Updated: Apr 17, 2019
This episode features Dr. Wei Luo, Chief Operating Officer and Head of Product at DeepMap, a startup aiming to solve the HD mapping and localization challenges for L4/L5 autonomous vehicles. Previously, Wei worked at Google for 9 years leading product and engineering efforts for Google Earth, Google Maps, Geo Enterprise, Location Intelligence and more. Wei received her Ph.D. from UC Berkeley in GIS and Remote Sensing.
Wei shared her career experience as a product manager and COO, as well as her insights on product development in HP mapping for autonomous driving vehicles.
How to define the role of COO?
Challenges in HD mapping for autonomous vehicles
Robin.ly is a content platform dedicated to helping engineers and researchers develop leadership, entrepreneurship, and AI insights to scale their impacts in the new tech era. Sign up with us to stay updated and access exclusive event, career, and business mentorship opportunities.
Wenli: Let’s just jump into it. So you received PhD from UC Berkeley in GIS (Geographic Information System) and Remote Sensing, and you could be an engineer, but instead you took a different route and became a product manager. What really interests me is that from my perspective, product manager is kind of a role that requires more soft skills and communication and business insight, perhaps is even harder than just engineering.
Wei Luo: Personally, I think it is harder.
Wenli: A lot of PhD graduated and started doing research or engineering, so what made you choose a different route?
Wei Luo: Actually for two years I was in two different consulting companies, first playing on a maybe very technical role. And then later on I started managing a team of people and then naturally I got into kind of the responsibility of managing not only people but also projects as well. After my consulting firm career, I started working at Google in Google Geo. So again naturally, I transitioned my role from a more technical driven role into team management and then communicating with various other teams that we have to collaborate with. It is more challenging, I mean for my personal path, that transition kind of happened naturally. For others maybe it's a harder jump if someone wants to make that career change. It's easier to go through a transition period rather than having a sudden change.
Wenli: Yeah. You worked at Google before. The product you worked on was Google Map and Google Earth. So what were some of the obstacles and the hardship that you had to overcome in order to become a successful product manager?
Wei Luo: Without citing all the hardship that everyone has to go through, some of the things that - let’s say I remember very clearly is that a very specific example is actually a fun project that I remember to date, which is overlaying - we call it a Time Machine project and then team came up with this fantastic idea that what if we can actually allow earth to display historical data as well so you can go back in time? Therefore that project was called Time Machine. Anyhow, I was very lucky that I was able to work with a very well-known historic cartographer. His name is David Rumsey. He has a huge collection of historical maps, some of them are hundreds of years old. We wanted to start building a mechanism to overlay these very old maps onto Google Earth, so that you can actually see earth back in time, right?
There were a lot of challenges related to that: Number one, how do you actually build this concept of letting earth go back in time and that technically was very very hard. And then secondly, just going through the launch process of expecting all the - let’s say if this project is very successful, there will probably be a lot of traffic or a lot of instantaneous traffic coming from users, where everyone wants to pick and look at a particular map, which - again, it's a very huge technical challenge, but also requires a lot of coordination not only with the engineering team but also with the marketing team. So that was one of the few examples in my earlier career path. And then suddenly, I realized that launching a product, oftentimes people think: Oh, this is a huge technical challenge and let's focus on that. But actually going through this big launch process made me realize that engineering probably only accounts for 50% of the success, how you actually define this launch from the right angle, how you sequence that, how we can actually coordinate all these different job functions to make sure everything goes smoothly and then have that perfect launch and big splash at the end requires a lot of forward thinking, a lot of meticulous planning and a lot of planning to deal with almost that aftermath.
Wenli: That project sounds really interesting. If I see it now I would like to see…Is it still up?
Wei Luo: I haven’t checked. I'm sure like David Rumsey himself still actually has his collection that can be viewed on Google Earth today. But yeah, I love it.
Wenli: Yeah. For each project, a product manager puts lots of heart and sweat into it. It's amazing. I know that you’re a bilingual. You speak Chinese and English and you are also a woman. How did you turn your language skills and work as a woman in tech into your advantages?
Wei Luo: Well, I think being bilingual is definitely a big plus. I think in today's world actually, the more languages you can speak, the more culture awareness that you have, the better. I think having that perspective or perspective from both sides really helps. Especially in a startup world, you have the chance of building a business that has global presence. Having not only the language skill but also the cultural understanding on why certain people behave certain way and why people want to conduct a business or engineering design in certain way, really help bring the teams together.
