The autonomous cars are coming.
On this episode of Recode Decode, host Kara Swisher hands the mic to Recode Senior Transportation Reporter Johana Bhuiyan, who spoke with Lyft Director of Product Taggart Mattheisen. Lyft is moving steadily toward a self-driving experience for its riders, but it is also working on all the other aspects of the business, including driver happiness.
You can read some of the highlights here, or listen to the entire interview in the audio player below. We’ve also provided a lightly edited complete transcript of their conversation.
Kara Swisher: Recode Radio presents Recode Decode, coming to you from the Vox Media Podcast Network. Hi, I’m Kara Swisher, executive editor of Recode. You may know me as the person who will try to backseat drive from her self-driving car, but in my spare time, I talk tech, and you’re listening to Recode Decode, a podcast about tech and media’s key players, big ideas and how they’re changing the world we live in. You can find more episodes of Recode Decode on Apple Podcast, Spotify, Google Play Music, wherever you listen to your podcasts, or just visit Recode.net/podcast for more.
Today, I am handing the red chair over to Recode Senior Transportation Reporter Johana Bhuiyan. Johana recently spoke to Taggart Matthiesen, the director of product at Lyft. Johana, tell us a bit of what you talked about.
Johana Bhuiyan: We talked about Lyft’s new self-driving ambition, working on actually building out the software, the brain of the car. You’ll no longer have to fist bump with your Lyft driver. There’ll be nobody there.
KS: I never want to fist bump my Lyft driver, not once. All right. Thanks, Johana. Let’s take a listen.
Thanks, Kara. I’m here with Lyft director of product, Taggart Matthiesen. Taggart, welcome to Recode Decode.
Taggart Matthiesen: Thank you.
We have quite a bit of ground to cover. Lyft has been really, really busy — some things that you work on, some things that you don’t — but for instance, the company expanded into more than 100 cities in the first quarter of this year. You closed a $600 million funding round, valuing at $7.5 billion, and you announced a new self-driving strategy. You’re also catching up to he who must not be named, Uber, in the U.S., in terms of market share. There’s a lot I want to talk to you about, but first and foremost, I would love for you to explain to our readers what it is that you do at Lyft. I did not intend for that to sound diminishing in any way. What does a day in the life of Taggart look like?
Yeah, so director of product, responsible for autonomous at Lyft. In terms of what do I do, we’re trying to solve a pretty hard problem, in terms of how does a car drive itself? Looking at it, we have a team that is responsible for the platform for our network. Most of the time it’s working with them to figure out how are we going to extend this in support of some of the partnerships that we have, as well as working with our other team that we announced recently, which is actually the brains, this self-driving system that we are planning to bring to market.
Great. Let’s talk a little bit more about you. Before Lyft, you were at Twitter for two years?
Talk to me about how that happened. How did that transition happen? You were at a social media platform company, and now you’re at a mobility platform doing self-driving, which is not an easy problem, nor is it an easy transition. Was that something that you signed up for coming into it, or what kind of interested you in that? What part of your experience lends itself to heading up a self-driving department?
Sure. I would actually take it even farther back. I spent nine-and-a-half years at Salesforce right around the area here. Most of the time I actually spent on building up that platform, that enterprise platform, working through with a number of partners, a number of solutions, and figuring out how do we extend Salesforce as a product, an enterprise product, and tying into a number of systems.
I then went to Twitter to do something very similar, which is how do we take Twitter and how do we provide value to businesses? This is a kind of data platform that I actually helped build out while I was there for two years. Really, the premise there is someone like Coca-Cola is not necessarily interested in the fact that there were a million tweets about their latest product. They were more interested in the demographic and psychographic key elements of those messages. How do you distill that into something that they can quickly evaluate? Those were the types of services that we started to build.
Peter Morelli, who is VP of engineering at Lyft, I was actually very close with him at Twitter as well as Salesforce. I had a quick conversation with him about two years into my stint at Twitter. Got an opportunity to meet John and Logan and Tali, who runs product. Amazing conversations. Love what they stood for and kind of the focus, and joined Lyft very soon after that.
In terms of what I started with, I actually started kind of on the enterprise side on the platform. Your question, “Hey, Taggart, what is it in your background that allows you to focus on self-driving vehicles?” The interesting thing is I’ve been building platforms for more than 12 years. Really, that is the key aspect. If you think about this open platform that we’ve built and how we’re tying partners onto our network, that is key to what I’ve been doing for over a decade. In terms of the self-driving car, the perception, the localization planning …
Yeah, you’re not touching that.
No. We have some very talented people that are focusing on that. I’m working with them in that area as well.
Great. I do want to talk in more detail about the different self-driving strategies that Lyft has right now, but just more generally speaking, has that been the large part of your focus in 2017? Are you no longer doing enterprise? What is the largest or the most important problem you’re trying to solve right now?
This is my focus, 100 percent, and actually has been the majority of my focus for about a year and a half now. In terms of day-to-day responsibilities, what we talked about earlier, again, it’s how do we extend our platform, basically our network, as well as starting to build out the capabilities in the vehicle itself.
