Christine Kim | Strategy at Scale

Hi, everyone. Welcome to Grey Matter, the podcast of Greylock, where we share stories from company builders and business leaders. I'm Heather Mack, head of editorial at Greylock. Today we're talking with Christine Kim, an investor here at Greylock. Christine joined the team in September 2020, following five years at Uber. Before that, she worked at Box Bloomberg LP and Credit Suisse. She's also worked as an angel investor and VC scout. But the focus of today's conversation is Uber, specifically Uber Eats. Christine led the team that launched that division. While it started out as a small project, it quickly caught on, and today Uber eats generates $50 billion in revenue annually. The pandemic further solidified Uber Eats as a critical component of Uber's business as the app accelerated demand for food delivery. Eats was also key to sustaining Uber's growth as its core rides business shut down overnight. The experience taught Christine some valuable lessons that are applicable to product development as a whole. We're lucky to have her here to share this with us. Christine, thanks so much for joining Gray Matter today.

Thanks so much for having me. I'm so excited for my first Gray Matter podcast, hopefully not my last.

So, Christine, a lot of what you learned at Uber has shown to be very helpful in your current role as an investor and partner to early stage founders. But before we get into Uber, let's quickly hear about what you're working on these days. Since you joined Greylock a year ago, what have you been focusing on?

I'm really interested in, you know, how technology changes and transforms sort of the tactile, real everyday things in our lives. That's one of the things that really drew me to Uber initially. But, you know, it's things like how we get and shop for things, how we move around in our cities, how cities are built, how receive healthcare and, you know, wellness and all sorts of things like that. So that's kind of the thing that has always drawn me to be interested in technology, and those are things that I focus on here at Greylock. I think those areas, they tend to boil down into commerce, health and wellness, entertainment. And I'm really interested in the creator and passion economy, especially as it represents a new way of working for our younger future generations. I like to look at things with a Gen Z lens or really like to think about the way that younger generations are approaching work and health and whether that's mental health or physical health, the way that they're approaching connecting with each other online. And so those are really the big themes that I spend a lot of time in here at Greylock. You know, with my Uber experience as well, I also look at a lot of marketplaces, those can both be consumer and B2B marketplaces. And I've led a number of investments, some that are still unannounced, that are in that marketplace category as well as in health and wellness. One recent one that was announced is Instawork. Really excited about that one as it represents a future way for most people to be working. You know, it's, it's actually not too dissimilar from one of the projects that we had at Uber called uberworks, but it's really focused on helping Americans get back to work. It's helping them get jobs that are temp jobs, that are flexible, that they can work on an on demand basis. And it's helping companies who need flexible temporary work during peak seasons, seasonal for high throughput events, last minute, to be able to find these workers in an on demand way.

That's quite a range. You also worked across several disciplines while you were at Uber. You, you started as an engineer, then switched over to product, and from there you were instrumental in growing Uber's API ecosystem. This was at the same time Uber was going through a period of significant expansion. Can you share more about the various roles you held at the company?

Would love to, yes. So I joined Uber as an engineer and then by the time I left five years later, I was in the product organization. And so I got both those experiences. I felt a little bit like product in a way, as being a jack of all trades and definitely working in that engineering capacity helped me in that role later on and even now in venture. One of the things I love about the job is that it's across several disciplines. On any given day, you're looking at a ton of different industries and evaluating it from all different lenses. So that's always an aspect about work that I've really enjoyed and, and really tried to be that, I should say, Jill of all trades. In my own role. I also got experience, you know, while I was at Uber, working on both the rides in each side of the business. I initially started on the ride side of the business and there I was working on our API and developer platform team. That was a super interesting vantage point because it allowed me to sort of sit at the seat of, you know, the intersection of a lot of third party partnerships that Uber was pursuing at the time. And so any partnership that we are pursuing that required a API integration between two companies, the systems, let's say Uber and another third party was something that was built by Our team, and so that could span a lot of different applications. And so one was an integration with Google Maps, another was an integration with Baidu Maps, which is sort of the Google Maps of China. And that actually paved our way to allow us to enter China as a territory. That was a very strategic partnership for Uber. It got me exposure to other areas, like new modalities like bikes and scooters, which is a big part of our strategy today. And through that team, I worked on our jump bike integration, which ultimately led to the acquisition of that company. We one of the later projects that I worked on on that team was an eats integration with McDonald's, which was our first enterprise partner. Through that, I got a chance to work with the Eats team super closely and loved that experience. And so I transitioned to Uber Eats. And that really kind of defined my last two years at the company where I was in product and I was working on the Iter Experience team. I was there from 2015 to 2020. You know, from my time at the company, having those roles across engineering and then product, working across rides and then Eats, I definitely saw the company go through a ton of changes. The company, you know, went from 2,000 to 25,000 employees. I saw the company launch Uber Eats. You know, we changed CEOs, we went through an IPO. We also went through a lot of struggles. We laid off nearly 20% of our workforce in the midst of a pandemic, which was a huge struggle and blow for our ride side of the business. And so it was certainly an interesting time to be learning and developing myself at Uber.

