
What we're seeing is really encouraging, which is multiple players are developing this technology. Whereas the technology has taken much longer, I think that most people expect it to develop, the software capabilities are absolutely getting there. Waymo is absolutely leading the pack. They are the, you know, first in. They have an incredible bar for safety and quality. We're very, very happy with our launch in Austin with Waymo. This was, we had ambitious targets as it relates to our Austin launch. The early metrics that we see are better than those ambitious targets, especially as it relates to the uptime of the vehicles and making sure that our fleet management services are as good as they need to be.
Thank you to the Waymo team because they kind of taught us some of their best practices that they've learned in the cities in which they operate. Of course, the utilization of the vehicles and the customer love for the Waymo product absolutely is there. Customers love it. It's safe. The utilization of the vehicles is very high. We are seeing that customers are opting in for AV at a high rate and the second time at an even higher rate once they've taken an AV trip. This is a really, really good product out there. Really what we're looking to do is continue to expand our scope with alliances with all of the players in the industry who pass our safety bar, so to speak.
You've seen a bunch of announcements, announcement with Pony.ai, WeRide, Momenta, Volkswagen, May Mobility, and hopefully more announcements to come and more launches in markets, both in the U.S. and outside of the U.S. The trend that we see in autonomous is very strong. We're a long way away from commercializing. We're a long way away from scaling, but the early signal that we see is quite encouraging.
Okay. Just as the number of vehicles in Austin ramps, and you've talked about the fleet going from nearly 100 today to hundreds over the coming months, how do you manage the mix of AVs and human drivers, especially at some of the lower demand times like early morning?
Yeah, definitely. One is that the number of vehicles in Austin is going to increase as the operational domain for kind of safety for the Waymos in Austin continues to increase. By the way, we're also launching Atlanta, which is something that we're really excited about. We have a, you know, kind of the way that we manage demand to various aspects of our marketplace is something that's controlled by this dynamic dispatch layer that we've got. For example, most human drivers like longer trips than shorter trips. You know, all the getting in and out is annoying, the waiting time, et cetera. Our Waymo vehicles are tending to do more shorter trips than longer trips inside of that operating domain.
We think as it relates to the number of vehicles that we have and the number of vehicles that we are scaling up, we'll be able to manage kind of human demand and/or human supply and Waymo supply in a constructive way as we go forward. I do think that during some, you know, kind of non-peak times, call it 2:00 A.M., 3:00 A.M., 4:00 A.M. in the morning, the number of human drivers that we will need are going to be significantly decreased. We can essentially message that to our drivers. We can shape our incentives in a way to essentially shape supply, human supply, because we only need to pay drivers when we use those drivers, when they're utilized. I think there's some supply shaping that we will undertake, but we're very, very experienced in terms of supply shaping.
You know, surge is essentially supply shaping. It's getting drivers out during the times that we want, in the places that we want. This is something that we've done for a very, very long time, and we're quite confident of the path forward, so to speak.
Okay. Sticking with autonomous, you provide marketplace as well as fleet operations in some markets with Waymo. How do we think about the current structure of the business model, how that could evolve, and then the degree to which you're willing to use your balance sheet to invest in the AV ecosystem?
I stress that we're very, very early as it relates to the ecosystem. Right now, we want to prove that this product works. We want to make sure that customers love the product. Above all, we want to make sure that it's safe. Ultimately, as it relates to the business model, we think that there are going to be very, very layers as it relates to the business model. There'll be the demand layer, and we think we're going to be the biggest demand pipe, so to speak. There will also be like certain players who have a direct channel. Waymo is going to have a direct channel to demand as well. You're going to have a software layer. This is the driver. Again, Waymo, WeRide, et cetera, might be the driver. You're seeing kind of this software layer develop.
As an example would be the Waymo relationship with Toyota. You know, they're kind of licensing their software to Toyota, which is one of the leading players out there. Then there's an operator. This is a fleet operator who is housing the cars, recharging the cars, cleaning cars, et cetera. We have very, very strong relationships with many fleet operators around the world. Fleets already represent about 15% of our overall supply hours, so to speak, on a global basis. This is something that we have scaled, and we are now, you know, demonstrating our fleet operations capability in Austin, and we will in many other cities, usually through partnership. There will be kind of the owners of the vehicles themselves, and then there'll be vehicle financing.