Gotcha. What about the company’s priorities? Lyft has been busy making announcements from tace mode to mini mode, but also relationships with transit partners, like Amtrak in New York. It feels like there’s a lot of movement on a number of different fronts. Is there a hierarchy of things that the company’s focused on now? Is it the current business? Is it down the road?
Yeah. I would say autonomous is one of the key pillars that this organization is focusing on. We are dedicating a significant amount of effort, and as part of this SDS announcement, we even have …
What is SDS, for listeners?
Sorry. Yes, we should work on that acronym. Self-driving system. Imagine it’s basically the brand. We are deeply focused on that and have basically built even, you could call it a unit, in terms of we have real estate down in Palo Alto that we’re starting to spin up and plan to have hundreds of engineers focusing on this problem in 2018.
Autonomous is front and center for the company as a whole?
Absolutely. We see it as an extension. A lot of people have asked, “Hey, Taggart, how does autonomous fit in with Lyft’s strategy?” John talked about this a little bit in his Medium post, but we see it as kind of an extension to what we do today. Today, we’re 100% drivers on our platform. Over time, we see kind of this hybrid evolution where autonomous will become, at some point in our business, very, very key to how we provide transportation solutions to people. I see this not something that necessarily impacts us today, but in the next couple of years, something that we’re going to start to weave into a core aspect of our product.
Gotcha. Before we talk more about autonomous — because that’s obviously what you’re focusing on, what the company’s focusing on — there are a few other developments. Lyft, for instance, is now in 350?
350 cities in the U.S. The plan was to be in 100 more by the end of the year. You guys were ahead of schedule in the first quarter. You expanded into 100 more. This happened at the same time that Uber was facing public scandal after public scandal. I think that that in conjunction with the original plan to scale has allowed for Lyft to catch up in market share. I want to break that down a little bit. What does the company primarily attribute that boost to? Is it just you think its difference in branding, skill? You guys also have a new marketing campaign. What do you think really led to that boost in market share?
We’ve continued to focus on the same kind of core goals that we’ve had as a company for a very long time. You look at the experience. You look at how we treat passengers, how we treat drivers. We haven’t really changed any of that. Yes, the question around the delete Uber movement, yes, there was an increase, in terms of passenger activations. We saw actually a 60 percent increase in passenger activations after that.
Oh wow, the first delete Uber movement?
Yes, that had an impact, but really I think the main growth here has been kind of our laser focus into how we create an amazing experience both on the passenger and driver side. Again, I like looking at what we’ve done with drivers. We’ve had over $250 million in tips that drivers have received. I look at some of the things we’ve done for passengers, scheduled rides to even the Amp, which is kind of our new beaconing tool that we’ve rolled out to all of the markets to the point where we … Again, you talk about growth, we’re doing over a million rides a day.
We’ve seen some pretty spectacular growth in the last year or so, and I think in terms of what do you attribute that to? Kind of focusing on our core competency.
Just for our listeners, the Amp is this physical pill-shaped thing that you put on your dashboard. It literally says “Lyft” on one side, and on the other side it says, what, “Hi, Taggart?”
Yeah, it says, “Hi, Taggart,” or whoever the passenger is. It’s a pretty interesting beaconing tool. The thing I like about it is if you’re in a crowded intersection or an area of the city and you’re looking for your vehicle, not only do you have this nice little emitting device in the vehicle, but on the passenger app you can see the same color that’s getting flashed. That helps you disambiguate between some of the vehicles that you may think are coming to pick you up. So it’s been very helpful.
Right. Is that something that you worked on, or is that a separate team?
It was a separate team. I actually was part of the original group working on that before the autonomous initiative. No, we have plenty of product people and engineers that are focusing in these areas.
Great. It’s one thing to get an influx of new users. It’s an entirely different problem, almost a harder problem — and you can correct me if I’m wrong — to sustain that ridership. Technologically speaking, what is happening behind the scenes as you’re seeing this new wave of riders? What happened on the ground when that first delete Uber campaign happened?
The thing that we continue to focus on is, I look at what’s the technology that’s supporting us. Think of mapping, locations, ETAs. We spend a significant amount of effort ensuring first, our system can handle this both on the supply side with drivers for kind of this increased demand, but we’re also improving the algorithms that we have behind the scenes to improve ETAs, to improve our comprehension of traffic in terms of routing, and our intelligence around locations as well. Yes, we’ve seen an increase, but we have been focusing on the technology behind it to make sure that our service can scale.
This was a focus of Lyft’s before the delete Uber campaign, right?
Oh, absolutely, and will continue to be for the entire life of this organization.
Did you have to dedicate any more energy to that once you saw … Who could have expected that wave of new users? Is that something that you had to actually dedicate engineers to?
We’ve always dedicated engineers to ensuring that we can scale.
Right, but more so in that …
No, I don’t see of any focused increase there. That’s something that we test every year.
Right. Is that something that Lyft would have had a game plan for, or is that just largely automated is what you’re saying?
It’s largely automated, and we do yearly tests on … We have certain days of the year. Halloween is one. New Year’s Eve is one where we see a significant amount of traffic on those given days. We do a significant amount of validation testing prior to that to make sure that those days are seamless. We expect spikes. It’s just the nature of our business. We’re able to sustain when we need to enough services to support that added demand.
