餐饮行业的经济学与趋势：DoorDash 的 Tony Xu

Oh, wow.

We got Dot coming.

It's like a TARDIS.

Tony Xu is the co-founder and CEO of on-demand delivery giant, DoorDash. He's grown the business from a Stanford side project into America's dominant food delivery platform, and is now pushing into everything from grocery to retail to autonomous vehicles and financial products for gig workers.

Great.

Great.

Cheers.

Cheers.

Good to see you.

Good to see you too, John.

As I think back to that era where you guys got started, what year was DoorDash founded?

2013.

Okay. As I think back to 2013 and those early days, the iPhone and the apps that were made possible by the iPhone were the defining tech trend of the era. Uber and Lyft in ridesharing, Instacart, you guys—all of the magic-wand apps that made your iPhone useful for bringing things to you and manipulating the wider world.

1. Why DoorDash Won

Loads of people were going after food delivery, and way more than are around now. As you look back on that journey—from, again, it was really kicked off by the iPhone, in my opinion. Not that there wasn't food delivery before the iPhone, but that really massively increased the market size—as you think about from then to today, why did you guys win?

The short answer is we got more customers than other people.

Sure, but let's have the five whys here.

I think these are very tough things to study because many things are happening at the same moment, but I'll just talk about it from the perspective of things that I think we got right. When you think about something like restaurant delivery, you actually get judged on multiple dimensions as a service. We get judged on what restaurants we bring you, certainly, whether it showed up on time and the quality and condition you expect, how much did it cost, and if we screwed up, what did we do about it?

It's not one thing that you have to be good at, actually. It's all of the above. Unfortunately or fortunately, this is literally the game that we're playing, where customers are judging us in all of these dimensions. I think getting that right better than anyone else, as measured by whether or not people are coming back to the app and using us—even when we had no money to market to them or discount offers to them—was the tell.

Are you saying it was a very complex, multivariate challenge even from early on, and you guys just embraced that complexity and said, “Okay, this is going to be super complex to get right,” whereas maybe others took a bit more simplicity?

I think if I were to become overly reductionist and purely looked at it—

That's what I like to do.

From the product perspective, yes, because at the end of the day, any consumer product is judged very simply by its retention and its usage. That's how you know whether you have a differentiated product. I think it's very easy to have differences in opinion about whose app you like more or whether or not certain apps look similar or different.

At the end of the day, though, if our app is performing at a much higher retention and frequency of use than others, that's how we know whether or not the things that we say actually are making a difference to customers. Getting all of that right very, very early and then building the systems to actually instrument that, as well as to repeat that over and again, I think was very, very important in the development of the company.

Do you think you guys were more focused on retention than others?

I don't know if we were more focused. I could tell you, though, one of the things that was happening, especially when you see a competitive fight, is you see everybody race towards it. Everybody is going to try to make offers to customers, try to give discounts, try to give coupons, free this, free that. One of the things that we had, looking backwards, is we actually did not have a large budget. In fact, between 2016, 2017, and 2018, we barely were able to raise a dollar relative to our peers. As a result of that, that made it a constraint.

One of the constraints is, okay, you can grow, but you cannot spend in order to do it. In order to do that, you effectively have to come up with ideas in the product to stand out and make a difference and have organic growth carry you. Then, once we were able to demonstrate to ourselves first that we had a product with higher retention than other people and higher frequency, and then we were able to raise capital, we made the decision to pedal to the metal and actually go and acquire customers because we had an unfair advantage.

Were you more customer-obsessed than others?

2. Customer Obsession Under Pressure

I think it's very difficult to measure us against other people because I didn't really work at those companies. But again, I really just look at the early DoorDash crew and what were the things that we did that were very useful. We all did deliveries. We all did customer support. We all made menus. We all sold restaurants. We all customer-serviced restaurants. We all worked inside restaurants.

I think at the end of the day, the term customer obsession can be fairly generic and vague and applied. It really just goes to show what actions actually demonstrate that. For us, some of the ones that came to mind—and frankly, the story that actually inspired the value in the first place—was when our backs were against the wall, and we had, I think, less than 2 weeks of cash runway, and we had a terrible night where every single order was late. This was a Stanford football game in 2013, and I had trouble raising the seed round. We made the decision to refund everyone. That cost us over 40% of the bank account. When you have 2 weeks of runway, 40% of the bank—

You issued a refund that was 40% of your funds' runway?

Yeah, maybe a little over that. We baked everyone cookies and delivered them at 5:00 AM before everybody woke up. Look, this was maybe 100-ish customers or something like this. But I think it demonstrates that this was a real value. It's actually my test, really, which is: what are the actions that are naturally occurring? What are the behaviors that are actually occurring inside of an organization? Then how do you actually articulate it versus the inverse, which is you start with maybe some articulation, and maybe some of the times it matches and other times it doesn't.

I've always believed that these cultural norms or behaviors are really 80% of what you've done. I think that's the best way to actually find what they are for your company. We're the only platform today to take care of Dashers when there are these increasing gas prices and making sure that we have their back and that we're actually trying to help them save $1.40 to $1.90 per gallon for Dashers. Or in COVID, we were the only company to cut commissions by 50% when we were not yet profitable as a company. That was an expense of over $100 million.

We then, on top of that, spent millions of dollars buying national TV campaigns to tell people to order, whether it's on us or our competitors, because restaurants have, on average, 17 days of cash on hand. You need to order. This is do or die for them. I can give you many of those examples where I think, at the end of the day, whether the actions first speak, and then you can articulate the actual words and norms and cultural behaviors that the company is trying to speak. That's ultimately, I think, how I think about customer obsession.

If I try to draw what these values are, it seems to be a very clear-eyed focus on the best product experience and taking care of Dashers. I didn't realize you guys had a time when you had only 2 weeks of cash in the bank account.

We had several moments where we ran out of money. We had that moment, obviously, in the fall of 2013. We had 3 very difficult years, 2016, 2017, and 2018, where we had 2 rounds of financing that were incredibly difficult for us, the Series C and Series D.

COVID was interesting because COVID started—again, I think it's easy to forget these things because I think humans adjust very quickly. COVID, the first week, actually, the business tanked 80%. It tanked massively because everything was shut down. Then the next week, all of the dining rooms were closed, but the kitchens were open. Boom, we saw this massive doubling of the business in a week.

There are a lot of these moments. You've seen this at Stripe, obviously, a lot, where you have these moments as an entrepreneur where you have very, very low lows. The highs don't feel as high as maybe they get, but that's what you get.

As I think about the funding rounds you mentioned being hard, one of the investors I know who has consistently had the most steadfast belief about DoorDash is Michael Abramson, who was at Sequoia back in the day. Why was he so convicted? What was it?

I think you're going to have to have him on the show to ask him.

Okay, yeah, we will. That'd be good.

But I think Michael always had a fairly simplistic view of the company, where if 3 things worked, the business works. When I met Michael, it actually reminded me a lot of our journey in Y Combinator, because I think in Y Combinator, we were just happy to be there, but a lot of people were really racing towards Demo Day, where the objective function was to raise as much money at the highest possible valuation.

For us, actually, the objective function was answering 3 questions.

It was: would consumers pay for this product, pay a premium for delivery? Would merchants actually work with us and pay us commission? And would Dashers partner with us for a wage that we could afford? I think Michael had a very similar set of very simple questions about the company. As long as those dimensions looked right to him, he kept on investing.

I think he said, “You have to do the math and work out that the economics of this business do work,” and they do. Therefore, a more compelling product experience will win.