Speaking of gender, one thing that I find being a women is that, number one, I forget there is a gender difference or at least I don't pay attention to it; number two, being a woman and actually being a mom of three, I actually do feel that this is a advantage that I have. I find myself to be maybe more sympathetic to other people's issues. In general, I do observe at least, women are more meticulous, they are more detail-oriented. And when you actually have a team, some of the best software engineers or product managers or engineering managers that I have met are very strong women. Having women in a team oftentimes changes the dynamic of the team.
Wenli: Any stories about the balanced team?
Wei Luo: Well, there's no magic ratio. The better diversity is always good. That rule does not just apply to gender or nationality. Diversity is always helpful because again, when you are building a product, you want to satisfy a broad customer base. And we are working with teams or your customers, you also want to have that diversity to help.
Wenli: You are clearly a very experienced product manager now. So you joined a HD mapping startup, DeepMap as their COO. I don't know much about the job description of the COO yet, but one thing I know is that you and the CEO have to be a good match at work. So what's very important about the relationship between CEO and COO, and what was your thought when you joined DeepMap?
Wei Luo: I think what's maybe particularly important between C-level executives is that the executive team set the tone and the direction for the entire company. so it's very very important that the management level have unity and have a common understanding on the direction of the company and the direction of the product roadmap. If there's a gap or if there's a divergence between the higher or upper level management, it will be very very confusing to the engineers and other people that are implementing the technology. So Trust is very important, to have consensus and common understanding. And how you do that comes down to a trust that you build over time. In this case, when I joined DeepMap, James Wu, our CEO, we were actually long-term friends. We have known each other for a very long time.
Wenli: How did he persuade you to join the company?
Wei Luo: Again, we have known each other for a really long time. And actually I not only knew him at the time but I knew pretty much everyone on their team at that time. But there's already that bond that we didn’t need to create.
Wenli: When we're interviewing CEOs, when we talk to CEOs, we talk about strategic plans, we talk about the motivation and the inspiration that the CEO can bring to the team and create these iconic operations. But the role of COO, the chief operating officer is kind of less clear defined. So what exactly does a COO do?
Wei Luo: So I think that's actually maybe intentional. At different companies, COOs can play different roles, and a COO can play different roles at different times in the same company as well. As the second hand in command, you have to be very flexible and be very adaptive to fit the needs of the company at that given time. For myself, for instance, my role has changed a lot in the last two and a half years. At the beginning, I wore many many hats. I was doing recruiting first hand; I was scanning all the resumes; I was setting up the phone calls and I was interviewing candidates. I was taking care of a lot of General and Administrative functionalities. And at the same time, I was wearing the hat of our product management, project management. And I did a lot of first hand negotiation with customers, so doing business development as well as sales. And I did all the customer support. So that's a lot of hats to wear. As the company gets bigger and bigger, I need to focus more. So today again, as the company grows, I will change my role from one job function, or my focus area at least will change from one job function to another. And today I still do a lot of business development, I also manage engineering teams, I also manage product management. And I can say in the future, depending on where the needs are and how the company morph itself in the future, my role will probably shift again, and that's totally okay as well.
Wenli: The job that you do is constantly changing with the product development. Does that only happen in the startups, or it happens when you work at big companies like Google as well?
Wei Luo: From what I observed, again in different companies the COOs may have different roles as well. And it depends on the background of this particular person. Just take one example, at Facebook, Sheryl manages more of the business and the people operation functionality, work on their relationship, because she has a bigger or stronger business background. And for myself, I came out of a technical background, so I do pay more attention to product development and engineering and so on. As a COO, you have to be okay with such changes and be okay with the flexibility or the demand to change over time.
Wenli: For startups and entrepreneurs, when do they know that they need to hire a COO and someone with the product background?
Wei Luo: When the CEO is too swamped. I think because the COO’s role is so flexible and can change over time, maybe you can always find a person that can fill the gap between what the CEO is able to do and the rest of the company. If there is a gap between what CEO’s either the capability or the bandwidth and then you need to bring a person to supplement either in a bandwidth or the capability. That’s probably the right time to think about getting a person to help you.
Wenli: Do you have an example working at DeepMap, like how you supplement the CEO’s job?