Right. Broadly speaking, passengers often make decisions on which service they want to use based on price and ETA, like you said, sometimes routing, so convenience in general. Now we’re seeing, to some degree, passengers are making decisions based on branding and governance, corporate governance. How long do you think that’s going to last?
It’s a good question. If you look at our strategy and how we’re evolving our product, that’s not something that we evaluate. We are continuing to try to improve the experience. At some point, that stuff will probably fade away, but if you look at the improvements that we’ve had on our service, I think that’s where we would continue to focus.
What kind of improvements? Give me an example.
Improvements in terms of ETAs, so having better supply as well as just improving the passenger experience as it relates to the app. Calendar integration is a great example that was recently released. I keep coming back to scheduled rides, because I use that almost every day. Yeah, it’s a continued evolution of ensuring that we provide the right services to our passengers.
Just for clarity, scheduled rides, the app sends out a signal for a driver right before the ride is scheduled, or is that something that …
No, sorry. Scheduled rides, imagine you need to go to the airport tomorrow at 6:00 a.m. The nice thing about this is I can go into the app, I can schedule that ride. Let’s say I want to leave my house at 4:00. I just realized that I’ve just created a very early flight for myself. Sorry. Yeah, let’s say I want to leave my house at 4:00 a.m. I can go ahead and schedule that’s when I want to be picked up. Then behind the scenes, the Lyft service will go ahead and manage that and dispatch a driver to ensure that they’re at my house at 4:00 a.m.
They’ll dispatch a driver right before you need to get picked up, it’s not that a driver agrees to it and then …
No, no, but there’s quite a bit of intelligence behind the scenes, in terms of how we identify those drivers. I can’t get into too much detail, but yeah. We ensure that you as a passenger have a driver come pick you up in the morning.
Okay, great. Today, Lyft’s brand seems to be its competitive advantage. I’ve actually argued in the past that Lyft seemed to try to set itself apart just by its brand. The services were largely the same, beyond the fact that there’s tipping. That’s something that’s fairly hard to scale with a pool of independent contractors offering the service. Today, given that Uber’s branding is so marred, Lyft’s branding seems to be a very big advantage. Uber is in the middle of trying to rectify that right now. It has its own driver improvement campaign. It’s working on its public perception. Does Lyft have a plan for creating a competitive advantage for if and when — not to say that it for sure will happen — but if and when Uber fixes its public perception, does it go back to eventually becoming a price war?
When you look at our competition, yes, they are going to continue to change, in terms of how they push their brand. Again, I look at what we’ve done, and that is continuing our focus on the experience. Yes, they will change. They will come up with new initiatives. They will come up with a different branding technique. We continue to look at how do we improve people’s lives, both on the passenger and driver side, and I think that just means we need to continue to evolve. We can’t just sit where we are today. We have to continue to innovate. I think that actually is one of the reasons why we focus on autonomous and trying to figure out how do we slowly evolve these self-driving vehicles into our service and our network. I think the best way to answer your question is yes, they will evolve. They will change their branding. We just need to continue to focus on both our passenger and our drivers, in terms of improving that experience and improving our network as well, especially with the technology that we have.
Do you have something coming for new driver updates, new provider updates? Let’s break some news on Recode Decode.
Yeah. Yeah. I can’t talk to those areas, but if you look at the past, we’ve got launches every few months or so, both on the passenger side and the driver side. We’re going to continue to do that.
Okay, great. We’re going to take a quick break now for a word from our sponsors. Here is Kara Swisher again.
Thanks, Kara. We’re here with Taggart Matthiesen from Lyft. We were just talking about the company’s growth and it’s bump in market share, which is very much focused on today’s business, but you work on the future business largely, at this point. I want to talk broadly at first, because not all of our listeners know what’s going on in the self-driving industry, but can you explain in just a few sentences top line of what Lyft will look like in the future, maybe 10 to 20 years, let’s say, because I know that there’s a lot of debate over when this is going to happen. What is the company working toward right now?
In 20 years, I see this as a transportation as a service, very similar to the other service apps that you have where it is an autonomous fleet of vehicles that provide the ability for you not only to just get from, let’s say, point A to point B, which a lot of people think that’s just it, but also an entirely different experience. What do I mean by that? You just asked for three sentences.
No, that’s fine. I said a couple sentences. That’s totally fine.
Sorry. Yeah, we are evolving into basically a service, a transportation service. In 10 years or 15 years, this will most likely be an autonomous service, in terms of vehicles. The really interesting and exciting opportunity here is changing that experience. It’s not so much about getting someone from point A to point B, but what they can do with that time. That’s where focusing on personalization … and whether that’s me going from this office back to 185 Berry Street in San Francisco and hopping in an autonomous vehicle, and that vehicle already being connected to my email or to my calendar, and for me to be able to take that time back and just kind of focus on work. There’s a ton of things that we can do in these vehicles that go far beyond just the very basics of transportation.
Right. You said evolve into transportation as a service. Is Lyft not today?