Yeah, he believed that if the unit economics were there, we could secure enough Dashers. I think his third comment was really that we can continue executing into the market opportunity and things will work out. He turned out to be right.

3. Why China Delivers More

Thinking more about this ecosystem, food delivery is very popular in the United States. It is even more popular in China. I feel like the experience of just being in a major Chinese city is one of seeing food delivery. You go to a restaurant and there are mopeds; you’re on the roads and there are just wild flocks of mopeds everywhere delivering food. The apartment buildings and the office buildings have dedicated zones to streamline it. Why is food delivery so much bigger in China than it is in the US?

Yeah, there are a few things. This is one where I’m always, on the one hand, so impressed by how far ahead sometimes behavior is in markets like China. Then, on the other hand, I have to remind myself that there are differences and why you can’t just clone some of these markets.

One of the first big differences is the eating-out culture in China is very, very high and very, very affordable. Eating out in China is about as affordable as cooking at home.

As a result, nobody cooks.

As a result, nobody cooks, exactly. There are a lot of reasons for this, and we can go super deep on retail history and grocery history in China. But anyway, that’s a huge phenomenon. I think, second, because of labor market dynamics, both the availability as well as the cost of labor in the Chinese market, that’s allowed a lot of these businesses to make this activity almost as affordable as if you were to pick up the orders yourself.

I think, third, especially—it’s not every city in China, but more and more—the Chinese… It started really with the Eastern Seaboard cities, where the density just allowed a very easy delivery setup, almost like every city was like New York City. I think in America, we think New York City is very special.

There are more megacities in China.

It has millions of people. In China, there are many dozens of cities with over 1 million in population, and the population density is even perhaps higher than that of Manhattan. Now, that’s actually morphed to the Western cities, too. It’s not just the Eastern Seaboard cities. If you actually study this over a time series, whether it’s the ’80s to the ’90s and certainly to today, because of how fast China is able to manufacture things, city-plan things, and deploy things—I mean, literally, they can build a train station in an entire day. Because they can do things like this, the market potential has really expanded, even as people moved now from Eastern Seaboard cities to other cities.

You’re saying a big part of it is lower labor costs mean food delivery is quite inexpensive. Does that affect then—

The cost of the food, I would start with that, is a lot lower. Then you have lower labor costs, and you also have mass availability of labor. Now you also have rising incomes because of how fast cities are getting developed. It used to be that most of the income for consumer discretionary spend, if you studied it as a percent of China’s GDP, was along the Eastern Seaboard, say, circa the early 2000s or even up to the 2010s. But if you looked over the last 15 or 20 years, that has now become a lot more uniformly distributed, kind of like the US, actually, whereas this is not true in other parts of the world.

Where I was going with this is, does that then predict the behavior of other markets? Is there also a lot of food delivery in other markets where food is cheap and labor is cheap?

And whether or not there are a lot of restaurants. People who can afford delivery, sure, but also the production of food. I think one of the things that I’ve always thought about is restaurants have 2 very special properties about them. I do think that, more and more, they are almost going towards the ends of these 2 spectrums. Same thing with retail, and we can talk about that, too. But restaurants, on the one hand, I think are centers of hospitality. It’s incredible, the level of attention to service they have. I think that will never get replaced by AI, that human-to-human connection.

On the other hand, where technology and autonomous technology can make a difference is that restaurants are almost like manufacturing production sites.

They’re a third space and a factory all in one.

It’s no different in retail. You have showrooms and warehouses. I think that, in the case of, back to your point around China, China has a lot of restaurants: over 7 million restaurants versus what we have here in the US, at about 1 million restaurants in comparison, for example. In China, you also have a very deeply capitalistic society, in the sense that wherever there is opportunity, you’re going to see lots of people go and fill that.

It doesn’t have to be a restaurant. It could be a kitchen. It could be coming up with different merchandising. If you’re a restaurant, you can sell not just food. You can sell T-shirts, and you can sell hats, and you can sell cookbooks. I think there’s a lot of that commercialism also in that industry, in a way that is much more advanced than other countries.

Yeah. Restaurants are very hard businesses, and they’re one of the most common businesses that people start, and yet they’re incredibly sophisticated, even on a small scale.

Yes, and they are also incredibly sought after. In fact, if you looked at the census data in America now—and this is also true in China—but if you looked in America, in the last 60 years, I think there are only a couple of years where the number of restaurants in the current year didn’t exceed that in the previous year.

Even though they are such difficult businesses, because of our love for food, for socializing, being with each other, and what a meal represents and how it can connect different people, it continues to grow. In fact, it’s the number 1 type of establishment that malls and a lot of shopping centers want the most of.

4. The Restaurant Industry Squeeze

Talk to me about the trends in the restaurant industry these days. If I think about what people talk about, there are dark kitchens. That was more a few years ago; people were talking about dark kitchens. There’s the fact that high-end restaurants can’t make any money these days because no one drinks. There’s maybe the emergence of new fast-casual trends, but also some criticism of the slop bowls and everything. But that’s just what people are… You’re in it. Tell me what’s actually going on in the restaurant industry.

One of the most difficult things is: how do you actually staff your restaurant? This is the number 1 challenge. This has always been the number 1 challenge. There are no easy ways around this, really. I think because the cost of labor only goes in 1 direction and only goes up, restaurants are increasingly making this choice on the continuum of service to manufacturing: where do I want to sit on that spectrum? I think that is 1 very big trend, that more and more restaurants feel like they have to go towards the ends of the spectrum, whereas before—

You have a local restaurant where people were sitting and eating their dinner and it was also fulfilling a few takeout orders. You’re saying that’s becoming less common?

Yes, because there’s a lot more pressure in order to make sure that everyone deservedly gets paid what is required to actually work there, especially inside some of these city centers. I think that’s a perennial trend. I think a second perennial trend, and this is why I think people love restaurants, is the innovation in restaurants is a little bit different than the innovation in, say, software, where innovation in a restaurant is just different types of cuisines and food. You’re always going to see the next type of restaurant.

I think this is best shown in what I was saying earlier around the number of restaurants. The number of restaurants always grows. That’s a perennial trend. That’s something that is not changing, even though the types of maybe food is changing, or what’s hot and what’s not, that’s changing. What’s not changing is the consumer’s appetite for these restaurants.

I would say number 3 is restaurants are thinking increasingly about how to scale, and this takes on different forms. In the case of some of these large brands, the big QSRs that you’ve heard of, which represent almost 50 cents on the dollar of restaurant spend in a place like the US, they’re thinking about, “How do I take my economies of scale and just keep going?” Because I have that as a big advantage.

You also see smaller restaurants who are trying to figure out, “How do I open up location number 2?” It’s really difficult. It’s actually really, really difficult in a lot of cities, even for the most popular restaurants, the hottest chefs in town, who you’d be shocked at how hard it is for them to raise the capital for location number 2 or even get the permits to start construction on location number 2.

I think that's another perennial trend that you see happening in the restaurant ecosystem. Those are the ones that I think a lot about: what are the things that are not changing, that are incredibly difficult, and that, if you can solve these big rocks, you can unlock even more value.

I presented a lot of obstacles, but if you actually look at the totality of the data, you would see that the amount spent on restaurants or eating out has increased over the last 75 years, every single year. These are rough numbers, but in the 1950s, when the government was measuring this, we would spend in this country, in America, maybe 75–80 cents on the dollar on groceries versus restaurants.

