This article was first published in CoFounder No 11 in February 2018.
Eric Gundersen, co-founder of Mapbox, pulled together a $164 million investment round for a startup that was born out of international development aid projects.
When Gundersen got into development aid, he soon saw the potential that technology could bring. “None of the stuff was digital, they weren’t using any tech. There was not a data-driven approach to huge decisions,” Gundersen said, describing the beginning of the business during an interview on the sidelines of Slush. “The whole point was how to get decision makers to touch the data and see this stuff. I mean the interface: the map is a canvas to analyse, and I wanted to put that in front of the decision makers.”
It did not work as well as the team hoped – mainly because of the lack of maps in those austere environments. However, there was a surprising benefit. “We got good at building the map part,” Gundersen said.
The team, working in remote areas of Pakistan and the Congo, invested proceeds from other projects to create better tools. Eventually, they built a tool called Studio, which Gundersen describes as Photoshop for maps, making it possible for doctors to show how malaria spreads, for example.
The tool was not only useful for special development aid projects. “By 2013 we were powering the maps for Foursquare, Evernote and US Today. There were a lot of people that wanted to be able to design custom maps with a custom look and feel,” he said.
“We realised that this was the moment to go further. We went out to raise funds and got insanely lucky on timing. In July 2013, Waze was bought. We then closed in September 2013. For our first money in, we raised $10 million,” Gundersen described the decisive year.
The team scaled quickly – growing almost overnight from a 35-person team to 90 people. “Every quarter started feeling like a year. The need for location in more and more applications was growing. It was very easy for us to start monetising on the web, getting some initial traction.”
In June 2015, Mapbox closed their B round, raising $52.5 million from a group of funds led by DFJ Growth. “That allowed us to really scale, not only going properly to market on the mobile side, but starting to build up a platform that’s now going to allow us to go into auto. That will allow us to go into augmented reality (AR) and virtual reality (VR). The best part was: we kept making money,” Gundersen described.
Mapbox has benefited from growth in a sector with only a few suppliers, with financing rounds further boosted by low interest rates.
In October 2017, the firm unveiled a whopping $164 million investment round, targeting the global market, AR/VR and autos. “It’s an awesome time to raise right now: welcome to what zero interest rates look like! This is not exuberance, this is actually rational economic behaviour, and this is a very good time to go out to raise capital,” Gundersen said. “It’s shockingly easy [to raise $164 million]. Series C was easier than Series B in a lot of ways; you only raise it from people that understand what the hell you’re doing. Series B cleared the way so that we only started talking to people that we trusted.”
The round was led by Masayoshi Son’s $93 billion Softbank Vision Fund. “Everything you’ve read about him is true. He’s crazy. We sat down and started talking not about maps, but about how cities were going to be planned: how people were going to move, how they think about the ride-share space, how they feel about the logistics space, how they think about the processor space. His investment thesis is simple: singularity. He’s not messing around. I was blown away, not just by his depth and how he sees all these pieces fitting together, but also by his team.”
The last four years have been quite a ride for Gundersen personally, too. “It was four years to the week that we closed Series C. That took me from being personally a quarter of a million in debt, with one working credit card left and about three or four hundred dollars on it, and all my lines of credit drawn out, to now having a 250-plus-person company.”
The firm has nearly a million registered developers using its mapping platform, with companies like Snapchat, The Weather Channel, Tableau, IBM, DoorDash, Mastercard and Instacart, among other customers.
Mapbox has carved out a unique market position – it is one of four companies holding world map data. Google is the dominant player in the sector; the other two firms – TomTom and Here – are heavily dependent on the auto industry.
Mapbox uses open databases, like OpenStreetMap and OpenAddresses, and combines them with a vast amount of sensor data it has access to through its partners. “Our maps reach 350 million people a month. Just like Waze, we collect anonymised, aggregated telemetry. We’ve been able to build up a kind of live map, and that puts us in a really interesting position when the world is about to become this insane, decentralised network of sensors,” Gundersen said.
The mapping sector has seen a lot of takeover rumours and some large deals. Google bought Waze, Nokia bought Navteq and then sold the mapping unit to a consortium of German carmakers, while TomTom acquired Tele Atlas. Here has a strong position in the auto industry, due to Navteq’s heritage, and it often competes against TomTom.