So what was the genesis of Uber Eats? Did the company just see a lot of demand for food delivery and make the natural move into the sector?

Yeah, the genesis for Uber Eats is really interesting. The company in general wasn't just focused on food delivery itself. We were really starting to think about, hey, we've built this system that allows you to move people around the city from point A to point B through this peer to peer network of cars. And theoretically you could imagine moving anything through the cities. We are kind of thinking like in abstract terms, what it would mean to literally move packages or real goods or like the building blocks that, you know, you kind of think about the city as this living, breathing organism. And we felt like we were part of this narrative and just helping people and inanimate things right, get around that city. And so food was one of those things that we felt like there would be a natural demand for. But really it was from that perspective of we have this logistical Layer that could theoretically move so many things last mile wise across the city. And so what else could we start to deliver in that capacity? So we were interested in retail, you know, delivering the goods that you bought online. We were interested in on demand food. And so one of the things that we first tested was on demand food. And that was what would eventually graduate to become Uber Eats. But early in those phases, I definitely recall us just sort of iterating on an let's deliver anything kind of vision. And it just so happened that product prepared food and meals was the thing that there was the most demand for. You know, at the time, I think we were also thinking about, okay, what do we have built and what do we need to layer and add on? And this kind of helps me as I think about now too, as an investor, as I'm evaluating marketplaces. You know, marketplaces can be so nuanced. I think there's two sided marketplaces, which is what Uber had built up until that time. A marketplace between riders and drivers and connecting them and doing the dynamic pricing and matching and optimization between those two sides. What we wanted to do with food delivery is introduce a third side of this marketplace and so building it into this marketplace restaurants. And now you have a courier picking up food from a restaurant and then delivering it to an eater. And you have now three sides that are all interacting in a pretty complex way. And so what we tried to do at the time was really test this concept without introducing too much change into the system. We could talk about what building that MVP looks like, but it was definitely thinking about what where is the system today and what are the small tweaks that we can make to start to validate demand for something new like food delivery.

Uber eats as it is today is known for convenience and variety, but you had to start very stripped down. Talk about designing the pilots. What were the considerations you had to make in order to launch and test quickly and without overspending on time, money or human capital?

I think having a small and nimble team is key to this and holding yourself to that bar of trying to ship something really quickly so you can validate the minimum concept and just get back to building if you know, as soon as you start to validate some of those early signals. I definitely recall that being an important factor for us in testing this early concept for food delivery. To that end, we didn't even think about developing a separate food delivery app. So the Uber Eats app that we know and love today, that was like super far off from our vision back then. You know, I think we knew that we wanted to get there eventually, but we were looking at the existing rides app that we had, where we had all our drivers, all our riders already downloaded this app. How can we actually just include Eats or Food as a tab or as, you know, the most minimum feature and introduce it into this existing app that we had? And if you recall back then, the product selector for Uber, you opened up the screen, there would be a map, you'd put in your address, and you would kind of toggle between do I want an Uber Pool? Do I want an Uber X? Or do I want an Uber Black? And so in that toggle, we just introduced a new icon that was like a burrito icon, and you just swipe and you're like, I want food. Users easily got it. Even though it was totally turning the app on its head in terms of the value that it was going to be delivering. What happened is instead of a car coming to you to pick you up, to take you to the destination that you put in, a car was going to come to you. And basically there would be an exchange and it would drop off a burrito. We really had it constrained to literally just burritos on a given day. I remember the initial pilot, we would choose one restaurant for any given day. And so we kind of had a weekly schedule. Like Monday would be burritos, Tuesday would be sandwiches, Wednesday would be salads. And we worked with a restaurant to say, hey, let's order 300 meals, distribute those meals in 10 cars, put those cars live on the map as like, you know, an available Uber X, but just put it under this food icon and then just have them fulfilling trips one by one as they're dropping off food. It was a real true hack on the Uber system just to get food out there. And what it ended up validating for us is that there was demand for this product. And that gave us the confidence to ultimately go back to leadership, put together a team, invest in an actual business around this, and build what would eventually become a standalone app that's more familiar to the experience that we know today that's a standalone Uber Eats app.