No. No. Sorry. Lyft absolutely is transportation as a service, but I think of it as almost like a subscription. That’s where I think we evolve. You look at Spotify. You look at Netflix. We become yet another kind of subscription service that is an afterthought. It’s a muscle memory of when I go outside, I need to get somewhere. I hop in my Lyft. As the technology improves, we have an opportunity to really kind of improve that experience as well. That’s kind of that tighter integration that I was talking about.
Gotcha. Why self-driving? Why is it so important for Lyft’s future?
One, one is safety. Autonomous vehicles are going to be far safer than anything that’s on the road today. Actually, if you look at vehicle fatalities, the last two years we saw the biggest increase, about 14 percent in over 60 years. That’s pretty significant. I think we have an opportunity, especially with these vehicles, to provide an extremely safe service, even more safe than what we have today, for passengers.
Two, as you get into an autonomous service, we have an opportunity to also look at improving the price point of the service itself. I think over time you could potentially see the cost of the service go down as it becomes something more of a subscription.
Then, I think, three, it will help us further our scale. Today, like we said, we’re in over 350 cities. This basically covers 80 percent of the U.S. population, in terms of access. We can even go further beyond that once we have fully automated vehicles that we can control, in terms of positioning and increasing the optimization that we have.
Right. How does it become more affordable in a subscription model? Talk me through that a little bit.
The goal here is that the vehicles, the technology over time and as we scale will become cheaper, and we can further optimize it to improve the pricing that passengers will get in the end.
Okay. It will become cheaper by virtue of what? How does it become cheaper?
It becomes cheaper, just the vehicles, again.
Also, you don’t have to pay commission out, so you can charge less?
I still think so. When we talk about commission drivers, I still think drivers will be in this equation. We can talk about the hybrid network. If you just look at the autonomous vehicles, there’s an opportunity for the car to become cheaper in terms of how it is manufactured, and then utilization of these vehicles. We can get these vehicles into a very high utilization pattern, whether that’s 60 percent or 80 percent, it’s anyone’s guess at this point.
Then also kind of positioning of these vehicles as well to ensure that we have the most optimal location for these passenger pickups. It’s a bit of a technology play, it’s a bit of a subscription play, and it’s a bit of kind of where these vehicles are evolving. We think the combination of those will help us lower these prices over time.
There’s still a fair amount of debate over when this is actually going to happen at scale. You talked a little bit about the hybrid model, we can go into that a little bit, but I think even two years ago people were saying either it’s never going to happen, or it’s going to happen in 40 to 50 years at the earliest. This year, I think we’ve sort of over-corrected and people are like, “Yeah, it’s happening. It’s going to happen next year.” Just given automaker’s production cycles, the fleet turnover rate, regulations, the infrastructure changes that we’re going to need, it seems like the reality is somewhere more in between. Where does Lyft stand on when we’ll see fully autonomous cars on the road en masse?
Fully autonomous, no driver?
En masse, we are still a number of years out. That’s why we talk about …
What’s the number, though?
It’s anyone’s guess at this time. You could say six to 10 years from now you start to see fleets of autonomous vehicles, but I think the better way to look at it, or at least in my opinion one way to look at it, you’re going to start to see these autonomous vehicles out in some of these urban areas.
Fully autonomous. The beauty of these networks is that … I’ve talked to a few people about this. We know the start and the end of the ride before it happens, which is actually quite interesting. We can look at the time of day. We can look at the weather. We can look at the traffic patterns. We can actually look at the segment, basically the route in which the car is going to navigate. We can evaluate based on those constraints or conditions whether or not we think it’s an optimal condition for an autonomous vehicle to be dispatched. What I think you will see is over time, you’re going to start to see these autonomous vehicles in pockets on the Lyft network. As the technology evolves and improves, you will see greater numbers of rides being serviced by these, or these vehicles being dispatched to our passengers.
The nice thing about all of this is that you, as a Lyft passenger, you’re kind of abstracted from all of this, meaning you’re always going to get a ride. When you need to go from this office to the airport, we’ll have a car in three minutes or less to you. In some cases, it will be a traditional Lyft, and in other cases, it will be an autonomous vehicle.
Right. That’s part of your pitch to automakers who work with you as well. You’re going to have an in-between when you don’t have enough fully self-driving cars, there won’t be that sort of pocket of time where consumers aren’t getting a ride. There will still be drivers fulfilling that, if there aren’t enough self-driving cars.
Right. If you look at … We have a number of partners that we’ve announced with the open platform, I think the opportunity here, and we want to collaborate. Again, this is very, very important, especially as you look at autonomous and where it’s going and what it means, I think, to society, we have the chance here to work together to slowly evolve the capabilities of this technology in a very isolated and safe fashion. That’s where we’re working, and each partner will be slightly different and will have a different experience as well, we will slowly and surely bring these on to our network in a way that as you, again, like I said before, as a passenger, it’s no different. Yes, it’s going to be an autonomous vehicle, but it’s going to slowly and surely be part of your day over time. Now, when will it be a fleet of autonomous vehicles 100 percent? I think that it may never be 100 percent. It will take a significant amount of time to get there.
It may never be 100 percent?
