Panel: Hear what Uber, Instagram, Coinbase and Super.com think about building Super Apps in North America
A Super App is an app that offers multiple services within one app, the best example being WeChat in China. That will not happen in North America — however we have “Theme based Super Apps” — multiple services and features under a unifying theme!
Last month I moderated a panel with Uber, Instagram, and Coinbase. Watch the full panel on YouTube or see some highlights below:
Watch the full 44 minute panel below:
Highlights from our panelists:
Loren Kosloske (Growth & New Verticals at Uber)
Believes there are 3 consumer-centric focus areas guiding their execution strategy around mobility: Right to Play, Operating Model, and Product Market Fit.
8:17 — Focusing on right to play, operating model, and product market fit
17:58 — Bridging services through personalization
25:37 — Targeting a core cohort to prioritize new use cases
Sanchan Saxena (ex Head of Product at Instagram)
Believes we have to recognize that customers across the world are not the same and they have different needs and circumstances, which drives how technology products evolve.
9:50 — Testing, failing, and learning to discover new use cases
23:40 — How to maximize user adoption and cross-selling
28:27 — AI as a tool to enhance consumer delight
Sid Coelho-Prabhu (Director of Product, Coinbase)
Talks about allowing their users to discover apps across the entire blockchain ecosystem will help them target the most effective use cases and build out their platform to support them. Watch some key segments below:
20:25 — Why building native interfaces for other apps is more efficient than competing with them
26:40 — Assessing activation rates and knowing when to kill development
42:15 — How to set up cross-functional teams to build the best products
Questions from the audience
Question 1: “Why are all the super apps in the US centered around one topic and why can’t they be like the WeChats and Grabs of eastern economies?” [Jump to 36:18]
Question 2: “What is your experience setting up teams to build a Super App — do you have separate teams building out each vertical or can you get away with one team working on your core product and one team testing all new use cases?” [Jump to 40:30]
Full Transcript Below
Hussein: Thank you everyone for joining. Today we are going to be talking about theme based super apps, and I have some fantastic guests over here with me to talk about this. But before I start I just want to give sort of a two minute introduction. So if you’ve read a blog post which I published recently, you’ll see that I believe we are not going to have a WeChat over here in North America. When you look at WeChat, you see something that encompasses chat, bill payments, and sort of anything and everything. It’s sort of like an operating system for all of China. Now I believe that in North America we’re not going to have something like that. But what we are going to have is what we call theme based super apps. So they’re essentially an app that starts with a core product and builds around that and it’s all built around a specific theme. And that’s where I believe we’re going to go. So, just a quick visualization over here.
We obviously have someone from Uber joining us, but you’ll see they built an app around mobility and a whole bunch of things around, you know, getting anything going anywhere, and we’ll talk about that soon. At super.com, we’re building a super app around savings and trying to help people save across the board. In a little bit of a different example, you’ll see Google Maps has built a super app around location. So in each one of these cases, that core theme lets the customer know why they’re opening the app, the reason they’re opening the app, and everything is built around that as a core theme.
And that’s the type of super apps we are already seeing in North America, and we believe it’s going to continue to grow. So what I’d like to do is introduce our three panelists who have come here to join. I will start with Loren, why don’t you introduce yourself.
Loren: Yeah, hey everyone and thanks for joining. Loren Kosloske, I’ve been at Uber for a little over three years, predominantly sit in our Seattle office, but I focus on kind of our new verticals and new products predominantly within the travel and airport space. So anything about using Uber outside of your home community, that’s predominantly the space that I’m thinking in. And I sit on our mobility side of the business so typically the app that has the black icon, not the one that has the green for Uber Eats, but there is some interplay as we’ll talk through later today.
Hussein: Awesome. Thank you, Loren. And then we have Sid from Coinbase.
Sid: Hey folks, I’m Sid, I’m a product manager at Coinbase. I’ve been at Coinbase I want to say, almost six years now so I’ve been here a little while, watching the company scale and helping grow it. I’ve been working primarily on the Coinbase wallet product, our lead wallet, which is our self custody crypto product that lets you store your own crypto anywhere in the world, it’s open to all customers everywhere, and then use your crypto. So we’re very focused on helping customers discover things to do with their crypto, including through things called decentralized apps. And over the course of my career at Coinbase, I’ve also worked on products like Coinbase Commerce, on Web3 and you know more broadly across DeFi and NFTs, etc. And in previous lives I’ve been an investment banker on Wall Street, I’ve worked in the Indian government, and I helped launch one of India’s first robo advisors. So I’ve had a few different experiences in tech and financial services.