In the early parts of this developing the business model, we or fleets will, you know, invest in these vehicles to kind of prove out the economics and, again, prove out the quality and the safety. Over a long period of time, we anticipate that these vehicles are going to be owned by financial players. You know, just like Marriott doesn't own their own hotels, and there's, you know, a whole financial ecosystem that owns these hotels and finances these hotels. We think a similar financial ecosystem is going to develop on the fleet side.
Okay. Related to the AV investments, earlier this week, we saw you monetize your stake in Aurora. Can you just talk about the rationale there?
Yeah, absolutely. We've been clear, first of all, that we, over the long term, we're going to look to generally monetize many of the equity stakes that we have on our books. Aurora was one of the largest equity investments that we had on our books. We thought it was opportunistic to monetize the part of our Aurora stake through what's a, I think, pretty creative kind of financial mechanism. To be clear, we are huge believers in Aurora. We're huge believers in Chris Urmson. I would expect us, you know, we still have a very substantial equity stake in Aurora. We got to monetize some of it at what is going to be premium over the market price of Aurora. That was attractive to us.
I anticipate being a holder of Aurora stock for some time to come for the foreseeable future because we really believe in what Chris and team are building. At the same time, you know, this was an opportunity for us to essentially kind of take some of the capital that we already have in our equity stakes and use it as funding for some of the investments that we may make in AV. If you think about the investments that we're making in the AV stake, there are kind of three categories of investment. One is we're investing in some of the software providers to give them support. These are strategic stakes. These are good alliances that we're building, and these are very capable companies. A WeRide or Pony or Aurora is an example. Waabi is another example.
We will have a portfolio of equity stakes and a number of these parties. Second, as I talked about, is investment in the AV ecosystem. This may be investments in fleets who operate or own cars, or it may be investments in some cars as we establish a financial model that we talked about. The third is operating costs of, you know, running Waymo in Austin or running a WeRide in Abu Dhabi. In the early days, we're not going to be making money on a variable basis on these trips. It's a de minimis number of trips, so it's not a big number overall as it relates to P&L, but that's another area that we're going to be investing. We thought it was an opportune time. Again, Aurora is going to be a very big investment for us for some time to come.
This is just part of, you know, some, I think, constructive financial structuring that Prashanth and team undertook that we're very happy about.
Okay. I want to talk about insurance for a minute. You've been advocating for insurance reforms in the U.S., especially in California. How should we think about the possibility of insurance rates coming down in some of those states? Are you seeing anything early on just as it relates to auto tariffs?
Yeah. As it relates to insurance, we think that the regulatory environment is finally moving in the right direction. You know, inflation is something that consumers rightly hate. Insurance has been a big part, has been one of the most inflationary parts of what consumers and businesses have had to pay. Businesses typically pass it on, as we have. We are seeing some encouraging regulatory trends in Florida, Georgia, and Nevada. California and New Jersey are quite problematic. About 30%+ of the average Uber fare is eaten up by insurance, much of which comes from kind of fraudulent legal practices, abusive legal practices, staged accidents, et cetera. I do not want to throw out a number, but it is significant.
You know, we think that we have a very compelling case to make to the legislatures that this is just a cost that consumers are bearing and they shouldn't bear, and that if we have regulatory reform, we can bring prices down, which means more business for our driver base and lower costs for consumers. I think that's something that's very, very easy to get behind.
Okay. Great. We're going to wrap up with a quick word association. Just whatever comes to mind. Affordability.
Hugely important.
Macro.
Don't care.
Waymo.
Great partner.
Uber One.
Lots of growth ahead.
Advertising.
Yummy.
Margins.
Higher.
Tesla.
Let's see.
Insurance.
Hopefully coming lower.
Grocery.
Very, very exciting and early.
Free cash flow.
Give me some.
All right. Leave it there. Thank you, Dara.
Thank you.