In the present day, those numbers are almost reversed, where it's almost 60 cents on the dollar on takeout and 40 cents on groceries. That number has only gone in these proportions, or this direction. Even though it is so difficult to make it as a single restaurateur, especially when I think about the days of washing dishes in my mom's restaurant, the totality of the industry continues to remain very strong, and it's moving in one direction.

When you talk about permits, famously, a lot of cities make it harsh to open a restaurant. San Francisco people have a lot of complaints. One of the innovations that you see these days in restaurant culture is the food truck parks and food truck collectives. I always find those bittersweet, where often you have really cool new concepts and stuff, but we have societally lost the ability to allow people to open restaurants and, therefore, we're reduced to a collection of food trucks in a parking lot somewhere. Are permits getting better or worse?

I think in general, similar to building things, they've gotten largely worse on average. That doesn't mean there aren't cities that are green-lighting and making it a lot easier. In fact, if you look at some of the fastest-growing cities in the country, they tend to also be the fastest-growing cities for restaurants. Not a shock. It's probably not San Francisco.

Phoenix and Austin—these places are doing a good job with permits.

Yeah. If you look at places in Arizona, like the tri-city area, Phoenix being one of them, Scottsdale and the Tempe area there. You look at the tri-city area in Texas, near Austin, or you look at what's happening in Dallas and the Dallas–Plano area, there are these pockets. In fact, if you look at the country, a lot of this growth is happening in the South of the country. That's been true for a couple of decades now.

They tend to be correlated, meaning if it's easy for me to build apartment units and to build construction in general, it tends to be a bit easier to also get the licenses to open up a restaurant.

That's good—that there are some bright spots amidst a nationwide, maybe negative trend when it comes to permitting.

I think when you're in any business, certainly the restaurant industry or even our line of work, I would argue, you have to be an optimist. You have to see the silver lining on the way out of how to get there, but it's a slog.

I do think that one of the responsibilities we now have as a company, as we're a little bit bigger now, is to help on behalf, particularly of these small businesses, to represent them almost as a class and actually try to even the playing field a bit and make it a bit easier to actually be a small business.

What do you argue for on behalf of small businesses? When you talk to someone in Washington, what are you saying?

One of the first ones is actually making it easier to open up location number two.

Permitting reform.

Permitting reform is one of those things that we certainly stand for. Making sure that labor laws are changed, or when they are changed, that they take into account the restaurants. We're talking about one of the largest sectors of job creation here that is always growing, by the way, every single year, which is a rarity but a truism in every economy, to take into account what the restaurateurs have to say.

Those are some of the big topics. Again, I look at the things that aren't changing in terms of the difficult challenges of being a restaurateur, and we try to fight on their behalf.

5. Technology Expands Restaurant Margins

Where has technology meaningfully changed how restaurants work? A very visible example to people might be in fast-casual restaurants. They have managed to change the labor equation, it seems. You might know the actual number, because it's self-serve ordering and pickup at the counter and stuff like that. That's much more common than it was, and so you need fewer people for the same volume of orders, or you can run much more scale with the same number of people.

I'm guessing there's supply chain stuff, but it's less visible to us as people who visit restaurants, and you'd know. As you think about where tech has made the biggest impact versus 10 or 20 years ago with restaurants, what would be your top three list?

Not to speak our book, but one of the first things that comes to mind is actually something like delivery. When you look at the economics for a restaurant, rough math: for every dollar of food that you and I bought inside of a restaurant, from the cost structure, you're going to have 30% of that in the food and the packaging, 30% in the rent, and 30% in the labor—something like that. Rough math. On that dollar, a restaurant can net 10 cents.

On a takeaway order, you're still paying for the food and the packaging, but you're, largely speaking, using the same labor and paying the same rent. There are some exceptions where some restaurants have become delivery-only and things like that. But by and large, on average, you're going to make 3–5× the incremental margin.

I think, number two, restaurants recognize that they need to build relationships with customers that are not just over the telephone or through in-person visits, where they're relying on their memory.

That area is still, I would say, somewhat incomplete, where there are products where you can manage your social media pages and maybe have a database of relationships that you recall. The challenge, however, is you don't get to see everything. You don't get to see the full visibility of your online orders and your in-store orders.

Another challenge is your staff turns over. The average staff may turn over every other day. Not every other week—every other day.

I do think that there's still a long road ahead there, in both making that product a lot more robust for restaurateurs to be able to have those relationships.

Back to one of the things we talked about in this conversation about the importance of retention in a consumer business: it is so important for a restaurant to build the concept of a regular, which is a retained customer. You and I, I know we travel a lot for work, and so we wouldn't always be great customers for a restaurant.

What's more important, actually, is if they knew that there was a regular who was actually coming and should be prioritized. Those are the kinds of things that I think, hopefully, we, DoorDash, can help solve.

Don't most restaurants do a bad job of recognizing regulars? You think about the airlines as an industry. They've made it highly irrational, as a frequent traveler, to travel on anything other than your preferred airline. They've really created a lot of lock-in there.

Whereas, as I think about restaurants, there is the super-mass-market punch card—10th coffee free and stuff. Then there's the really old-school restaurant where the owner might greet you and might comp your dessert or something, but it's very sporadic and very dependent on the owner being there because, as you say, when the new waiter or waitress is there, they don't know you.

But there's nothing in between that's systematic, and it feels like some way of rewarding the regulars is missing.

It's a hard problem. Let me ask you: what is your favorite food to eat?

It depends. At the weekend, we went and got good Taiwanese food, and that was fabulous.

All right. How often do you eat that?

Once a month, once every two months.

Once a month, once every two months. That's the challenge. Even your favorite food is probably not something you're going to eat every single day. When you're talking about an activity like eating, which is 20–25 times a week—

It's not like coffee.

Yeah, it's not like coffee. It's even different from travel, where the product may be the same every single time.

I think it is tough to create what you're describing, which is this program in which you'd be known as a regular. But that, I think, is the opportunity that at least we're thinking about at DoorDash, because if you think about it, our product, with now over 100 million-plus annual customers and tens of millions of monthly customers, serves people who shop at various frequencies and are building relationships with all of these businesses—sometimes many times a day, sometimes once every other month, sometimes once a year.

I think because there are 20–25 moments per week, there is an opportunity in which we can create that. Now, we recently acquired, last year, a company called SevenRooms that manages a lot of the in-store bookings of restaurants, particularly higher-end restaurants.

If we can democratize that and give it to everybody, then also layer on the DoorDash data set and actually tell you that John is a regular, that he tends to like the following things, and here are other things that he tends to do, perhaps we can crack open that question.

You want to help restaurants take care of regulars?

Yeah, because that is how you build a great consumer business.

You need the retention. That is true for every restaurant, every consumer business.

6. The Limits of Ghost Kitchens

We were talking a lot about ghost kitchens, I don't know, 5 or 7 years ago. You don't really hear about them that much. How big have ghost kitchens become?

They're relatively small.

People thought the model was going to take over, right?

Yeah, the model sounds really reasonable, and I think logical on face value. On that spectrum of restaurants, where you have high-end service on one end and hospitality only, to perhaps delivery only or more of a manufacturing concept on the other end, it seems reasonable that you can borrow a small square footage of space, not incur a lot of not just the fixed cost but also the labor cost of actually running your “restaurant,” and then sell through a delivery platform, acquire your own customers, or do something like both.

It just turns out it's extraordinarily difficult, however, unless you're a large brand or a house of brands, someone like DoorDash, to attract enough customers to make that math work.

Is another way of putting it that it's hard for a ghost kitchen to be as efficient as a restaurant that's also churning out orders across both in-restaurant and delivery? It's just hard for them to compete on the economics?