Hussein: Then we have Sanchan who was previously at Meta and worked on Instagram, but I’ll let you introduce yourself.
Sanchan: Yeah, happy to. Hi everybody. Back in the day I was at Instagram, but Instagram was much earlier. We had $0 in revenue back in the day and you can imagine how early that was. And I worked on many things that Instagram did become for example Instagram shopping, Instagram creator tools, and through that journey we tried to also build our version of a super app which is, if you come to do something in Instagram, how can you do other things while in Instagram as well. So we will share some tidbits around that. And after that I was the head of product and general manager at Airbnb and I spent the last three years at Coinbase as VP of the retail product as well. Excited to be here today.
Hussein: All right, well we have great people here and only 45 minutes. So we’re going to jump right into it. Let’s start with Loren. So you know the theme of mobility, going anywhere, getting anything. Can you tell us how this has evolved over time. Where did Uber start, where did UberEats start, what are you adding and how is it all coming together.
Loren: Yeah, I think first, like it’s more to know like Uber started back in 2009 so it’s been a long journey. It’s not something that you do right away, but if you think about it, we started as UberCab in March of 2009. A few months later we rebranded to Uber, but then fast forward five years later and we launched UberEats, then fast forward another five years and in 2019 back in May, we brought UberEats to like the airports and venues experience so today you can utilize UberEats post TSA or airport security at Tampa Bay Airport as well as Toronto, but you can also do UberEats at Yankee Stadium directly in your seat so we’re continuing to go deeper there. We’ve gone even further in July of 2020, we brought groceries to the UberEats platform and have continued to bring that out so not only can you get food, but now you can also get your groceries that you need every day.
When COVID hit we continued to innovate throughout that time so actually in August of 2020 on the mobility side of the business we launched boats in the United Kingdom, brought rental cars in April of 2021, package delivery during like Q4 of 2021 as well. And they also spun up UberPark back in December of 2021. That one is something that we were no longer doing but it was a great little thing. The big thing to know is test and iterate and don’t be afraid to fail was like a big learning that we did throughout COVID. And then, in May of 2022 we brought the ability to book charters, so like limos, like the longer limos charter vehicles for like bachelor and bachelorette parties various activities there and for weddings. Then in August of 2022 we brought long distance and short distance coach buses in the UK, launched flights a few weeks ago, and then also announced our launch of boats in Greece.
Just recently we also announced our launch of car share, which is currently live in Australia, and more to come here throughout the next several months. So overall we’ve done a lot. If you look at that journey, you’ll note that it took a couple years for us to really accelerate everything that we’ve been doing. First we started with a very solid core product and wanted to continue to expand and make sure, at the end of the day, when you want to go anywhere get anything that you can utilize Uber to do that consistently across all those product experiences.
I think going forward I would think of that about, you know, wherever and whenever you are typically utilizing Uber to get things or to go to and from all those services are technically fair game for us to think about as future opportunities. I can’t tell you specifically which ones coming next it’s really a prioritization question. But I thin, if you were to put yourself in my seat and think through how you would prioritize it’s probably the same approach that we’re thinking about it. I will say too, while this topic is super focused on super apps. The super app is actually a topic we don’t talk about internally at Uber, it’s not something that we’re focused on and what are we missing. Really focused on using data to drive what products make the most sense and how we can continue to deliver the magic of Uber within those spaces as well.
Hussein: Amazing. Thank you, Loren and there’s a lot of people who don’t really talk about it as a super app so even if you talk to Google and the Google Maps team. They don’t necessarily call it a super app, but it is a way as an industry term for us to understand how you’re building things around the core theme. I do want to switch over to Sanchan, but just one more question Loren, you mentioned I believe you said park is something that you decided not to continue with how does the company decide when a test is, you know, working, and they want to double down on it versus when a test is something to shelve.
Loren: Yeah, I think we kind of look at it, regardless if it’s successful or not across some similar fields. One is, do we have the right to play in that consumer mindset. So what does engagement look like. That doesn’t mean just are you transacting and like actually spending money on it. But, are you clicking into the experience, going, coming back again and again. And what do those interactions look like not only for that core product but also at a platform level. The next one that we want to look at is the operational model, we want to make sure that it’s validated both from the user perspective and the business. So that can mean, do we have the right partner. Are we setting it up for back end capabilities to scale. What are we learning throughout that to continue to think about like what it takes to go from maybe a small city to a global experience. Then the last big ones are really around, you know, product market fit, it’s a big mental leap to think about I’m going to take an Uber to go to or from dinner to now be like, oh wow, where am I going to park my car there and trying to change that. We learned a lot out of it and it was super fun. It was a very different space for us to think through.