In the sense that it’s not … If you look at autonomous vehicles, one great example, we do non-emergency medical transport, which is if I need to go to, let’s say, the doctor’s office and my leg is in a cast and I can’t drive, we have a service for that. I think if you get into the world of autonomous, we may need someone in that vehicle to help that person, help them get out of the door. I think there are things that we’re doing beyond just getting a passenger from point A to point B where there’s additional services that we can, as a company, look at.
Yeah. Speaking of which, a few years ago when I first starting writing about Lyft’s autonomous efforts, I talked to a few people at the company, and there was this idea behind creating experience pods. Each car can have sort of a different theme. One’s a working pod. Another one is maybe a manicure, pedicure pod. Is that still kind of the direction you guys are heading toward?
It’s a fun aspirational idea of where can we take this. Yes, if you look at John Zimmer, he came from the hospitality vertical, and this is something that he’s talked about as well is how do you go from just taking someone from point A to point B to making it an amazing experience? Yes, over time you could have maybe a different car that takes me into work, versus the car that I take to, I don’t know, go to a concert or maybe go to Tahoe or go with my kids to soccer practice. These are all different experiences that we could potentially provide as part of our service when you start to think about the pod. I think there’s a lot that we need to do first.
Right, before that, yeah. Let’s catch everyone up on what Lyft has announced. For those of you who don’t know, Lyft has self-driving ambitions. Those ambitions were made clear for the first time when General Motors invested in the company and struck up this partnership where essentially GM cars would be plugged into the Lyft network. Fast-forward to today, Lyft is now working with Alphabet’s self-driving car called Waymo, a startup called Nutonomy, Jaguar Land Rover, have you announced anyone else?
No? He’s smiling, because he almost said a company, a partner. Do you want to tell us today?
You and I have talked about this before, but it does seem like there was a clear shift in strategy at some point, because now you’re also developing your own self-driving tech, which is the most recent announcement. Lyft at first was going to serve as this sort of agnostic platform that people could just plug their cars into, and the company was well positioned to do that. Now Lyft is also developing its own technology stack. That’s not cheap, nor is it easy. I do wonder what led to that decision to take on that cost.
We’ve been working on autonomous for about two years now. I would actually argue the strategy hasn’t changed. What you’re seeing is us slowly and surely opening up new pieces of that strategy. The first thing, as we talked about in the past, is this open platform and working with a number of partners to bring their self-driving technologies onto our platform to our passengers. That is something that’s very, very core.
It’s fairly organic for this, yeah.
Absolutely, absolutely. If you look at our efforts with the brain, the self-driving system, it’s actually tied to that very, very closely. The reason why we announced this and the reason why we think we’re so well-positioned is a couple of points. One, we have a significant amount of data that we do a million miles a day, in terms of trips. If you think about the self-driving problem for vehicles, a lot of people will talk about mapping. There’s 930 some-odd miles of navigable road in San Francisco, and yet we have enough information to help us identify a very small segment of that that would allow us to get our vehicles up and running. We don’t need to map all 930 some-odd miles of road. We can map a very small segment of that and bring these cars on those.
One aspect of that is the data. We know the routes, and so we have an opportunity to work with our partners to ensure that the technology that they’re bringing to market fits those requirements. The second piece on that is our core business is really about mapping and ETAs. We kind of talked about this a little bit earlier but it’s absolutely critical to get drivers to passengers as fast as we can. There’s a lot of kind of mechanics behind that. That actually is, I would argue, part of the autonomous problem. We’ve been focusing on mapping. If you think about how do you get a self-driving car to drive itself? You start with mapping.
Start with maps, yeah.
You then get into perception. You then get into localization. Then you can start talking about actually driving the car, which effectively is path planning. We have a number of people that have done this in the past. If you look at Luke Vincent who is VP of engineering, who’s running autonomous, came from Google, basically I call him the godfather of street view, but has a ton of experience in the world of mapping and perception. These are core aspects that the companies need to build in order to have a self-driving car.
If you look at our network, yes. Like you said before, it seems like a very natural extension to go talk to some of these providers to bring their technologies onto our platform. What we realize is that we also have a significant amount of both data and expertise, and our core competency as it relates to mapping with locations and ETA is actually key to this problem. We have some of these pieces, and that’s what brought us to the point of let’s continue to build that out, which is why now in — like we said — Palo Alto, Level 5 Engineering, we’ll have a large …
Level 5 Engineering is the new name of that policy?
Yes. Very cute.
Yes. I know. Yeah.
We’re probably coming up on time for this part of the segment, but explain why it’s not the smarter strategy for a company like Lyft that — I’ve said before — that you guys are well on your way to profitability. You already have all the ingredients to be a great platform for these self-driving tech companies. Why isn’t it smarter just to do that and let GM and Alphabet take on the cost of building out the brain of the car?
I think there’s two reasons. One, we actually have, I think, key aspects and assets that we want to help provide to these organizations, to these partners. We think we can improve and speed up those efforts. Again, we know the routes. We have been working on mapping or a significant amount of time as well. There’s an opportunity here for us to help those efforts. Then two, we need to make sure that we have access to this technology in the future. It seems almost — I guess in my mind, I’ve been thinking about it too long — but kind of an obvious extension to what we do today and an investment that I think is quite critical for our organization to continue to thrive and grow as a company.