Yeah, it's slightly different. Perhaps we can look at 2 types of examples. You can have 1 type of example, which is perhaps my mom's Chinese restaurant. It's an SMB where the name value outside of literally its locale isn't that well known. It's very difficult for that restaurant to acquire enough customers to make that math work because they still have to—remember, I mentioned that staffing was the number 1 challenge that restaurants face—they still have to staff these kitchens. That's 1 of the challenges.

You go all the way to the other side. Pretend you are a large QSR with a big brand presence. You certainly can attract lots of customers, but you're going to have to think for yourself, “What's my opportunity cost?” My opportunity cost might be to open up another restaurant. Why not do that and take my economies of scale and actually apply it to more orders, in-store orders as well as to-go orders? Because I can do that with a restaurant. It's hard to do that with a delivery-only kitchen. I think because of that, it's a tricky model to scale to every type of restaurant brand.

I think people maybe thought that big chains were going to do it more, where you're right, the billboard effect for an actual retail presence is hard to compete with. You get this very cheap customer acquisition from the fact that people know your name from the high street or maybe have been to you, and as a ghost kitchen, you need to make up for that awareness gap.

But if you're Chipotle, maybe people thought that the orders get fulfilled out of a dedicated Chipotle central factory, and then the retail spaces are separate. But that also hasn't really happened that much.

That happens occasionally. I think these are not decisions where we're accounting for all of the variables. One variable we're not accounting for is, “What else could I do with the space?”

You have to remember, for businesses like Chipotle that tend to identify real estate in fairly expensive areas, there's a high opportunity cost to what you do with that space. Yes, you're right. One choice is to turn it into a kitchen or a delivery-only kitchen, and sometimes that happens, but there's also a massive opportunity to recoup the expense. If you start introducing, for example, franchises—Chipotle is not really franchised, but if you look at other types of brands where it is franchised—I think the economics become even trickier.

There's a next-best-use math. Which restaurants have actually invented something that's hard to copy?

I actually think any restaurant that's been around for, let's say, 2-plus decades probably has very interesting IP. Obviously, some of this information is not public. You can imagine when you go into a McDonald's around the world, that french fry—perhaps they don't sell the same exact items in every single store in every country, every city—but the french fry almost always tastes the same. That is an extraordinarily difficult feat to accomplish.

There's a lot that goes behind the scenes, just like there's a lot that goes behind the scenes at DoorDash in terms of getting you 1 order on time to make that sentence true. The same can be said about a lot of other businesses that have been around for very long periods of time. That's on the big-brand QSR side, where a lot of the innovation, if you will, is in the process innovation. Also, how do you run large groups of people and have very high and consistent standards of service? Extremely difficult. That's the IP, I would argue, for a lot of these large brands.

On the other side, there are small restaurants, some of whom have been around for almost a century, actually. There aren't that many of them. But when you look at what makes them special, it tends to be the same things that actually make startups special. It tends to be a small group of people that really believe in a certain idea, some commitment towards some standard of excellence. It could be excellence in the food, the service, or takeout. There are some takeout-only businesses that have been around for many decades. As a result, they have high retention, high usage, and they're very efficient.

Can you speak about any specific examples of the impressive IP that these chains have conducted over the years? It could be the McDonald's french fry, how they actually make it taste consistent—I don't know the answer to that—or any of your favorite restaurants?

One of the things that I'll never forget was going through a hospitality training at Chick-fil-A. I want to say it was 2017, maybe it was 2018. I don't remember everything on the checklist, but one of them that stands out is that the floors must be so clean that babies would lick them or could lick them.

The babies will lick them regardless of how clean they are.

Exactly.

Is Chick-fil-A still the highest-grossing fast-casual restaurant per square foot?

I don't know, actually, but it's definitely up there.

People talk about it. It's up there.

It's definitely up there.

Why is it so successful?

It actually goes back to, I think, the standard ingredients of a great consumer business. It has a highly retentive following, which it spends very little money acquiring, and they come back often. They're extremely efficient with all of the scale economies behind the scenes around equipment, which is quite custom to that company, the service training, and how they treat their staff.

One of the most impressive facts about that business is that many people I met with have spent over 2 or 3 decades of their careers at Chick-fil-A. We find that to be a novelty, actually, here in Silicon Valley or in technology.

And in the restaurant industry.

In the restaurant industry, one where we described all of the churn and the difficulty of staffing, Chick-fil-A has extremely high retention of staff. It has so many repeatable stories of where it gives somebody who comes from not a lot of means, frankly, the opportunity to become the general manager of one of these stores and earn a really healthy living, pay their entire family through college and beyond.

In many ways, a lot of these restaurant stories, Chick-fil-A being one of them, still have a lot of the essence of the American dream.

We're talking a lot about QSR chains here, and people often like to ask the question, why is there not a Chipotle of my preferred cuisine, where it's slightly elevated versus regular fast food, with very good, consistent quality and good execution? There is no chain that serves you really good Neapolitan thin-crust pizza in whatever city you're in, or Indian food, or again, pick your favorite cuisine. Why don't we have more chains delivering just a really good standard version of a cuisine?

Have you ever cooked for a large group of people? Or do you cook often?

Yeah, we cook, but not for huge groups.

One of the things I found is that as the number of people you have to cook for grows, the difficulty of keeping the standard of service—whether it's the taste, the temperature, or the speed at which you receive the food—goes up exponentially. That is what's so hard about exactly what you just said, around keeping that quality control really high.

It's maybe easy to build you the Chipotle for 1 type of cuisine for 1 restaurant.

That’s just from the skill perspective. Although I do have some ideas on how perhaps you can do that, the second reason is to think about who you’re competing against. Let’s say that you want to make a Neapolitan pizza. You mentioned that—or any kind of pizza.

I do.

Great. I can’t wait to taste it, John. Did you know that there are over 1,000 different pizza choices in the Bay Area on DoorDash? Just on DoorDash. We don’t have all of the restaurants. That is just one of the many things that you’re going to have to do to be able to create this concept and actually scale it. Then don’t forget all the other challenges we talked about: the staffing and the permitting.

7. Scaling Restaurant Quality

What determines whether chains succeed versus whether they don’t? Because there are a lot of pizza restaurants in the Bay Area, but at the same time, there are a lot of Mexican restaurants in LA that are really good, and yet Chipotle still has plenty of locations there. It doesn’t just seem to be whether local options exist. Sometimes a chain can break through. I also don’t think it’s a skill issue because it’s very processable to make a good Neapolitan pizza. It’s about the ingredients, the temperature and humidity control in proofing the dough, and then cooking it at a very high temperature, but it’s not rocket science. You could make a process around this.

Potentially. I think the fact that there only exist so many of these QSRs—a countable number of these brands that have actually scaled to the size that we’re talking about, where they’re ubiquitous—should be some evidence—

It’s harder than you think.

—that it’s a lot harder than you think. Because of what I just said, there are a lot of process skills that you have to be extraordinarily good at. In fact, I remember we partnered with one brand where we launched this brand to great fanfare and sold a lot of burgers on a weekend, testing this idea: can you make a healthy-tasting burger if you put a brand name around it?

The answer is yes, and it’s really hard to retain any of those customers.

Because it’s too healthy?

No, because half of the reviews said, “Your burgers were great,” and the other half said that they were inconsistent in quality. It’s because it’s extremely difficult. That exponential challenge, as you try to grow and scale while keeping up the quality, is very, very, very difficult. You have to get the service right, you have to get the production right, and you have to get the pricing and the packaging right. There are lots of things you have to get right.