Hussein: We’re going to jump into bridges and how to get a customer to think about one product and then the next product but I do want to flip it over to Sanchan. Sanchan, you were at Instagram early you were part of the team that was talking about bringing payments and commerce into Instagram. Tell us about that journey.
Sanchan: Yeah, happy to. I think let’s rewind back to 2012, 2013 and 2014. Instagram was a one trick pony. All you can do is add filters to your beautiful photos and I think to Loren’s earlier point, you got to build an incredibly successful first use case before you even earn the right to talk about super apps. You’ve got to have something that is so amazing people fall in love with it and they’re using it tremendously and then you can think about okay what adjacencies what other things I would like the customer to do in my app, but that first right is really really important. So that was our right, and then we started thinking okay people are spending time what other activities they’re doing. One example of this was we noticed that customers and businesses were using Instagram for ways that we never thought we designed it for. If you’ve ever traveled to Thailand, or any of the southeast Asian countries, many businesses never had websites so they would create an Instagram profile and put a photo saying this flip flop this T shirt is available for $15, and people were like wow I love that. And they’ll say okay message me your PayPal and your address and I’ll ship it to you.
Now I always say you know the customer leads the way what I mean by that is, sometimes your customers are showing you things that they want to do with your app, and the real genius for product to then figure out is is that the right thing to invest in or not. So we decided that it was worth investing in because we saw engagement coming up earlier. And to Loren’s earlier point we felt, do we have the right to play there. Meaning Instagram was already the visual catalog of the world. You know you could actually open it up, and you can see all the visual stuff on Instagram so we did feel that when it comes to visual commerce, not when you buy toilet paper where there is no visual commerce associated with it, but when there is design aesthetics, you know, feel look, etc, that’s visual commerce and that’s where we had the right to win. So we started adding that and Instagram shopping became a huge success. And then we said okay this is by the way this is just an example of a use case that made us into a super app but again we never talked about a super app all we talked about was what other problems customers have that we can uniquely solve. And then we started thinking what else can we do and this is what I want to talk about a failure use case as well because I think oftentimes we will not find all use cases working in the app and here’s an example.
We have this assumption or hypothesis that you are sitting down at a restaurant in New York City. What do you typically do you pull out your phone, you take the photo of that beautiful food that you’re eating, and you post to Instagram. In fact, you would tell your friends, don’t touch it. Let me first take a bloody photo before we even eat right and that’s what people would do. So the natural conclusion from that point onwards you could potentially draw as a hypothesis is that well, the person is already there. They’re already eating. They’re already taking a photo they should be able to pay right there using their Instagram credentials. And you have, we had hundreds of millions of users around the world, whose credit card information we as Facebook had. Right, so we tried that. And that didn’t take off. Right, because the leap from going from what they were doing, putting the phone down and then actually eating and then remembering to open Instagram to pay again was not the right leap. And that use case never took off.
So I think it’s important to recognize in these cases that while it’s important to do everything under one roof for the customer around the theme, but within that theme as well there might be adjacencies where you will have success faster, quicker, deeper than adjacency where you will take forever to get to. I’m not saying Instagram cannot solve the payment use case. Probably they can. But at that point in time, that wasn’t the right adjacency to pursue.
Hussein: Thank you Sanchan. Sid, I know Coinbase does talk about a super app. Brian Armstrong was just in the news last week mentioning a super app. So maybe for the newbies out here, a lot of people think of crypto as I’m just going to come in and buy and sell crypto. So could you expand on what other use cases, what other consumer value there is and how Coinbase is thinking about that.
Sid: Yeah, for sure. So, you know, if you look beyond the trading of crypto, totally get that, you know, everyone’s first experience with crypto is usually just buying their first Bitcoin or buying Ethereum. But if you look beyond that, like our vision, and this is not just at Coinbase, there is actually a huge ecosystem of builders in crypto. And what we’re all trying to do is create real utility and use cases for crypto. So once you own your crypto, can you then go and use it for X, Y, and Z different use cases. And so one big example is just financial services, right? What crypto offers is for the first time, anyone in the world can onboard and own the same currency as anyone else in the world. And they can send it seamlessly.