In order to know whether Uber Eats was working, what kind of metrics were you looking at?

For one, there were sort of like top line signals that came through super strong. And that was also something that initially motivated us to test this in the first place. We were seeing what was going on, innovation and food delivery and seeing that, you know, customers were willing to get this and. And for a long time, Grubhub was around as well. Right. They were working with restaurants in New York, your original grubhub Seamless. And they were delivering through restaurants that largely already had their delivery fleet set up. And so for us, it wasn't really as much about do people want food? Do they want it on demand? But there were sort of a level of like second order questions that we were able to test. You know, for a long time we had this hypothesis that people wanted more elevated food. They wanted more than Chinese takeout and just pizza. And we felt like that model had been largely proven to be successful, but we wanted to kind of test, could you get your Halal guys and your Korea now through Uber Eats? Was there demand for sushi and salads, things that historically you would have thought like need to be fresh and aren't really suitable for delivery? I think the pilot really allowed us to test a lot of different cuisines. It allowed us to pressure test delivery times. How long were people willing to wait? What price were they willing to pay? And so all those things ended up being really important levers for us. We're a very data driven organization and I think those second order things were things that we were really trying to validate with the pilot. We really tested the pilot for about a year before then going back to the market and launching our standalone Uber Eats app. The first order, things like pure simple demand for pizza and Chinese takeout and even just some of the more basic foods. We already had some confidence in those. I would say Ubereats today is now doing $50 billion in revenue a year. It's definitely a lot bigger than I think any of us had imagined back when we were testing it as this little alternative product selector inside the rides app for sure. And it has proven to actually be a huge part of Uber's success and survival through the pandemic. Really. I think going into the pandemic, we were just fresh out of going through an IPO process. We had this big commitment to being profitable. A lot of that was hinging on the rides business, carrying through that moment. And then, you know, 2020 serves us a once in a lifetime global pandemic and travels down our rides businesses down. But actually the eats business was way up. And so having that diverse placement of both the rides and the eats business within our company was huge to our success, I think, in weathering 2020 and has proven to be an interesting tool in our toolbox. And I would say the eats business today has much grander visions than also just food delivery. We're thinking about Alcohol delivery, we've just acquired Drizly. We're thinking about on demand grocery. I think we're thinking just about in general how people's attitudes and shifts around acquiring and getting food are rapidly changing. I think for a long time we were only delivery and now we're doing pickup. And so there's a lot of other vectors. When you think about food, food is really something that we do three times a day, seven days a week. And so it's amazing to see the story of eats and how it's really grown to be so much more than what it was initially.

I want to talk about the foundational strategies that were at play that made it possible for Uber to pull off this very complex undertaking. And all the while, as I mentioned, like the company was growing rapidly. Aside from the specific factors that we just talked about, like the combination of increasing demand for food delivery, the willingness of drivers and restaurants to participate, what else was instrumental in making this all work?