“In the mapping space, the upside is scarcity. If we get to the point where it goes from 4 to 3, or the 2 competitors that aren’t collecting live sensory data become stagnant, it will get super-interesting,” Gundersen said.
With the constant inflow of live data, Mapbox sees itself closer to Google than to Here and TomTom. “We have different business models: they sell their data, where we provide services with data and software all in one, because the map needs to learn. The only one that really gets the living part is Google. With Waze they only understand the sensor play,” he said.
Mapbox has positioned itself more as a mapping platform for anyone to use. “We provide building blocks, almost like Lego, for you as a developer to build your own custom world,” Gundersen said.
The move into autos will probably be welcomed by consumers – just imagine if your dashboard map worked as well as Foursquare. “We are trying to give Detroit the same tools that we give to developers, so that maps stop sucking in cars,” Gundersen said.
He explained that even though everyone speaks about self-driving cars and connected cars, very few of those vehicles are collecting live data for maps. “It’s shocking how few connected cars there are on the road today. We have more sensors on the road today than the entire connected car space will have in 2020 … people have been way too pragmatic about who pays for the bandwidth going into the car.”
With a strong followership in place, and a lot of apps using Mapbox data, why has no-one bought Mapbox?
“We’ve not been for sale. This has been a long play. We are super-fortunate in our capital structure, being independent for so long that it reduced a lot of pressure.”
“Where you see the classic mapping play, you get to 20-30 people. We were able to get past that cliff and work really hard to hit revenue milestones in our growth round. It worked. I think the defining moment happened last summer. Now we can go and build a long play.”
Today’s Mapbox is a well-funded startup that is seeking top experts to help it grow further, and Gundersen is not nostalgic when he looks back at the early days of the team bootstrapping, using their income from consultancy projects to develop the mapping tools.
“It sounds really cool, but it sucked. Making payrolls sucks. The reality is that you are working so much that you barely have any friends outside those you work with, and your ass is on the line for running their payroll.”
At the same time, it helped to build the team. “We cheated for the first two years, after getting funding. Going from 35 people to 90 that quickly was crazy; doubling again after that was crazy; but we had enough strong core leaders on the team and trusted each other that they could help create confidence,” he described.
“When you’re running a team, you’re basically getting people on your team to suspend reality for a second. What you’re doing has not been done yet, and you have to convince them that it is simple, despite how daunting it is. The fortitude of that team in the early days was amazing.”
Unlike many early-stage companies, which focus on processes and command-lines early on, Gundersen and his team built Mapbox into a 200-plus-person company before creating a leadership structure in early 2017.
“I think younger companies roll out structure very early, making it complicated to do reorgs, stifling teams and so on. That said, I think we took the other extreme, which was not healthy,” Gundersen said. “The trick is to have smart people; they will make the right decisions. But you do need some structure.”
“I want to bring in people that have been there and done it before at different levels. What I’m interested now is the global side. I want people who have played this game in other arenas, because we’ve got a lot more building to do,” he said.
With the mobile industry going deep into AR/VR and the logistics field around the company changing fast, Gundersen is convinced that the company is in the right place at the right time to switch off reality for a while: “It’s a junction point where everything is changing.”
The tech industry has talked for years about the promise of AR and VR. Gundersen believes that the steps taken a few years ago by Apple and Google to bring technologies closer to users made the decisive difference to the situation. “People now realise that this can be done right. That this is really interesting. I think a lot of the ‘done right’ part is not being as visually obvious,” he said.
“A successful navigation experience will get you there in time. If we were going to hop in the car and I was driving, you would give me directions. You would give me the perfect amount of information to get there, and if I did something wrong you would correct me. That’s what AR will allow us to do.”
Gundersen believes that this technology leap will make the whole sector more interesting again. “It’s going to be fundamentally different. The mapping space has been boring for a while. We haven’t seen a lot of innovation.”
Photos: Mapbox; and Samuel Chan, Deva Darshan, Samuel Edwards, Ivan Bandura through Unsplash.