You don’t want to be beholden to a partner’s timeline?
It’s not so much that. I think it’s more, we have a ton of experience and data that we can bring to the table to help our partners evolve their technology as well as evolve our own.
Explain to me, for our listeners, you could have done just the platform aspect of it, and then also provide your partners with mapping data and riding data, things like that, because that’s what you … Like you said, Lyft already has all of that. Why go that extra step? Are you guys trying to ensure that you’re not left behind in this?
It’s more of, we think we can do more than just provide the mapping data. We have the experience to help with perception, to help with localization. I think that, if someone asked, and I have been asked that question a number of times, it’s not just about the data. Yes, we could work and help some of our partners with the data, but we can also provide significant experience beyond just that.
Okay, great. We’re going to take another quick break. Here again is my boss, Kara Swisher.
Thanks, Kara. We’re here with Lyft director of product, Taggart Matthiesen. We have been talking about Lyft’s autonomous strategy. I want to continue on that a little bit. There’s still a lot of questions about how this is going to shake out in the end. It’s still very, very early. Lyft has only been working on it for two years. I think that’s about around the same time that Uber has been working on it. Waymo’s been working on it for nine years, but that’s them. It’s still so, so early. There’s still a lot of questions about things like which aspect of the self-driving supply chain, whether it’s the car, the brain, or the path to market, which is Lyft and Uber, which becomes the commodity.
What do you guys think is going to happen? Explain to me, what is the ideal model for Lyft? Obviously not becoming the commodity isn’t the ideal model, but what does Lyft’s business look like if every carmaker, every self-driving tech company, is on both Uber and Lyft? How do you get a competitive advantage there?
Yeah. The hope here is that a number of providers actually get to the point where they have level five vehicles. We’re actually quite excited about getting a significant number of partners, which is why we built this open platform. Your question around, “Hey, are you going to get a little concerned about this being commoditized?” I actually look at it as no, what our focus will be is how do we ensure that our platform is flexible enough to take any of these vehicles from these providers and then on top of that, I really think it’s going to be the experience is where we differentiate.
We look at today, “How do we differentiate?” We differentiate on the experience. I think in the future, that’s the same focus that we will have. Clearly, we’re going to want to make sure that we can integrate these to the point where we can optimize in certain conditions in cities, but it’s also going to be a point of how do you evolve beyond just looking at this of going from point A to point B. We talked about it a little bit earlier, but I think there’s a huge and exciting area of personalization around the experience that we can improve and an area that we’ll focus on. I think that’s also a number of years in the future.
I think getting there — and that’s where we’ve been focusing, is helping evolve this technology — that’s going to take time. I think doing it right instead of doing it in a contrived fashion is also something that we’re focusing on as well. We’re not looking to partner and then just release vehicles or experiences that don’t fit our brand. That’s why we’re working with a number of partners and we’ve spent a significant amount of time on this, and we’re waiting for the technology to get to a point where we feel it matches the level of experience that passengers get today, and hopefully over time improve on top of that.
Right. You and I have actually talked a lot about this particular aspect of it, but talk more about how Lyft plans to not control, but have sort of a say on how consumers kind of interact with the car. Is the plan to have control over the in-car experience at all? Is it going to just be through the app? What are you guys thinking in terms of consumer experience?
I think first we’re very flexible on that, but I think one of the things we should focus on is that experience. Today, we have the passenger app. We have the driver app, and then clearly you have the experience in the car. Once you take that driver out of the car, you go from this very familiar environment to a very foreign one, especially in the early years of autonomous vehicles. That, I think, is really an area where we can focus on not only providing the experience, but also a bit of education as well for passengers so they help understand how the car is perceiving the world, but that also gives us yet another opportunity to improve all of the kind of experiential stuff that we’ve been working on.
It’s absolutely about the pickup experience. Long before the car even arrives, how do we evolve that to ensure that it’s a very, very clean pickup, to getting into the car, to the validation that it’s, in fact, the right passenger, that all the passengers have gotten into the car, to then maybe some experience stuff around whether it’s your Spotify or Netflix account just magically turning on as you get into the vehicle, so it may be a quality ride. It may be that I get motion sickness, and so what I want is a slower ride. Or maybe it’s on Friday and I’m going home, and we know that, and so instead of going down the freeway, I’m going to take some side streets and kind of enjoy myself. There are all these things that we can do, and I think pretty critical, in terms of evolving the experience.
Yeah, and that seems like a huge, huge opportunity for Lyft or from any company working on this, creating a customer experience so that your brand is kind of the touchpoint for the consumer throughout the ride. Let’s break down the passenger validation aspect of it, just because that’s something that people know today. How is a self-driving car going to make sure that this passenger is the right person?
Yeah, one of the things we absolutely need to get right is to ensure the safety that the right passenger is getting into the right vehicle. There is a number of technologies today that allow us to do that handshake.
Not a fist bump?
No. No, no. I won’t get into the specific technologies, but what we want to do is ensure that there is a validation from the vehicle and a validation from the passenger that yes, this is, in fact, my car, that the passenger can get into the car, that there can be enough, kind of again, a handshake of sorts to ensure that the passenger has closed the door, has fastened their seatbelt. If there are other passengers in that car as well, ensuring that there’s enough validation there focusing on that safety aspect.