Do you think reliability is the thing that people underestimate about the restaurant industry?

Yes, or about a service like DoorDash. I think when you have high volumes of activity, keeping the reliability as reliable as the electricity we have or the water inside our buildings is extraordinarily difficult.

8. Building the Future of Delivery

Can we talk about the future of delivery and how you see it playing out in 10 years’ time? There are a few different modalities being discussed. There’s drone delivery, there’s sidewalk robot delivery, and you guys have Dot. I’d love you to talk about Dot. I don’t know how much you’re doing in drones. Maybe you can talk about that, too. Maybe there are autonomous vehicles, like autonomous cars. People don’t talk about that as much, but maybe in the suburbs it would kind of work. This is what we have at the moment. What’s the mix in 10 years’ time?

Even before we go there, I think the first question might be: what might be delivered? I think that’s actually still not that well understood. Today, yes, DoorDash is probably best known for food, and potentially now increasingly groceries and household items. But that is a fraction of the tens of millions of items inside every single city.

One of the biggest things that we’re going to have to do before we can just fulfill the items, which is what we’ll get to, is figure out where the items are and what the items are. There are tens of millions of items literally inside these cities, whether it’s in the US, different countries within Europe, or other parts of the world, and they’re not cataloged. There’s no structured data that you can scrape in order to figure that out. I think that’s going to be a long journey.

Sometimes the inventory isn’t even close enough, and so you may have to house the inventory so that you can actually bring those goods. A lot of what I believe makes for great logistics is that you need a great setup. Because if you’re always trying to figure it out on the fly, pulling some magic trick out of thin air, and having to be the hero, that is not a system. That is an exception.

I think great logistics, or great reliability, requires building systems. You have the catalog, you’ve got maybe the inventory close enough to where people live, and now we can talk about the fulfillment. The entirety of that is what you have to build before we can get into the great technologies.

The staging of what it is that you’re delivering.

Because if you don’t know where to get the items, or if you can’t get the right item, what difference does it make if you have the right vehicle?

You’re talking about the fact that you don’t actually know what’s on a supermarket’s shelves, or the fact that you have to actually be able to predict when the restaurant will have the stuff ready—all that knowing what is ready to be delivered.

Yeah. Or take when the Cheeky Pint becomes an established bar, and maybe DoorDash can deliver from it one day. What’s inside the Cheeky Pint? What’s on the actual menu? What’s off-menu that you can actually order? These are complicated issues before we even get to the fulfillment of the actual items.

Then you can get to fulfillment. I think that’s one where we’re still exploring. To me, the most important question we always ask on any technology, but certainly a technology like this, where there are long cycles of development, is: what problem does it actually solve?

Obviously, in the case of drones, as you mentioned, they can cover a lot of ground very quickly. Longer-distance orders seem to be a great job for them to solve. In a busy, high-rise area where you’ve got short, dense orders, which is the majority of these orders, maybe they’re not the best solution for that problem. This is actually what led us to the creation of Dot, which was when we started the Autonomy Project in 2018 or 2019, we didn’t set out thinking that we needed to build anything. Surely there would be someone who would—

We’ll just pick the best one, and I’m sure we can buy some tech off the shelf.

Yes, we were out there raising our hand, asking for a partner to the dance, so to speak—someone who could specialize in that, while we could specialize in the operations and things like this. It turned out that no one was that interested. A lot of people were interested in building robotaxis, and obviously, we’ve seen that come to fruition. A lot of people did build, to your point, sidewalk robots, but we found those to be too slow.

Back to what problems are you solving? Because you’ve got to remember, humans are very good at delivery, actually, it turns out. We’re also very good at driving, too. The bar is actually fairly high. It’s not as simple as just saying, “We can make the technology; let it rip.”

Then we said, “It has to be fast enough. It probably doesn’t need to be the size of a car because cars aren’t very good at parking inside these busy locations where a lot of these restaurants are.” We had to think about all these things. We had to think about how to load the vehicle, and we had to think about unloading the vehicle because, unlike a robotaxi, the passenger solves all of those problems, but in our case, we have to solve that. That’s how we ended up on this.

Where are you in rolling out Dot?

We’re primarily testing right now. We really want to get it right in one market. We’re doing a lot of this in Arizona because it’s not just the technology. One of the things I’ve learned and come to appreciate about building something like Dot is that these autonomous vehicle companies are almost many companies inside one company, or certainly many different types of skills you have to be good at.

Yes, I think everyone obsesses over the autonomy, but there’s also the hardware, there’s also the manufacturing, and there’s also the operations.

There’s hardware reliability.

There’s also the maintenance. There’s also the regulatory side. There are a lot of different components that you actually have to be good at. Then we can talk about the stack required to actually build the inputs to some of these things. For example, there’s also understanding how to procure supply when you have geopolitical tensions or supply chain disruptions.

Where do you do assembly versus where do you do construction? There aren’t super-obvious answers to some of these questions at times, especially when you have to make these decisions sometimes years out. I think that’s the stage that we’re in right now: “We’ve got to figure this out correctly before we can actually scale it.”

Where do drones fit in, if at all?

They definitely do. As I mentioned, going back to jobs to be done, I think drones obviously can do a lot of these longer-distance orders.

We've been doing drone deliveries for a couple of years now, mostly outside of the United States.

Whereabouts?

Places like Australia. We're going to bring them to Europe and the United States as well. Again, you have all of those problems you have to solve. The autonomy is a little bit easier. You still have a routing problem, hardware problems, permits and regulation, setup, loading inside different stores, and things like that.

Very interesting. We're talking a lot about some of these scale tech problems, and one of the ones that people might not think about is fraud. I feel like Stripe and DoorDash obviously work together in many ways, but one of the things we work together on is fraud. I feel like the number of potential fraud vectors is much more complex than people think. Maybe you can just talk about trust and safety at DoorDash—all the different ways that people try to game this complex system, and what it takes to run that.

9. Trust and Safety at Scale

Many times, I think about my freshman-year electrical engineering course on state machines, where DoorDash, in some ways, is like a large state machine. You have many systems in place, and they're observing what's happening. There are things that must happen when things are green to keep them green. But as we start degrading gracefully from green to yellow to orange to red, something like fraud, as you mentioned, means other systems have to kick in and do things like exception handling.

Obviously, you're spending most of your time building prevention systems, but of course, it's tricky. You also have to have the 911 response system that responds within minutes, because sometimes we're talking about physical safety. We're not just talking about fraud. Given the volumes that DoorDash carries—billions of orders per year—the one-in-a-million incident, unfortunately, does occur. That's DoorDash as a system.

Now, in fraud, you're right. We partner, for example, a lot on online fraud. You have the largest database of all of these transactions, and therefore, you can give us lots of information that we ourselves couldn't get on our own. On the flip side, there's also offline fraud, which is very tricky to catch.

What do you mean by offline fraud?

For example, let's say that a consumer suggested that an order was never dropped off. Let's say that a Dasher said that an order at a grocery store was already shopped and picked, but we're not sure. Or let's say that a Dasher said an item is missing inside of a store—any store, maybe perhaps an apparel store. All of those, I would argue, have offline components that are difficult to verify.

Building the eval, if you will, for this set of conditions and use cases is actually extraordinarily difficult. A lot of what we're trying to do is build those signals ourselves. There is no perfect science here. I wish there were, but in many ways, we have to invent this. This is actually why I'll give you a few examples.