So you can save your money, you can earn interest on it, you can send that money around. So in that sense, it’s offering something that could never have been done before. And that’s really where I think the core use case, the first like super app use case comes in, which is kind of being part of this global network of people that are all owning and storing this currency, they’re able to send it amongst themselves, etc. And then on top of that, you have a bunch of applications. And so this, I think the second kind of journey that crypto is going through is the application phase. And so, you know, with blockchains like Ethereum, you can actually build any program that does anything with crypto. And so it’s like this infinite space, anyone can build an application. And anyone who owns crypto can use that application anywhere in the world. And so what we I think our conception of a super app potentially would be that everyone starts off by buying some crypto. And then we provide a range of all these different use cases that you can use with your crypto inside the app. And this includes financial services.
This includes things like buying NFTs, buying audio NFTs, storing your crypto identity. So you have a decentralized identity that you can use in any of these different crypto applications. All of that really comes together inside one app. And that’s I think how I would see things playing out.
Hussein: Got it. And that actually goes back to what Sanchan said about first having that core use case that gives you the right to build on top of that. So obviously, Uber with rides, Instagram with photos and Coinbase with the buying and selling of crypto. And then once you have that core base, you can build off of that. In our case, we started the business around hotel deals. And you know, we have 10s of 1000s of customers coming in every day buying hotels at great rates. And now there’s ability to say, hey, you just saved a bunch of money on your hotel, what else can we help you save money on and what can we build around there. So we’re constantly testing other use cases.
I want to talk about some of the core principles on building a super app. So I’ll just share my screen for a second over here and I’ll talk through these six. So the first one is building a bridge between services. And that’s, you know, when a customer is using one service, how can you bridge them to get to the next service to this friction free activation. So even when you do bridge them, how do you activate them, how can you make it as easy as possible for them to use and try the new service.
Three is use case frequency. So if you’re going to build a super app, you want to have an app that people are opening up all the time, not something that people open up just once a month. There’s a brand promise. So every time you’re trying a service, it should live up to that brand promise and have a consistent experience. Number five, which is one that a lot of people don’t talk about is expectations on usage. We simply cannot expect every customer to use every service. And if you believe that they’re going to use every service, then you’re living in another land. And then finally, number six is AI. So we’ve all seen the impact that AI has had in a short period of time and we know what it’s going to let us do in the future. And one of the criticisms of super apps is that there’s just too much going on too many services too many offerings, and I believe that we can use AI to leverage that as an accelerator for super apps. Let’s talk about bridge between services.
So that’s one of the hardest things which is you come in for one service, and you need to get the customer attuned to or used to using another service and having a bridge from one to the other and not just being like here’s a grid of all the things you can do. So I’ll open it up to the panel. If anyone has any examples they’d like to share.
Loren: So you’re right, bridges are super important. They’re also, it’s pretty easy actually to throw a bunch of products in an app. You can think about the Uber app. Two years ago even pre coded. It was we had these little buttons that would have various different types of products. You had ride, then you’d have, you know, like we have a rent one now we have a travel one technically like there’s various different ones but really for us bridges requires personalization to be really consistent that it’s a hard thing to do. And it’s one thing we really just started playing around with a lot. So if you think about what we’re doing with Uber travel, for those of you outside of the UK if you click on the travel icon in the Uber app they’ll ask you to link your email and will ingest third party travel bookings and help automatically nudge reserve or rental cars depending on where you’re going. And that’s really where we’re starting to play around with the bridge is knowing when and where you’re going. How can we make the best recommendation for you for products that are maybe outside of our core on demand offering.
However, as when we’re thinking about our bridges on our end, we’re also super focused on making sure that that value prop for that bridge works for, you know, our drivers on the mobility side, our couriers on the UberEats side or even the restaurants or grocery providers. So the bridge is one that’s super top of mind for us. So I think the fact that it’s listed number one for you, it’s our priority number one as well.
Hussein: For me, by the way, for Uber, I think one of the best things you’ve done is pretty simple. It’s when you’re ordering food from a restaurant, you’re prompting to order something else from somewhere else like, you know, also get some something from the convenience store while your driver’s already on the way.
Loren: Right. And the first time you do that, now you’re like, oh, wait, maybe every time I need something from the convenience store, I should use Uber.
Hussein: Right. But for most people, they wouldn’t think about that. They’re just like, oh, I’m just going to use Uber Eats to get my restaurant food, not necessarily my groceries or my convenience store items or flowers or anything else.