Wei Luo: Yes, a lot. So I think between our CEO and myself, we oftentimes switch roles. To give you a specific example, we actually closed our Series B maybe a few months ago. It was a very successful round, but it took a long time, a lot of time or time commitment. At the beginning, I was doing some fundraising activities with our CEO, but then very quickly, we decided, that's not gonna be scalable. So he ended up spending more time doing fundraising, versus me like taking over the entire operation of the the team to make sure our R&D was going smoothly, our customer projects were going smoothly, and then we had a clear road map. And then the other fundraising activities were sort of over, he could actually spend more time and look at the internal operation as well and then we would divide out the responsibilities again.
Wenli: Speaking of fundraising and also speaking of the startups, I know that DeepMap was invested by lots of venture capitals like a16z. But I also know that they approached you first. For some startups that are trying to pitch messaging all the VCs or making connections, what is your experience and suggestions for them?
Wei Luo: Silicon Valley is all about networking, which is again probably a universal rule everywhere. But in particular, as a start-up, if you’re trying to make the very first step, there are a lot of incubators as well as VCs who actually regularly host these startup pitching sessions. Depending on which space you're looking at, you should go there and then do your lightning pitch. For self-driving, for instance, there’re a few organizations that actually organize these bi-monthly pitching sessions. Another observation I have is that, again good people know good people. If you are trying to make the very first step, do talk to people who are already in the start-up world. It’s very likely that they already know some VCs then can connect you right away. Taking my personal example, now I’m in the startup world, actually quite some people come to me and ask me like: Wei, I am particularly interest in this space, do you know any good VCs that are investing in this space? And if I know the right people, I'm more than happy to make the intros for them as well.
Wenli: DeepMap is aiming to solve the HD mapping and localization problems for L4 and L5 autonomous vehicle companies, but there are still a few similar companies out there. So how DeepMap differentiates itself?
Wei Luo: Honestly, I think we have the best team in the world to tackle this problem. The reason is that mapping requires very special talents and a very special knowledge. It's not a technology field where you can just dive into it without prior work experience. And very fortunately, we have been able to assemble some of the best talents who have worked for many years on building maps, building not only maps for consumers but also building maps for enterprises as well. The other aspect is technology. One thing we realized from the very beginning is that building the right HD mapping technology takes a very sophisticated design, so we did not dive into it right away, at least we did not trivialize how hard the challenges would be. We know HD mapping for self-driving is extremely hard, it's harder than anybody can imagine actually. So we took a lot of time to design the right product from the beginning. And our company's mission is that we want to build a map product that is extremely safe for self-driving cars to use, and therefore our design is targeting true level four for self-driving cars. We want to solve the hardest problems first, so that we know that our solution can be safe and then we look at how we actually scale this up and bring the cost down. What we have heard from our customers is that our technology is great. They like to work with us not only on the small-scale projects, but they want to work with us to scale up their operation.
Wenli: Nice. I'm so glad to hear that. What are the current problems that you're working on right now?
Wei Luo: For our company? Well, without diving into the technical details, I think the biggest challenge for HD mapping is solving the quality problem. Because HD maps are ultimately maps made for self-driving cars to consume, and it's absolutely not allowed that you have mistakes in your maps and therefore leading to mistakes in self-driving decisions. So having extreme high quality and then the ability to verify or check the quality, meet the requirements is a forever challenge that will go on for a long time.
The other challenge that we’re working on is scalability. And by scalability what I mean is that, we not only need to make maps for a small area, but we also need to make maps for a very large geographic region and be able to maintain the data. So mapping a small area is already hard, mapping a big area is harder, and the ability to maintain this map and accurately reflect all the changes that you see on the road and do it very fast is very very hard. So these are the two technical challenges that we are working on.
Wenli: Thanks for sharing that. My last question is that, I would like you to share your experience and your knowledge with our engineers in the community and also the future entrepreneurs. Any advice you can give them?
Wei Luo: I think there’s only one: Be very very passionate about what you do.
Wenli: How do they get promoted?
Wei Luo: Be passionate. If you're not truly interested in what you're doing, you are probably working on the wrong thing. Find something that excites you, be very persistent about it. If you deliver high quality work, people will see it, your peers will see it, your manager will see it. It's not about building a fancy PowerPoint, it's not about lobbying other people. I mean having visibility is definitely necessary. You should not hide yourself in a corner and work for a year and then disappear for a year. Communication is definitely important. Transparency is definitely important. But at the end of day, it's the outcome of your work and then you need to be very passionate about it. You should be working on it because you want to work on it, not because you are working on it just to get a promotion.
Wenli: Well, thank you very much for sharing everything.
Wei Luo: You’re very welcome.