There is a few things I can reflect on that defined us as an organization or were really strong first principles that I really reflect on my time at Uber. And I would say every single product team and every product manager, engineer, data science designer really was instilled with this mindset. And so for one, we were a very data driven organization. I think in the product which itself, if you think to the rides product, it's a pretty simple user interaction. You press a button, a car comes. But under the hood there's this huge data optimization problem that's happening. Matching the closest driver to you, optimizing what that price, what is the navigational route going to be. And like I talked about with food, you introduce another player into that market, restaurants and food, and it just becomes this super complex engine. And it's great for anyone who really loves to nerd out on data because there was so much data to work with at the company. It was a very data driven company and I think that informed all of our decisions. And so how that manifested when it came to building product and some of those foundational strategies is everything was measured, everything was AB tested. You know, so every single feature that we launched and then, you know, really down to every single iteration on a feature, when it came to even copy or placement of a button or, or changing major changes like changing a checkout flow, every single one of those was AB tested and optimized to make sure that we are constantly shipping the most optimized version to our end users. So I think there was an element to letting data drive our decisions and then Mixing that with some of that intuition of like, where do we think the future is going? What are some of the things that we have theses around? We believe that alcohol or grocery or food delivery is going to be the future, but then we also can test it along the way. And so I would say that kind of mix of injecting some intuition into the problem, but then still testing and using data to drive those decisions was a huge part of our culture. Another thing that really defined the way that we worked was just really kind of having all ideas out on the table and even being willing to revisit very old ideas. Part of that is, you know, this narrative of just, you know, well, the environment that we're in today is different than the environment yesterday. Let's put it to a test, let's a b test this thing. Even though it didn't work a few years ago, maybe it'll work now. And so I think it's in line with this overarching thing that we are very data driven. But an example of revisiting an old idea that was a failure in the past and then turned out to be, you know, a way for the future is our going back to this concept of Uber thinking about delivering everything, not just delivering food. You know, at one point we were thinking about being the last mile infrastructure to move all things, people, packages, goods, groceries. And I think you're seeing some of that manifest with Eats. You're seeing eats get into grocery and alcohol as we've discussed. But we're also really getting into retail. One of the last projects that I worked on at Eats was getting into things like pet food and you know, retail goods that you might buy, electronics, toys, things that we were once interested in having in the Uber platform, but for, you know, a variety of reasons, actually didn't work out that well. And so when we first initially tried this concept of delivering things, I should say even non food things, that was in 2014. And we were building what we called the Uber Rush API, the idea that you can kind of rush, last, smile, deliver anything. And it was an API based solution. So the vision then was that we were going to partner and distribute an API to retail partners who could then use the Uber network to deliver goods. And so we actually lined up some pretty good launch partners. We were working with cvs, we were working with Walmart, they were thinking about shipping pharmaceuticals and over the counter medicine and groceries through this pilot. And I think the key was that this API based solution actually meant that these merchants, like a CVS and Walmart had to provide the engineering to basically provide what Uber was providing before that front end application for users to interact with. And I think at the time it was just a little bit early and they weren't quite ready for that. And so largely we looked at this endeavor to deliver things as a failure. But I think later on in 2020 we actually revisited this concept and with a slight tweak which was instead of delivering an API based solution and expecting Walmart to build that application layer, why don't we actually just build an eats like experience to deliver over the counter medicine, your toilet paper during the pandemic and your pet food. And so I think now we've actually reinvested in this delivery of things business. Slight tweak in that we're owning the application layer and the UI and it's turned out to be a 3 billion GMV business for eats, which is incredible.

So it's a classic, you know, you listen to the feedback from customers and users, you incorporate it, focus on how you could leverage your existing strengths. And once you were able to get all that working, how did you go about scaling it?