Sid: I think, you know, I’d reflect that we’ve probably taken a slightly different approach on wallets to what bridges to build and where and when. And the way we do that is we actually work a little bit backwards. And what we did first was we opened up the platform where users can discover anything that’s going on in the crypto app ecosystem. So there’s a bunch of crypto app developers, they’re building stuff anyway and putting it on the blockchain. We let our users access all of them. Then we see what’s succeeding, like what are the sparks, where are people organically going, and we’re not necessarily directing them there. They’re finding about the stuff on Reddit, on Twitter, through friends, etc. And this app platform is totally open, so we don’t pick who’s succeeding, who’s winning. And then when we see something is working really well, we don’t try to compete with it. Instead, we build a more native interface for that application that still uses the underlying business, but just provides a more seamless experience, like there’s a dedicated button to it in the app. There’s a much better flow that’s better than using a website because these apps all start as websites and then we graduate them to very clean, native, built-in experiences.
So the examples I can give you are Dex trading, right? One of the first big successes in crypto has been these decentralized exchanges like Uniswap, for example. We found a bunch of people three years ago were going to use these apps and connecting their wallet. We brought that into Coinbase Wallet in a native experience. It still uses Uniswap and all these other Dexes, but users have an easier bridge to access it. We did the same thing with staking, right? Users are going and doing this anyway.
We’ve done the same thing with NFTs. People are going to all these marketplaces. We bring that into the app. So I think it’s something that’s a little bit unique about crypto in that it’s permissionless. And so people are building all these products anyway. And so we can then as a super app, we can pick and choose what to bring in in a very first class, first party way when the time is right. And we get good signals before we make any firm commitments to build something.
Sanchan: Yeah, I think these are great examples between what Loren and Sid are saying, because if you think about it, there’s a tale of two different stories. Uber, first party experiences owned and operated by Uber. Everything is at Uber. Coinbase, not like that. Coinbase super app strategy is more like the Apple strategy, but you know these apps are already there. How can I make it easy for you to access? And when it becomes successful and maybe there’s a better way to do it, they build it natively. So I think it’s important to recognize there are two different tales to how to build your super apps, not just one way forward.
Hussein: That’s exactly what I was thinking as you were speaking. And we fall more under the Uber scenario where we’re managing the experience. If a third party wants to come in and offer a savings use case, you know, we want to help your customer save on gas, we want to help your customer save on groceries. We’ll work with them, we’ll channel the experience, we’ll work on the API integration, and we’ll try and make something work. So, where Sid, you have this great benefit that you can just have this open app store and like, you figure out what people want. Loren and I have a lot more work to do because we have to assess each one and build each one, or at least a smoke test for each one, before we can learn and see what customers actually do. That’s a great, great position to be in.
Let’s talk about the expectations on usage, which is one of the core principles. Again, I don’t think we can expect every customer to use every service. So, wanted to see what a cross-sell rate you could hope for. Like, when you have a customer using one product, what percentage of customers do you expect will use more than one product or more than two products? If you have some numbers you can share, that’d be fantastic.
Sanchan: I don’t recall the exact numbers, but I can share some ballparks here. So, Airbnb, when I was there as a head of product / GM, we went from a home sharing company, where you can only book homes, to also launching experiences, where you can actually book local experiences while we’re doing that. And it was a similar journey where we were like, okay, you’re booking a home anyhow in New York City or in Paris. What are you going to do over there? Well, you might want to go entertain yourself.
So, here are a bunch of experiences that you can do that. And I think there was always this question of what is the attachment, or that’s an internal technique, but the idea is like, if a user were, if you had 100 products, how many of the products will the user use? And roughly, I think it’s kind of interesting to note that you will not expect a vast majority of your customers to use in the beginning for sure. So, for example, for us, it was like less than 10% would book an experience in the beginning days. And our goal was to continue to increase that.
But the way to do that was not force feeding just through notifications, notifications, notifications, but rather realizing, okay, what are the barriers to adopting this new experience? And what other kinds of things can we latch on that will appeal to the next cohort of users from the initial 10% as well. I’ve heard anecdotally, at Amazon and other places, if you get an attachment rate of five or six, meaning if you use five or six other products, and Amazon has a lot, that’s a pretty good success. But that’s an anecdotal thing, and I can’t vouch for it, but that’s what I’ve heard anecdotally.
Loren: Yeah, I’d say we’re thinking in the same way, like each product differs. I think even if we look at like our mobility and delivery products fundamentally we want everyone to be 100%. We’re not there. We have things like Uber One to help accelerate that and think about the cross platform benefits, but with the things we’re doing and travel, if you’re buying a flight on Uber, I for sure, 100% want you to try and book a ride to or from, or once you’re on the way to the airport also utilize Uber Eats post security to skip the line. But realistically it’s not going to meet everyone’s needs.