One of the things that we shipped a couple of years ago was called SafeChat. We noticed that prior to any physical altercation that may unfortunately happen between people, 90-something percent of the time, it's always preceded by a verbal altercation. Knowing that alone allows us to prevent a lot of tough situations. Or building alert systems.

For example, sadly, there are shootings, particularly in the United States. There was one shooting in Manhattan where several people died in the lobby of a large office building. Within minutes of a Dasher seeing the shooter walk into the building, we have an alert system that doesn't just alert all Dashers, consumers, and merchants within that vicinity, but also all local law enforcement. Actually, we were the first to report that to the NYPD, who tried their best to respond in time.

There are a lot of those events. It's not always about fraud, is what I'm trying to say. You have to build extremely responsive systems, and you also have to build as many signal-detecting systems, too.

You're running a fairly large slice of the economy, and you're trying to have everything run smoothly. There are quite a lot of failure modes that people might not think about when you're seeing so many economic interactions all day.

Exactly. We're trying to be the digital representation of these cities, whether it's a city, a suburb, a neighborhood, or a rural area. Sometimes it's not always about what's transacting, but just what activities are occurring.

One of the hardest kinds of frauds to prevent is the classic, say, credit card fraud—the one that people think about, where your card got stolen and it's actually someone overseas buying stuff for themselves rather than you buying it for yourself. We've now gotten very good at detecting that. It's quite hard to do. There's just so much data we can bring to bear on that problem.

A much harder one that we see businesses face is the kind you're describing. The industry term of ours, which is funny, is “friendly fraud,” which is a bit of a euphemism. The customer buys something and just says it never arrived when it actually did, and they're just straight-up lying. You don't really want to call them on it. You also don't have proof that they're lying, but it definitely does happen to some degree. How do you deal with that? Maybe there's a photo of the Dasher having delivered it, but I'm curious how you deal with people just defecting from the system.

There's no one answer, as you know, with fraud, because as soon as you figure out how to measure and contain the risk, obviously, the incentive for the fraudster goes toward the next axis of breaking in. I don't think there's a perfect solution to the problems that you're describing, but you're right. A lot of this is about how you build systems where the metric is around reliability, consistency, and great customer outcomes. The inputs are a lot of measuring things before they happen.

I talked about how one of the most important things about logistics is the setup. That happens even before a delivery, when somebody hits “place order.” The setup is extraordinarily important. Did you get the right catalog? Did you label the catalog correctly? A lot of that sort of stuff.

In the case that you're describing, we can track several things. We obviously can track what is happening with the Dasher. You mentioned some of these ideas, like taking photos of receipts or orders being dropped off. We have our own mapping system, for example, that we built.

Why did we build our own? It's because we care much more than any third-party mapping system about exactly where the last two feet—not 20 or 200 feet—of some apartment unit door is, for example. We know if we reliably deliver to that door and saw the pin hit exactly where it should, so we have a little bit more fidelity on whether something was dropped off.

We build profiles of customers, I'm sure you do, too, based on their behavior and what they tend to say. There's lots of things that we're trying to do behind the scenes to build these signals and figure out the likelihood of fraud.

Makes a lot of sense. Tell me about DoorDash Tasks.

10. Cataloging the Physical World

DoorDash Tasks is something that we shipped recently. Actually, it started with a project where we were trying to solve our own problem. It started with trying to figure out where every item inside the city was. We have millions of Dashers, and they're usually frequenting different types of stores. We know that consumers move items around a lot, which makes it extraordinarily difficult for any store to track its own inventory. Dashers sometimes are collecting this information. That was one of the problems that we were solving.

Interestingly, while we were doing this, we started getting calls from people you probably would expect—retailers and CPG companies who might be interested in this information. Then, especially as companies—I think it's well known at this point that a lot of the large LLM companies have been building lots of repositories of online information—what about the offline information to bolster some of those models, or just to help some of these companies building robotics and things of the sort?

We said, “Interesting. We're building a catalog for ourselves, but we ought to be able to help other people, too. If it's an opportunity to help Dashers earn more and give them more choices to earn more, great. Why not build that solution?”

That's really the story of DoorDash Tasks, which started organically probably a couple of years ago as we were solving our own problem. We started realizing that there may be other people who have similar tasks.

It's a platform for having fairly small tasks done? You can give the task to DoorDash and you guys could complete it?

Yeah.

Do you get excited about an AR angle for DoorDash? Are you in the best use cases of the Meta glasses, where now you can get directions within the store and everything? You'd actually speed things up quite a bit rather than wandering around lost.

I think we definitely could be. Again, back to the question of asking, “What problem does this solve?” The human is pretty good with their eyes at some of this stuff.

The bar is pretty high in order to do some of the things you're talking about.

As I think through AI applications for DoorDash, one thing I'm struck by is that LLMs are pretty good recommenders just by tossing things into context. You don't need to train a custom model. I don't know if you've ever tried book recommendations or TV recommendations. You just tell it a bunch of the stuff you already like.

Restaurant recommendations are like, “Here are 15 restaurants we like to go to. Give us other recommendations,” and it'll do a really nice job. Yet within products, if I open DoorDash, it's the same categories, and it's less personalized to me than if I took my DoorDash history and put it into an LLM. Shouldn't we be somehow using the fact that LLMs are pretty good recommenders within products? This is not just DoorDash; it's every product I use, it feels like. Those recommender capabilities are underutilized.

No, I think you're definitely right that there's an opportunity here. There's almost the traditional school of thought, which is to use the information that you have and build the best personalized models that you can. One of the things that LLMs obviously do is throw efficiency out the window and throw as much compute toward it. Interestingly enough, one of the things that spits out when you put in enough tokens and have big enough context windows is—you're right—it has actually better models because it's just using much larger data sets.

And it's a world model.

Exactly. I do think there's an opportunity in what you're saying, which is that products like DoorDash, like consumer products, will definitely have their own ordering agents. I think we're all trying to figure out what the right modality is, because it's not obvious to me that everything's going to be done through text. We don't buy things through text. We don't always get inspired through text, either. I think it's something that we're all figuring out.

What else are you excited about these days from a product point of view? What's the new part of the product that's close to your heart, or where do you want to take things?

We have a few missions that we're on, and we talked about several of them, I think, in this conversation. One of them is that we've got to bring you everything inside the city, and we're a long, long, long way from that. We can go super deep on restaurants and food, but as you start going into category N+1, we are just a smaller and smaller and smaller drop in the ocean. We've got a lot of work to do.

A second mission is that we also want all of the businesses to have their own software. DoorDash's goal is not just to grow your business by sending you customers. We also want you to learn how to do it on your own—that's why we have products like DoorDash Drive, Storefront, or SevenRooms, which are B2B products, and is probably why consumers don't think of us for them—and help you build your own omnichannel business. We have a mission in which we want to actually bring you inside stores. That's actually something that we're trying to do. We started that about a year ago.

I mentioned earlier that one of the perennial challenges or trends for restaurants is, how do you grow your own brand? If we can also help you with that, both by helping build for you a CRM with a 360-degree view of guests and by sending you customers inside stores, we'd love to do that. We're starting with restaurants with 2 products. One is called Going Out; the other is Reservations. That's a big mission that we're on. We have several of these missions that I'm pretty excited about.

You mentioned reservations. There are a number of quite established companies in that space, and restaurants are famously not exactly eager to wake up and switch all their systems one morning. How do you plan to unseat the existing guys in reservations?

I think it starts with the idea that no restaurant should own its own reservations technology, ideally. It starts with that. A reservations book—we won't get into it super deeply here—is actually quite complicated to run. I mentioned earlier this idea that restaurants' North Star really should be to build regulars, but that's not how every reservation system is designed.