I think there's a few points that I would want to bring up to address scaling once you have something that you have found that works. And so one of those is just admitting that Uber did an incredible job early in its stage as a company and focusing on international expansion. And so if I reflect on the cities that we are launching in, we first launched in San Francisco, but really we were just still a series a company when we launched the very second city in Paris. And so we didn't really have this period where we got to US or American domination and then started to expand internationally as some people might think. Those international cities were part of our mix from day one, really. And actually very famously, Travis had the inception for Uber while he was in Paris, kind of observing that the taxi system there was not ideal and it was super hard to get around in this otherwise very delightful city. And so I think to give the early team at Uber that credit, that they always had that focus on international expansion very early on that was definitely key to our success. What that meant for us is you actually had to have very high operational excellence built into the organization. And so I reflect at Uber as being, yes, a tech company and I had my place as an engineer and as a product manager in that company. But equally as important as the tech was to the success of Uber was operational success. And so I definitely think about this whenever I'm as an investor looking at Companies that have that element, where there is geographic complex rollout or there is a heavy degree to operational scaling to the business, it's just really having that excellence with how you scale, how you build those teams. For us, that meant having product and engineering and R and D headquarters in a few areas, mostly in San Francisco. And then later on we had offices in New York and Seattle, and then really this huge distribution of city offices. And by city office, that might mean, you know, we had three people in Paris and one person in Amsterdam. And, you know, really every single city where Uber was, there was someone on the grounds in an operations capacity that was signing up drivers, answering support tickets, really being the boots on the ground. That was a business model that turned out to work for us.

Wasn't adopting an international operations playbook a huge use of resources for what was a really pretty small company at the time?

I think some of it was just incredible ambition on the company's parts to be the international company and be everywhere. And so to some degree, that worked, and in some places it didn't. You know, there's along the way, it's been a lot of places where we tried to have a presence and then ultimately pulled out of that region. But then in other areas, I would largely say Uber having that being on that international stage has been a huge part of our success as a company. And so I reflect on those moments for us as a company, like, you know, we talked about how eats is this important balance to have in our platform alongside rides. And we talk about how, in addition to that, having international presence in many different markets was very critical to our success and not just being concentrated in the U.S. it certainly provided a ton of challenges from the beginning. So it meant that everything from localization to regulatory rollout and even just functioning in a distributed way, I think now I think about companies, it's very in vogue to have distributed fully remote teams, especially as we've learned in the pandemic, that we can all work that way. I kind of think back to 2014, 2015. It's kind of crazy to me to think that Uber was already building in a way where we had 300 offices to serve all our 300 cities. And that was always part of the model, obviously regulatory. And every single country has different rules in terms of, you know, mostly coming down to how taxis are treated there and what legal protections they have to operate in the respective country. That affected and was certainly a challenge for us to roll out. There are some countries where we tried to expand very aggressively, some worked out and some didn't. And I think that's kind of a simplistic way to look at the situation. It's not a matter that, like, some worked out and some didn't. You know, if you look at the example for. With China, for example, we tried to enter China, be the dominant player there. It was a hugely, fiercely competitive market. We were competing with two fierce competitors that ended up merging to be Didi. And then we ended up pulling out of China for a stake in Didi. And. And so I think at the time, I remember that being largely viewed as a failure. We had great ambitions to say, you know, we could be the Uber of China. China was looked at as this green field, huge market for us, but it actually turned out to be a really valuable decision that resulted merger. We now own 12% of DD. That's now worth 8 billion for us. Just that stake in that company. We would have never had an opportunity to even be part of that company or have, you know, that financial stake if we didn't even attempt to go after that region in itself. And there's a lot of stories actually like that. If you look at Careem, we've done it in Saudi Arabia, we've done it in multiple places where there's some countries where we continue to operate and continue to lead in category position. There's some countries where we may have historically tried to operate and now as a result have merged with a business, bought a business, or sold to a business. All, I think, are good strategic chess moves along the way. And then, you know, there's some companies that, you know, for some reason or another, like South Korea comes to mind. Uber has always been illegal. Taxis have always been the favored model. And we just, you know, tried to fight there, but kind of pulled out eventually. It definitely gave us a framework for evaluating, okay, for certain markets, is this a market where we want to invest in? Is this a market where we want to maybe strategically partner with an existing operator? Or is this a market where we don't see a path to either one of those and we should be pulling out?

Speaking of complexity, Uber has a lot going on under the hood, but the user interface has always been really straightforward. Obviously it has to be or people won't use it. But as you've described, testing out different features and services usually means having to build a standalone app first. But then when you want to bring it all together, how do you do that?