So with flights you can book direct, you can go to other OTAs if you want. So it’s thinking about how do we still bring value to the platform with these other services, regardless of where you book but realistically we aim for 100 but it changes for each use case. And we always focus first on who are the people using Uber Eats and Uber let’s start there as our cohort, and then work towards the latter tail in terms of prioritization.
Sid: I’d say that, you know, I don’t have hard numbers to share but I think maybe in terms of informal heuristics if I have a certain offering inside the app. If after a certain period of time we haven’t hit a 10% activation rate where I say 10% of our customers are using it at least that to me is like okay we need to start having a conversation about killing it.
If it’s hitting something like a 30 to 40%, I think that’s healthy. Not every individual app you offer inside a super app is going to have 100% adoption. Right, if it is then really it probably needs to be its own app altogether. But apart from the core offering in our case of storing crypto for example every every user should do that. But then beyond that for certain use case, if like 30 to 40% of users are getting value out of it then you can say that for every different app I think that’s like pretty healthy.
Hussein: Yeah, thank you so it sounds it sounds like we’re centering around that like 10% number is somewhere you want to get to, you know, if you get to 7%, 8% that’s pretty good and 10% is okay but like, ideally you want to get that much higher. And so you want to get to 20% or 30% but I don’t think anyone can hope to get to 100% so that’s it and obviously it depends so much on the service you’re offering and on the customer depending on the bridge and a whole bunch of different things right so even when we think about it, you know there’s certain products like a cash card or gas that people are more likely to cross over to, but something like insurance is much harder, because that’s a much more considered purchase and takes a lot more time and there’s a lot of complicated offerings before you can make a decision like that.
Okay. We have about 15 minutes left and I want to talk about something very important which is AI. So you know we’ve seen obviously the big impact it’s made in a very short amount of time. Some companies are already using AI aggressively some companies are only in progress. Some plan to use it. So we’d like to talk about AI and it can be around the super app and how you’re using it to customize the experience and decide what offers to show to the customer or it can just be your usage of AI in general.
Sanchan: I think maybe we can break down the wide spectrum of AI the word AI into smaller chunks as well, because personalization has been there, like machine learning based personalization has been there for 15 plus years you know Facebook was the first one to do that. And I think at one layer everybody needs to personalize the super app experience that’s one of the basic layers, you can’t just throw a hundred things at me and expect me to learn every hundred things. You gotta tell me how that is relevant to me, you should be able to understand who I am, and in context, offer that service. The context piece is really important and you were talking about for ordering food and it prompts you to say, you should also get something delivered from the store, that in context personalization awareness of the users activity etc is the basic thing we all need to do. But then you take it up even further and the other thing that I’m very excited about is, how can you then take this to the next level where the user actually has all these experiences, pretty much serendipitous.
Let’s take Airbnb as an example and again I’m going to cook up new services that don’t exist and just to make my point easier to understand, you know you are in the Airbnb home, you’re sitting over there watching Netflix and how many of us hate logging in with our own Netflix credentials right and we actually just go over there and I could actually be able to air stream everything from my app on that thing. And now I’m actually able to buy movies and watch those things. Now that is an example of personalization over time that is using anonymous data but at the same time it’s giving you a seamless experience using personalization at a next level. So I think that’s the spectrum we’re going to play through. And I’m very excited to see where the next level comes in, but also recognizing that personalization and offering in context experience has always been the case in many of these so called super apps for very many years.
Loren: I fully agree with what you just said. Personally I think generative AI is super interesting but it’s a bit overhyped at present. Like, it’s something we’re always constantly evaluating. To be honest we’re playing around with it quite a bit with hackathons and things like that. But ultimately the core question for us at Uber is like, how does this provide value to the end user, what pain point are we solving for. So if I had to invest a dollar in one place or another right now I’m skewing personalization because we can see that a little bit more versus connecting and paying for tokens and having to then solidify that and think about it when we bring it to the Uber app and if we bring it to the Uber app, it’s gonna be focused on what pain points are we solving for the consumer. That will be what ultimately defines where and how we use it.
Sid: In terms of AI, you know, we haven’t really been doing a ton of stuff there but it’s always something we think about as product managers generally and builders, and I think the areas that seem interesting are how can you provide a more human and personal experience when it comes to things like problem solving, things like customer support.