I think the most important thing is that restaurants can have reservations on any platform. That's what I believe. That's one of the main premises behind SevenRooms, too: It's agnostic to who generates the demand for you. What it's trying to do is make sure it achieves your objective function. It may be to grow profits one year, to grow locations the next year, or to get a certain type of customer the next year, and so on and so forth. I think it starts by building what's best for the merchants.

Hopefully DoorDash can make a difference, too, where we just have more customers than people that play with restaurants, and also more information that we can supply to these restaurants on how to make more money and achieve whatever outcome you want. If we can help achieve a restaurant's objective function better than anyone else, then hopefully we have a seat at the table.

Tock in reservations carved out a successful niche in high-end reservations. I think about one of the beliefs that informed their products most strongly: no-shows are terrible for restaurants, and a restaurant reservation should be like a flight reservation, where you don't just get to decide you're not going to go to the restaurant.

You put a little deposit down.

Exactly, you put a little deposit down. Do you subscribe to that philosophy? Do you think it's maybe less relevant because they were really dealing with very high-end restaurants? Do you think it's less relevant for regular restaurants? I'm just curious what you think of that viewpoint that we're doing reservations all wrong, which was, I think, one of the founding insights of Tock.

I think it makes a ton of sense. The implementation of the idea is what's difficult. Restaurants that have very limited seating and very long seatings have a very high opportunity cost for a cancellation, versus a restaurant that may have infinite production capacity, if you will. So I think the implementation of the idea is what matters, but I think the idea itself makes a ton of sense.

How do you, with reservations, plan to ensure that people aren't no-showing to restaurants?

We're working on that problem right now, but I don't think there is one thing. I think the first thing is we have to help you find reservations that you want to go to. I think there's a lot of room left in the space to innovate there. I don't think there's been that much innovation in that area.

So discovery is almost a bigger problem?

I think discovery is very difficult. How do you find out about restaurants today? I'm curious.

My wife finds them on Instagram.

I think that's right, because there are so many different sources now. Instagram is one. I'm sure other social channels are another. LLMs, as you mentioned earlier, are about throwing things in there and getting something out. Friends, I'm sure, are another. I think it's really difficult still. No matter the sources of input, most restaurants still struggle to generate their own demand. I think so long as that's true, we have a shot at making a difference.

You talked about your European expansion. You just acquired Deliveroo. In what ways are the Deliveroo and DoorDash businesses different?

Fundamentally, they're more the same. I would start with that.

Hence, the acquisition.

Instead of the—we talked a lot about systems and logistics. Have you been to London? I'm sure you've probably been there many times. You probably know it better than I do. It is an interesting city in that it doesn't actually have a lot of the grid-like hub-and-spoke properties of a lot of cities. There are a lot of reasons for that. We could go into the history.

It's mostly age.

No, there are other cities that are actually designed differently in other parts of Europe. We can get into other countries and cities, but age is one of them. Obviously, it's not meant to be a city where you drive a lot of vehicles. No vehicles are advantaged.

I think it's older than even a lot of other European cities. Even as European cities go, London is really old.

It is. As a result, that alone makes the logistics problem very, very different. The logistics—not just the algorithm, which I think is just one part of the system, but the signals that you'd want to collect—are very, very different. Vehicles are very different. We predominantly are non-auto in the City of London, which is very different in the States, as an example. That's very different.

The regulatory setup is different. It has parts that rhyme with the US, but as you well know, each country is slightly different. Even though, I guess, the UK is not under the EU, within the EU each country has its own local version of regulation. That's very, very different. You know this probably better than anyone in the world. Payment processing is very different. There are many more card types in parts of Europe than, say, in the US. Industry dynamics in retail are very different.

In the US, we have a fragmented retail industry that has a lot of different players.

That's more concentrated in the UK?

Yeah, in the UK, in Germany, in France, and in a lot of European countries, this is more the norm than the exception. Whereas in the US, you have maybe hundreds of different big brands that are strong in certain regions. But in Europe, that's less true. Those are some of the immediate differences that come to mind.

A mystery to me is why the pharmacies are so much better in the UK versus the US. You go into a Boots versus a CVS or Walgreens, and it's like night and day.

One of the things I would say, just as a standard matter, is because of the concentration. It's really concentrated by category. Pharmacies could be a category. Apparel can be another category, et cetera. This is true in Australia, and this is true in Germany.

The standards are higher, and the service is more consistent. In the US, when you have tens of thousands of supermarkets and hundreds of big brands, it's different. If you look even more upper-funnel, say CPG, there are a lot more CPGs in the US market competing than in some of these other countries, which makes things like inventory management, SKU counts, the size of the stores, and all of that different.

You referenced some of the stuff we're doing together. I think of Stripe and DoorDash as almost having grown up together, founded at very similar times, and then you guys were early users of all manner of products, be it Stripe Connect, Radar, Issuing for cards, or other things.

What would you like to see Stripe do that it doesn't today, either because you guys would like to use it or just because you think it's something we should be doing? I've been giving you all my silly product suggestions. This is where you get to turn the tables and give me yours.

I would always start with what our teams would say. Probably the first piece of advice always is, “What problems are you solving for us?”

I think especially as companies grow in size, success, and variety of products, there's a natural temptation to start imposing what the company wants to do or wants to solve or wants to sell, versus saying, “What problem does it solve for the customer?” I am sure one of the things that could be true is making sure that you're solving our top problems, versus just introducing a range of products.

The next thing.

That's probably one thing that we probably both share as advice for one another.

Wow!

We got Dot coming.

This is the—awesome. This is Dot?

This is Dot. There you go.

Wow! It's kind of like a TARDIS.

It's very cute, but I can't believe that we actually have Dot with us today. It's traveling backwards, but here you go. You can see it there.

The LED screen is for the delivery?

Yeah, the LED—

Is for the person?

Exactly, where we can send different messages. We can even give messages through the eyes. We thought a lot about the eye contact between Dot and a human. You see part of the sensing stack there in terms of the cameras, the lighter up there.

This is a lot bigger than I expected. It's like a small car, not a big robot.

Yes, although it's still—

It's like a Smart car.

Yeah, it's like one-tenth the size of a car, but it can travel up to 20 miles an hour. When you think about how that compares to what a car travels in traffic, it's about the same.

Wow!

This is how we load it up. We got you some non-alcoholic Guinness.

Have you had Taytos?

I actually have not had Taytos.

Ireland's preferred. Here, have some of those.

These are the chips?

These are the Tayto chips. Very good. That's cool.

That's awesome. The compartment is totally custom and modular, which means that depending on your needs or what we want to fit, whether it's heating units, cooling units, or all of that, it can be configurable as well as built modularly. But yeah, this is Dot.

Where will this first be most useful?

There are a lot of deliveries. You can think of almost suburban-like environments, where the deliveries are probably under 5 miles, and there's some challenge for a Dasher, for example, parking. We talked a lot about that earlier.

So suburbs are literally a great place for it?

Yeah, and then I think certain city centers, depending on the actual area, though. Those are the two types that we're thinking about: distance and the types of environments.

One of the coolest things that we're actually building in Dot, that's actually not physical, is what we're calling our autonomous development platform, which is basically the algorithm that determines which orders go to which types of vehicles. Dot, for example, obviously can also travel mixed routes with humans. For example, if it's a longer-distance delivery, but there is a part of that journey where Dot is more advantaged, say, parking, it can do the first leg of the order.

First leg and then hand it off to a Dasher?

Yeah. Thinking carefully about—

Can it go on the roads?