This is another great example of how what's old is new again. And I think about eats in the very beginning. As this spawn outside of the ride or within the Rides app, I should say how we initially kind of hacked the Rides app to just see if we could deliver food. And then as soon as that we validated demand there. We kind of had the green light to go build a standalone Eats app. Now I think we're back to, not to say square one, but we're kind of back to this strategic vision where we're thinking about, well, what would Uber be as a super app? Definitely, as you think about international regions where super apps are a lot more common and popular, we're getting really excited about Uber as the launchpad for your public transit, the food you need, your groceries, booking a hotel, booking a flight or getting that last mile ride. And so how do you actually unify maybe this portfolio of products all together in one cohesive place? It's not going to make sense to have a dozen apps. And so we're actually thinking about merging these apps into one unified experience. And so in some ways it feels a little bit full circle. Of course in some markets it's going to make sense to have two separate experiences. I don't think it means that the Uber Eats app is going away, but we're certainly thinking about platformizing our Rides app. So it's a little bit more multi purpose. And so, yeah, very much core to Uber's strategy and ethos is a super simple, deceivingly simple experience on the surface, but really under the hood, there's a lot of complexity to make that happen. And I'm sure for a lot of companies and products out there, that is the case. And so it's really a testament to great engineering and hard work to make that a super great experience. But also I think thinking about the complexity that goes into that. So for example, when we're thinking about bringing in some more components into the Rides app, let's take Eats as an example and bringing that into the Rides app. At the foundational level, there's a lot of core components that are shared across Uber.

And what were the trade offs to the platform initiative? How do you figure out where to compromise?

You know, one would think that it's just super easy to merge these two product experiences into one because we have profiles and accounts and wallets and payments, even customer support and receipts built in a way that it's all unified. You know, your Uber account is the same whether or not you're ordering food or you're ordering a ride. But as you actually think about some of the other kind of features that we have built Some of them have been optimized for certain use cases. So, for example, for rides, we built this loyalty program and wouldn't it be great if that loyalty program actually worked across all our products? Or for Eats, we built a pretty sophisticated promotions program. You know, with Eats, because you have so many different SKUs you can, we have all these sorts of clever buy one, get one free or buy three and then get 20% off your total order types of promotions. Wouldn't it be interesting if you could layer some of those promotions into the ride setting or scooter setting? And so merging these two systems and these two products actually required a lot more under the hood. And I think that's an example of one way in which we're thinking about still marching forward and creating that very simple and elegant experience for the user and kind of picking and choosing all those things that we have to do under the hood to really make it work.

You really got a lot out of that experience. What are some of the ways it's helping you in your work today?

I think some of the lessons that I think about for myself and I think about for companies as I'm interacting now on the investment side of things is to one, just never get too comfortable in your successes. It's amazing and you should be proud of all your accomplishments. But it can be a double edged sword, right? I think behind every success there's something that's also, you know, maybe getting overlooked or not done or maybe you're cutting corners. And so to have that moral compass or to really think about, you know, what are all those other things that we should be paying attention to besides just growth and revenue and new users to just pay attention to that as well. I think the other side of this is, and this has been something that I think the pandemic has shown a lot of companies as well is it's amazing and it's super fun when you are growing and when you're doing super well. But are you still around and are you still excited about the company and are you still showing up to work when the company's not doing well? And you know, Uber really had that challenging moment in 2017, culturally, we also had that moment alongside a lot of other companies struggling through the pandemic. And I worked through those times and it was really amazing to see the grit and perseverance of, of employees, of our leadership to really push through those challenges. And so it's really inspiring to me whenever I meet founders that have companies that have weathered their own tough times through the pandemic and it's something that I always look for in founders, too. It's great when there's a success story, but I love it even more when there's that founder that has that challenge story and you can sense that they have that grit to get through it.

Awesome. Well, thank you so much, Christine. It's been a pleasure having you with us on Gray Matter today.

Thank you so much for having me. I had a lot of fun.

That concludes this episode of Gray Matter. If you enjoyed this interview and want to hear others like it, please hit subscribe. You'll get new episodes delivered directly to you, and you can catch up on episodes you may have missed, like Reid Hoffman's discussion on the true value of an MBA degree for entrepreneurs. You can subscribe to gray matter@soundcloud.com greylockpartners. You can also subscribe on Spotify or wherever you get your podcast. You can find new episodes and blog posts every week on greylock.com blog and you can follow us on Twitter reylockvc. Thanks for listening.