I think many initial attempts at things like chatbots across the entire customer service industry have not been as good as they could be when you see a chat GPT type bot. I think that’s a good way to think, could you provide some kind of concierge inside your super app that is really powered by AI, and this conversational technology that’s kind of pretty groundbreaking 10x versus where it used to be. I think that’s a great thing in the crypto space specifically as AI grows and becomes so easy to do and it’s not related necessarily to an individual app that we’re building.
But the fact that AI lets people create deepfakes and create so much content that you don’t know the actual origin and source of it. We think that crypto technology could be very helpful in providing provable verified creators who actually made this who actually said this and you know this interview was produced on a verified device that you can sign off on cryptographically. And that’s the kind of use that will help with things like disinformation and fraud, and things like that. So that’s almost aside from anything you build into a super app but I think greater impact on society.
Hussein: Yeah, and that’s great and that’s definitely worthwhile mentioning my initial thought process on using AI and super apps is just to improve personalization and personalization has obviously always been there. So we always want to be as personal as possible, we want to try and get the right service to the right customer at the right time. But even internally as we’re building there’s a bit of a, hey, should we offer the customer this, should we offer a deal on gas, should we present them a notification for the next greatest hotel deal because they’ve had a search recently. So there’s all these inputs, and as product managers we’re trying to do things to try and say hey we should put this here, we should have this much of the home screen for this or for that, and this many notifications per day, but I imagine there’s going to be a point where AI will be able to make better decisions where we can tell that to the AI and say hey here’s the ultimate golden one.
So we’re not going to be like, hey, we’re going to put this in this amount of retention or this amount of revenue or this amount of usage, and let the AI decide what notifications send and how to customize the home screen, so it doesn’t feel like well there’s too many things going on here. It’s just the right amount to sort of maximize conversion and minimize people who are getting annoyed and blocking emails blocking notifications.
Sid: When you think about like setting up an ML model for a given problem. It requires a lot of grooming up front, a lot of constraints and parameter setting, you know, it’s not something you can just say hey I want to do this you got to work with a team to get it set up over the course of a few weeks at least. And I think you’re right in that hopefully the AI can be more autonomous in figuring out what problems to solve and start solving them. And maybe that’s the next level and helping business builders do things directly. Instead of having to work with say, a whole ML team that needs to set this up. It’s a little bit more of an interface where it democratizes the ability to set up personalized experiences inside apps by letting you do it more easily like the AI is able to take instructions and human friendly language and set and do things for you.
Hussein: That would be amazing. It’s always the dream to be able to get more of these experiments live without requiring engineers. So, that would be fantastic. I want to actually switch it over to a couple of questions which came into the Q&A. So the first is just more of a statement. We have Fahad, who mentioned that we at Zillow actually talk about super apps. So they call it that and they want it to be accessible publicly. So just wanted to throw that out there and maybe we should connect after this panel. But that’s interesting. I never thought of Zillow in that way.
We have a question that says all the super apps in the West are centered around one topic, Coinbase — crypto, Uber — mobility, PayPal — finances, vs. the OG super apps like WeChat and Grab that touch dozens of verticals at once. Why do you think that is, could it be responding to demand in the West, rather than creating demand in the East.
I can sort of take a first shot at that. And mentioned at the start of this webinar, WeChat was just made at a different time. There were a whole bunch of things that had to exist in order for WeChat to be able to grow in terms of the powerful mess of the smartphones, the government regulations, the infrastructure over there that allowed for that to be what it is. And now it’s become so central and core to what everybody does. There’s no real replicating that. Grab is also an interesting use case, not as extreme as WeChat. But they’ve had sort of an interesting set of conditions that allow them to do that.
I think in North America, what you see is that it’s actually not so difficult to switch apps. Everyone has powerful phones. Everyone’s been sort of accustomed to be like, hey, an iPhone, one phone, and it’s like all the apps you want, and you can kind of quickly switch between apps. And then you’ll see companies who say, okay, I’m going to focus on a particular use case. And I’m going to go and try and expand that and sort of grow from there. Whereas in the East, they do a bit more of, let me focus on the customer. So I have a bunch of customers, let me understand who they are, and let me see what else I can do to help them. So it’s a bit of a philosophical thing and a bit of an infrastructure piece. But I’d love to pass it to Loren, Sid, Sa chan, if you have any thoughts on that.
Loren: I agree with everything you just said there. I think the regulatory piece is definitely like a big one as well. I think what’s also different, though, is if you look at the super apps that we’re talking about today versus those that are more so in the East. They’re getting much more complex here, I think, when we think about bridges, AI, and how we’re like trying to weave it together. It’s very focused on the user experience, I feel like more so in the West than it is in the East. So I think like, yes, they may have gotten bigger, but I think we’re really starting to think about it differently and have higher barriers of entry to thinking about making sure that we’re weighing the consumer and have the right deploy than you do in other parts of the world.