Yes.

And the sidewalks, too?

The sidewalk and the bike lane. It's the only vehicle in the world that does that.

Thinking carefully about the handoffs is very, very important to actually making autonomy work because this is very different from robotaxis, where obviously a passenger can solve a lot of these “edge cases,” but we actually have to think about all of it. I actually had no idea. This is cool to see Dot here.

How far can it go on a single charge? Or how many deliveries can it really do?

It could do hundreds.

Really?

Yeah, it really could. We probably wouldn't test it to failure, so to speak. A lot of this is making sure that you build it ready for mass manufacturing. We talked about one of the things you have to do in building autonomous vehicles: you have to build it ready for mass manufacturing. You also have to build it for easy operations. The swapping in and out of the right vehicle parts is very important.

Where is the battery? Is the whole base batteries?

Yeah. There's some compute in there, there are some batteries in there, and there are some more sensors.

The autonomy is all local?

Yeah.

On the device?

Yeah. There are a lot of other things that we're going to have to keep doing to test Dot in different types of weather.

This is a much more sophisticated sensor stack than I expected.

If you're traveling on the road and you're going up to 20 miles an hour, there are some physical parameters that you're solving for, which is what guides the development of the actual stack. You're right in saying that—

Just when you see it, it's surprising. It's not a baby sensor stack.

It's not a baby sensor stack. I guess we're not trying to build any particular technology. We're just trying to make it work.

Is the fan for the fridge or the compute?

The compute. There are a lot of different versions that we're working on right now, both in terms of what the temperature settings could be internally, what the heating systems are, and making sure that we don't overheat. There's some interesting braking systems that we're building as well in Dot.

Why is the braking system interesting?

A lot of the question is, what type of brake are you going to make that makes sure that, at different speeds or in different types of weather, you can actually stop on a dime, especially if you're not going just on the road? For example, if you're going on the sidewalk or when you're making a turn.

You need to be able to stop much more suddenly?

Yes, because if you're going to move from, say, the bike lane into the sidewalk, or make any quick turn like that, those are the types of things that are more sensitive.

Dot, you're dismissed. You did great.

See you, Dot.

Good robot. That's so cool.

These are really good, by the way.

They're good.

I've never had Tayto.

Yeah, there you go.

Wait, you're saying this is like—

This is Ireland's—

This is like the—

This is Ireland's potato chips.

Lay's chips? No.

Yeah, it's the Lay's chips of Ireland.

Delicious.

Very good. I'm introducing you to all my culture. We're talking about—what are we talking about?

One of the things I'll never forget talking to you and Patrick about is that we grew up together in the years of building our companies.

You've always had the idea that Stripe is a technology company that happens to be in payments. It's not a payments company. You've always challenged, I remember, your customers to think about, “Don't think about just accepting payments. There's more that you can do.” I'm curious: What more can we do? Should we be doing things like crypto or stablecoins?

11. Commerce Becomes Agentic

Two topics that are on the minds of all our customers today are stablecoins and AI, specifically with respect to changing buying and agentic commerce. On stablecoins, I think maybe DoorDash is not where I would start. Two places where we're seeing tons of interesting stuff are, one, any cross-border use case where, if you want to send money to 100 countries, it's already the case that a stablecoin is a much better way to do that than anything else. That's where we're seeing a ton of adoption, where that long-tail coverage really works.

The second, and this is where Tempo, our new initiative that you guys were nice enough to work with us on, is coming in, is very scalable crypto payments. Obviously, agents are what we have in mind as we're developing Tempo. If you want to pay for your API consumption, or if your agent wants to be able to pay for things around the internet, you actually need a really good, scalable blockchain for that. Okay, so that's on the crypto side of things.

The other thing that we're thinking about a lot is, how does commerce change? Now that people's expectations are increasingly that they ask their AI for stuff, it feels like you should be able to ask your AI, whether that's ChatGPT or Gemini, or whether that's Siri or whatever your interface is to AI, “Let's just order again what we ordered on Sunday.” You should be able to make that natural-language query and have it take care of interfacing with DoorDash, payment, and everything like that.

I'm curious how you guys think about the interface changing. I mean, you just had a lot. Do you want to enable people to order within the AI apps? Do you want to have more of a natural-language input to DoorDash? How is that going to happen, given that that is presumably what people want in terms of the low-friction AI experience?

Well, I think what people want is definitely low friction, but I think people also want the products to show up on time.

You only get a right to do this cool UI stuff if the logistics is right.

Totally. The challenge with a product like DoorDash is you have to do the whole thing end to end, or you have to have great coordination or communication mechanisms. We definitely agree with you that there’s going to be a place in the future where, as these AIs develop their—whatever we want to call them—app ecosystems and relationships with companies like ourselves, they can do exactly what you just said: the jobs to be done in their regular life. I think that will be a source of top-of-funnel for us.

I think the hard part is making sure that we can actually deliver on the promise. As a cautionary tale from the past, I remember Google food ordering launched in, I forget, 2015 or 2016, and they allowed you to order delivery, for example, from various Google surfaces: Google Maps, Google Search, et cetera. While, on the one hand, it was low friction and they could send lots of traffic, they couldn't get the retention. This is where we're talking about consumer businesses, and retention being the name of the game. The reason was because they didn't know how to contact the driver, they didn't know if the driver couldn't find you, if something was taking a little bit longer, or if something was out of stock.

There's a lot of things that happen after the checkout button. I think this is one thing, back to what we were talking about earlier about what is the correct UI. It's not obvious to me that it's just a chat interface. There are all of these different components that have to be solved. The way I think about it is, okay, regardless of whatever we ship, whether it's on our own surface or other people's surfaces, are we solving the end-to-end job? If we are not solving the end-to-end job, it's not going to matter.

Last question. What is food that you can get in other places when you visit London, when you visit some other city, that you wish you could get in the Bay Area, but you can't?

Actually, I really like tea sandwiches. I was literally in London—

Afternoon tea?

Yeah, little cucumber sandwiches. I'm always looking for a healthy snack in the afternoon. Today, I had an awesome—

Healthy snack?

No, delicious.

Have you done the proper afternoon tea with the three-level platters and the little cakes and everything?

I did it once. I took my kids a couple of summers ago. That was an experience, I think, in part because it was new to me and my kids, and in part because I had kids with me playing with all the different things that they gave us. I think one of the things that's really interesting about some of the grocery stores, for example, in European markets, is that prepared meals are a massive focus, massive. We have some of that here in the US, to be fair. But in London, I just found myself finding that in every store, as I went to shop, you would see so many.

There's a few of these things that I find are so much better back home in Ireland, the UK, and Europe. Prepared food is one of them. A sandwich you get in Heathrow will probably be pretty good. A sandwich you get in JFK, don't go there. On a shelf in a shop. Some of the stores we have—the cookies, obviously—are much better back home.

There are some delicious cookies. The teas—there are a lot more choices of teas inside restaurants, I found. I think one of the coolest parts, and greatest privileges, frankly, that I have at DoorDash is that we get to interface with all the businesses of all these different wonderful cities and really get to meet the people who are building the GDP of those cities. It's not just the cool economic impact that they're having, but you get to find out about all their passion projects.

Everyone is passionate about this stuff.

Everyone's passionate about something. That's one of the coolest things about my job, where I don't just— You asked me a bunch of questions about different IP and process and things like this. I promise you there is that for every single business inside every city.

It's a cool job where you get to work with people who are incredibly passionate about what they do all day long.

Yes.

Well, Tony, thank you.

Thanks so much, John.