Sure, one may argue that you don’t have to spend it on delivery, but what we tend to see is that there’s also the macro trend of convenience increasing in the direction of greater convenience. And so if someone has some dollars to spend, it tends to start with the category of the highest frequency where they also seek convenience. And there’s just more opportunities to do that given that someone consumes food 20 to 25 times a week. So we are not seeing any of the activity perhaps that I think some other categories in commerce are seeing, but I think it’s really multifaceted. There’s the product improvements there’s the fact that people have to spend on food, and it tends to, therefore, lend itself towards the category where for eCommerce, it’s the most resilient and the highest frequency.
I mean, you can say this for the past 60 years, which obviously has a lot of macro dynamics as well. And you see that in only 2 out of the 60 years, I think, did the restaurant industry in the U.S., at least ever decline. And so I think it gives you a sense of the robustness of this category. And then I think the final point is we are still a lot earlier than people think in the category than perhaps our scale suggests. And that means that we are not where we need to be at in terms of where our selection of merchants are, whether it’s restaurants or new categories, the affordability of the program, the quality of the service and our customer support. We have to make improvements in all areas.
Ravi Inukonda: Doug, just a few more points to what Tony touched on, and then I’ll talk about the second question that you asked. We obviously have a lot of signal in the business. I mean we watch this very closely. When I look at the underlying cohorts, I mean they’re really strong. If you look at the performance of not just Q3, but actually the first 9 months of the year, cohorts continue to improve, and I look at retention, it’s very stable. Our users, in fact, at our scale are growing double digits. Order frequency, like we talked about earlier, continues to grow. So we are very pleased with the performance of the business and a lot of strength that we are seeing in the business. To your second point around GAAP profitability itself, as we’ve mentioned before, GAAP net income in and of itself is not an explicit target that we are managing the business towards.
Financial has always been how do we maximize the long-term free cash flow in the business. And what you’re truly seeing in the business today is every single line of business is becoming more efficient. We’ve dramatically improved the unit economics across the restaurant business, new verticals as well as our international business. I do expect every business to be profitable over time. And as and when that happens, the overall business is going to get to be GAAP net income positive as well. But that’s not an explicit target that we are driving the business towards.
Douglas Anmuth: Thank you both.
Operator: Our next question comes from the line of Brian Nowak with Morgan Stanley. Please go ahead.
Brian Nowak: Great. Thanks for taking my questions guys. Tony, I wanted to go back to one of the comments you made earlier where you talked about investing appropriately towards the stage of the product. So can we talk a little bit about grocery through that lens? So give us some examples of areas of investment that you made this year based on the state of the product that maybe you — I would say, okay, 1 year later, we can do more of this in ’24. I’m just trying to get an understanding for what are the areas you’ve pressed on hardest in investment in grocery? And what are the areas where you think that there could be more investment to come to really drive that business faster for longer in ’24 and beyond? Thanks.
Tony Xu: Sure. I mean — okay, we can take many flavors of this question because, I mean, this is probably something that I spend the majority of my time on, which is initiative and product reviews, because I do think it’s really hard to apply the correct judgment certainly correctly every time, depending on the phase of the project. And — but take a few instances, right? For instance, we really didn’t — one of the first choices we made on grocery was, well, what’s the right product to build. And I think there were many external beliefs of perhaps what you have to build in order to create a sustainable grocery business. But we took the view of what is the actual problems left to be solved, first and foremost, coming from our customers.
And one of the things that we heard, for example, was that while what’s really frustrating about grocery delivery is that the items that either perish the fastest or that we consume as consumers the most frequently. So think about your berries or your milk or your eggs or your cereals or your coffees. It’s really annoying in the middle of the week to have to go back into the store and somehow go and get it. Whereas for other types of shopping behavior, if I’m already out on the weekends anyway, and I’m buying other things for the week, maybe it’s less of a big deal. And so the first product we’ve built was actually to solve a smaller basket topic use case, which really introduced ourselves to consumers and helping address a problem we heard.
And then I think had the byproduct of introducing ourselves to grocers, too, because this was a use case that they hadn’t been investing that much into. And then I think it surprised everyone how large that opportunity occurred. So that’s like one version of this, which is before we kind of just look at what everyone else is doing, that’s first and foremost, start by building the correct product. And I think building the correct product has a lot of implications on do we use engineers and business people and other functions correctly and therefore, get the most out of the team’s potential. I think there’s a version of that shows up. I think number two, another example would be, while we didn’t really spend much on marketing when it came to building the grocery business.
Now Ravi echoed some of the kind of reasons why because, well, one, we do have the largest audience of local commerce customers coming to our app shopping as well as just coming to the app and including those that don’t shop. And so we had a bit of a benefit there. But at the same instance, we also didn’t want to invest in marketing because the product wasn’t good enough yet. And so I think those are two, I guess, examples I would call out where you have to make the right call on how to make the most efficient use of your resources. And first and foremost, for us, it starts with solving the right customer problem correctly. And whenever we see something that’s not being solved correctly, and we think there’s an opportunity to build something 10x better, we are going to go for it.
And then the question becomes, well, is there an efficient way to grow, then we are going to think about that and if we have an advantage there. And then finally, what’s the path towards building a very large cash flow generating business that will maximize total profit dollars in the long run. And I think that’s kind of the sequencing of how we thought about things. I think so far, we’ve made good decisions in grocery and a lot of these paths, I think you’re seeing some of the combination or the middle of impact of the results in this quarter. But we’ve got a long ways to go and we have to keep making good decisions.
Brian Nowak: Great. Thanks, Tony.
Operator: Our next question comes from the line of Michael Morton with MoffettNathanson. Please go ahead.
Michael Morton: Thank you. Two questions, if I can. One, start with for Tony and then the second for Ravi. On the advertising business, kind of following up to the prior question. How do you feel about your product’s ability to serve some of the larger enterprises that goes for the QSRs, who kind of demand sophisticated solutions, but also now moving into the grocery business? In our conversations with CPGs, there’s a concern about incrementality in advertising for grocery retail marketplace — retail media networks because the fact that there’s not a lot of unit growth there. So whoever can win the incrementality measurement kind of like wins the market for advertising. So curious to how you think about your product there. And then just second for Ravi.