Sanchan: Plus one to everything. I think when we did a lot of user research back in the day, what we understood was that the customer behavior and mindset is also very different between the two markets. So I always remind people just because you build a successful product in America doesn’t mean it’ll work in India, and even in India doesn’t mean it’ll work in China, right? And plus, we have to recognize we no longer have one uniform customer in the world. They all have different needs, and they all have different circumstances. And I think we talked earlier about the circumstances of India and China where the mobile phones were not as powerful and there was a necessity. So to say, you know, all innovation happens because of a necessity and the necessity over there was, man, I can’t download 100 apps. You know, I just cannot I don’t have enough space on my phone, and the necessity in America wasn’t that where we had the iPhone and other phones that allowed us to download many, many apps to I think culturally.
I’ve also noticed that the customers are more comfortable with a barrage of information in India and China and other places, you know, versus America. If you ever walk around their version of New York Times Square, you know, there’s only one in America, every city is like that in those countries. So the customer expectation and the comfort level they have with a barrage of information, how they manage that is also very different, and the annoyances that come in America are also different. And last but not least, I think we touched upon regulatory. I think another way to frame that would be there’s only so much a Facebook or Google can do to become a super app before a lot of other things start to happen in countries in which they operate versus not.
So I think keeping all of those things in mind, I think there’s no right or wrong answer. The answer is what is contextual to the geography and the customer base you’re serving. So I don’t imagine another WeChat coming out of the West. Something new will come up that neither is WeChat, nor is a singular use case app. We just don’t know it yet.
Hussein: From Cody, what’s your experience in setting up teams to build a super app. Do you have separate teams for each individual vertical you launch. Could you get away with one team working on your core product and one team building and testing multiple verticals.
Loren: So for what we’re doing in like Uber travel, for example, we have a team for train and coach, a team for flights, a team for our bridge smart itinerary so we keep them separate in terms of like separate product managers, separate engineering squads, but overarching vision is tied to all three. So we do when we go through our planning cycles and our sprints, we’re making sure that we’re talking and everyone’s building in the same direction, but in terms of staying focused and providing best in class capability, not only on that per product level, we’ve just found that a little bit of separation is good.
Sanchan: Yeah, I’ll plus one that and I think one thing to remember is a super app is an outcome of building amazing individual products. What I mean by that is if you’re building Uber Eats, you want that thing to be kick ass the best eating product out there. If you’re building gas, or helping people save on gas, you want it to be the best in class on its own. And guess what we have amazing distribution because of travel or whatever the other thing is, plug this amazing new idea into that distribution. So I think it’s important to have single threaded owners as we call it at Coinbase, Amazon, and other places so you wake up every morning and want to become the best ordering food app, best decentralized exchange, best gas saving app, and then leverage the distribution you have with the overall app and plug that into that.
Sid: I would just echo exactly what they said, the way we do it on our team is all the individual verticals have a team that’s really dedicated to making, whether it’s NFTs or defi, making it great and they make it the best possible experience. Also having the technical competency to really know the market, the technology, all of that. And then we have other smaller pods that are horizontal and they are responsible for things like the app architecture, engagement, gamification discovery and their responsibility is to pull all of these products into really cohesive experience to get the right bridges built like you’ve said before. Get the right introductions at the right time the right notifications to support all of these products to make them all successful. And so I think setting up that matrix and the incentives for each team that’s really important and that’s the general structure that I think works pretty well.
Hussein: Again, I agree with all three of you. We do the same thing, so we have dedicated travel people, we have dedicated people working on the on the super pay card because that’s a whole thing on its own. And then we do have a team that’s doing the horizontal stuff but also experimenting and running smoke tests and trying to say you know what could be coming next what else do our customers want to save on that’s combined with customer research and doing all these things to try and be what comes next. If you start to see something really take off and start to see things really do well, you can see about splitting that off into a separate vertical so you have a separate team working on that.
Okay, we are pretty much out of time. Wanted to really really thank all three of you, really appreciate it. Like I said, we are seeing super apps in North America they’re just very different. So having a theme based super app is what we’re seeing, and we’re going to see this continue to grow. Thank you, Loren. Thank you Sid. Thank you Sanchan. Really appreciate your time and hopefully this was helpful for everyone watching!