Right. Do you envision that being something that the car is manufactured to do? Do you think it’ll be via the app? Will there be some sort of in-car console that does that?
It’s definitely going to be an integrated experience, because if you think about it, you sit down, and there’s going to be weight on a seat. A sensor is going to recognize that. A sensor would need to then send it to … Imagine you have a Lyft experience on a console. That console needs to recognize the fact that there is weight in that seat. It will need to be a very, very integrated experience between the vehicle itself, the sensors in the vehicle, and then the experience that we plan to build out.
Automakers in the past, for instance with Alphabet and later on with Uber, have been sort of resistant to letting these tech companies take control of the in-car experience, because obviously they don’t want to become metal benders for tech companies. How have your automaker partners reacted to the idea of working closely with Lyft on things like how a sensor communicates with a passenger and things like that?
I think each provider of this autonomous tech, there’s going to be an aspect of their brand in that vehicle. It’s not like this is going to be Lyft only. If you look at the open platform, and I’ll use an example of one of our providers, it’s the Nutonomy vehicle. It’s the Nutonomy self-driving car on the Lyft network. We want to make sure that when I, as a passenger, get into that vehicle, yes, I know that I got in through my Lyft app.
We’d like to have the experience in the vehicle consistent across these partners for the passenger’s sake, but at the same time, we also want to make sure that there is an awareness where our partners can highlight their technology. I look at it as it’s a collaboration. It’s not just Lyft only. It’s absolutely kind of a Lyft, for example, a Lyft Nutonomy experience. I’m on the Lyft network, I get into a Nutonomy car. That’s where I think you need to have tight integration on that as well, but I don’t see it as just boom, just Lyft.
Right. This is something that I think you guys have probably fielded a lot of questions about when you announced that you were doing your own self-driving system, but it does appear that you are now kind of competing with your partners — Nutonomy, for instance, Alphabet’s self-driving arm. Talk to me about why that is or is not true.
Yeah, I don’t think it’s competing at all. I look at it as we’re building out this SDS. The goal here is that as we build out this technology, we have an opportunity to feed this technology back to our partners, whether that is mapping, whether that is localization, whether that is path planning.
Right, and you’d be able to sort of piecemeal it out. They don’t need to get the whole stack.
It’s early on, to be fair, but the goal here is to offer up as much as we can to kind of improve their tech.
You’re not competing because …
No, I don’t see any competition whatsoever. We’ve talked to our partners as well, and we still are working with these partners. Clearly, they see this as a collaborative effort, not as a competitive one.
It wouldn’t be the case where in a Nutonomy car, Lyft has its self-driving software. For that particular partnership, I know you guys haven’t talked too deeply about the details of any of your partnerships, but Newtonomy would be creating the entire self-driving stack and would just kind of be plugging into the platform, correct?
That is correct.
It wouldn’t be the case where Lyft would be inputting their own self-driving system in there or adding it in there.
No, not now, but if you think about where vehicles are today, and it may be … Let’s talk about perception stack. Depending on the city that you’re in, there are different pedestrian patterns. It may be that a given partner’s stack has been trained on a given set of cities. Maybe it’s San Francisco or maybe it’s New York. I think what you’re going to find is that there’s going to be very, very unique things that you’re going to want to train on as you go to different cities. It may be something that we can provide as a service or as an add on.
It is, again, early on with all of this stuff. Is you ask where we are today, we’ve built this open platform. We’re working with our partners. Now we’re also starting to build out this self-driving system that over time will get to the maturity where we can work with our partners to see how we can further collaborate. It may be, as you mentioned before, bits and pieces, but maybe it could be the entire kit.
Right. How important is it for Lyft to be first to market with a fully self-driving car?
I think it’s critical for us to be one of the first to have self-driving vehicles on our platform, but …
It doesn’t have to be Lyft’s self-driving software that’s powering it.
No. No, no, no. Again, if you look at our strategy, it’s both. Yes, we want to be the first to market that have self-driving vehicles on our network, but it’s important to note we’re less interested … You’re going to see for a number of years, pockets. To me, that’s not the exciting piece. The exciting piece is once you have tens of thousands, 50,000, 100,000 of these self-driving vehicles on your network. That’s where we want to be one of the first. That’s going to take time for us to evolve both our network as well as our partner’s technology, but that’s our core focus.
Why is it important to be one of the first to market at that scale?
Because our whole business, I think, over time is tied into that. Again, we talk about this hybrid network, and it’s not just 10,000 vehicles. It’s just making sure that we can evolve into it. The key here is to ensure that we have the sophistication where we’re in a city like San Francisco, and let’s say we have 10,000 autonomous vehicles, that those 10,000 vehicles can be weaved into our hybrid network to ensure you, as a passenger, will always get a ride, and that also as we introduce those into the platform, allows us to improve ETA. It allows us to improve our routing.
There’s a whole host of things that we can do, and it’s quite critical when we talk about those numbers, because those numbers, that’s where you really start to change the service and potentially change the experience for passengers as well.
Right, so the competitive advantage there is not so much … There is the aspect of it where there’s going to be increased profitability given that you’ll have fewer drivers that you need to pay a commission out to, but you’re saying that you …
I don’t think we’re going to have fewer drivers. I think that’s …
Oh, those cars in particular, it’s just all the profits go to Lyft is what I mean, certain models?
I think it’s too early to tell exactly how that will play out, but the goal here is to get those vehicles onto the network to tie them into our service in a scalable way, not just if you have five or 10 vehicles floating around a given city, that to me isn’t scale. It’ll take time to get there.
Right, but the competitive advantage of it being one of the first is that you will be earlier to be able to test and validate, get more information, and then improve the service. Then that improvement is kind of what makes …
Yeah, absolutely. That’s why we’re working with these partners. That’s why we’ve opened it up so early is that we want to be one of the first to market to have these relationships to help build this technology so it fits our platform quite well.
Right. We’ve talked a lot about self-driving. Obviously that has very real implications for drivers, although Lyft plans to have this sort of hybrid model where drivers may always exist on the network. This is a question I’m sure Lyft has gotten a lot as well, that your entire pitch is that you’re friendlier to drivers, that you care about your partners. How do you navigate being able to continue to communicate to drivers that they’re valued, but also say we’re working very hard, and we’re very serious about something that may, in fact, replace you at some point down the road?
I think the first thing is, drivers have always been part of our family. They have been core to our service. As far as I’m concerned, they will continue to be that. Yes, over time, technology will give us opportunities to potentially provide additional services on our platform, whether that is a concierge service, whether that is an in-vehicle experience where a driver helps out in that capacity. These are all things that we will slowly evolve and work with our drivers on. The other piece is if you think about hundreds of thousands of autonomous vehicles, there is an aspect of cleaning. There is an aspect of maintenance. There is an aspect of kind of fleet management. I think over time you’re going to see a lot of services that come up that kind of surround this area that provide us yet another opportunity.
Right, another opportunity to provide jobs for the drivers who may no longer drive for Lyft.
It’s another service that we can provide. In terms of drivers on our platform, we don’t see it being any less than it is today. We see them absolutely core, and we’re very early on, in terms of our growth. We see drivers being a critical component of this equation. I think over time, we can provide additional services as part of our platform to our passengers that these drivers may be interested in.
Is this a conversation that Lyft has had with any drivers?
Yes. We’ve been proactive with our drivers. We have an advisory council that works with them. I think it’s one of these things that we want to continue to work with them. Again, this is a transition that is going to take a significant amount of time, and I think there’s a real opportunity here to work with our drivers as we always have on what this future is going to look like.
What has been the feedback so far? What have you heard from drivers? It’s actually interesting you brought up the car maintenance part of it, because in conversations that I’ve had with drivers, they think that self-driving will never happen, because you need someone to clean up the vomit. That, in your mind, is the biggest obstacle to self-driving, which is interesting. What have you heard from drivers?
I’ve had a number of conversations with drivers. On one side of the spectrum, they’re dubious. They’re not exactly convinced that this technology will evolve to a point where it can service passengers at the level that they think it needs to. Then I think there’s others that are actually quite excited about this technology change and actually want to be a part of it, and do see the opportunities that may come out of this as these autonomous vehicles come to market.
Gotcha. This is more related to drivers than it is to autonomous, but Uber is kind of infringing on the driver-friendly space. They have this 180-day driver campaign, rolling out a lot of features, some of which Lyft already has. First, what do you make of the campaign in general? Do you feel like it’s a genuine effort? What do you make of the changes that they’re rolling out?
They seem to be following a lot of the efforts that we’ve done for a number of years. I can’t speak to their efforts, but when we look at drivers, they have been core and part of our family. We’re going to continue to improve that experience. I look at it as an opportunity for us to continue to differentiate beyond I think what they’re trying to do.
Right. Lyft does have plans to roll out new driver products and features and things like that?
Absolutely. Yeah, we have an entire team, if you look at our product organization, we have an entire team that’s focused on the driver experience.
Today, let’s say I’m interested in driving for Uber or Lyft. Convince me to drive for Lyft.
I think the best question would be, do you have a car?
No, I don’t. I’m a New Yorker.
Great. Fantastic. We have this thing called Express Drive. Express Drive, we can provide you actually with a vehicle. You can almost imagine the minute you sign up, I can hand you the keys, and you’re basically driving that vehicle now for Lyft, unless you have a car.
I think we’ve already highlighted that. No. The other piece is, again, if you look at the community that we have, the driver education, there’s also a number of things that we do, in terms of driver incentives to help you as a driver to get up to speed on our platform. There’s tipping, of course, as well, which I know our competitor is now jumping into the game as well.
I love that none of you ever say the name of your competitor. This always happens, but yes, go on.
I don’t know what you’re talking about.
Last question, what’s next for Lyft’s self-driving announcement? Are you focused just on establishing partners? What’s coming up?
No hints, other than stand by. There’s a lot of exciting activities that we have planned for this year, both on the platform side and the self-driving system.
Great. Taggart, it was great talking to you. Thanks for coming on this show.