Formed in Marseille in 2010, peer-to-peer car-sharing startup Drivy has grown into a business serving 1.5 million users with around 50,000 cars available for hire across France, Germany, Austria, Spain, Belgium and -- most recently -- London. The company's next priority is to expand its operation throughout the UK during 2018. Drivy currently employs 100 people and is headquartered in Paris, with regional outposts in London, Berlin and Barcelona.
In February, ZDNet met with founder and CEO Paulin Dementhon, and chief development officer Patrick Foster to learn more about Drivy's business model, growth to date and plans for the future.
The basic idea of Drivy is to increase the utilisation of an often-idle asset -- your car -- via peer-to-peer sharing using a web-based/mobile platform. How has this market developed?
In Europe there were 50 or 60 companies trying to do this, and 15 in France alone, but we managed to make it. Now it's down to around 10, but very few are of any significant size -- we're number one by a long way. These marketplace businesses are a bit 'winner takes all': maybe there's room for two or three, but not tens of companies. So once we started raising a lot of venture capital money it was hard for the other players to keep on getting finance and investing in large teams and so on.
Cultural acceptance of a new business model is important: did customers quickly 'get' the idea of peer-to-peer car sharing?
Of course the first question that comes up is: "If a stranger is going to drive my car and it gets damaged, what about my insurance?" The number-one problem was finding insurance, and then convincing car owners that we were a serious company. Probably half of the 50-60 startups I mentioned failed because of insurance and operations issues. Many had good tech teams and a nice app, but the back-office operations -- not getting the car stolen, not having too many incidents etcetera -- lots of tech people didn't really care about that.
Insurance is obviously a key success factor: how did Drivy's relationship with Allianz come about?
Actually Allianz said 'no' when I started Drivy, and I had to run for three years with a local French insurance company. After seeing how we were developing, Allianz accepted -- they were actually watching us more closely than we thought. What's interesting is that the relationship has now totally changed around: I can see the CEO whenever I want, and Allianz has recently created a Sharing Economy department inside the company. The whole industry knows that within five to ten years everything is going to change, from insuring cars to insuring usage. So now they consider it strategic, and they want to insure businesses like ours.
Would it be accurate to describe Drivy as 'Airbnb for cars'?
That's true: we are a marketplace, we have reviews, our business model is we take rate on every transaction, we have listings, owners set the price -- it's very similar to Airbnb. But actually, our mission is more to become a 'Spotify for cars'. What's the difference? Once you go on Spotify, you can ditch your CD player and CDs and you just use Spotify to listen to music. Our objective is the same: making it so convenient and so easy that, if you've got, say, 10 cars within walking distance, then you don't need your car anymore.
For Airbnb, it's not their objective that you live 100 percent on Airbnb; it's a travel thing. Half of the car rental market is in airports -- another travel thing -- and we don't target that. We do a kind of riskier bet: if it's convenient enough, then you can replace ever buying your first car and operate only on demand. That fits really well with the future of connected and autonomous cars.
My vision is, it will be less peer-to-peer and increasingly fleets of cars -- people who buy or already have fleets of cars. Today, peer-to-peer is probably 80 percent of our business, but I think it's going to be the opposite within a few years: 70-80 percent fleet and some private owners to complement these.
How does Drivy compare to a company like Uber?
What we focus on is having the best experience, plus stickiness and customer relations, so it is becoming more similar to Uber -- without the company culture issues! But it's a different use case: the big fight today in on-demand transportation is getting from A to B in cities, but we focus on outside cities -- weekends, trips and so on lasting two days on average. We don't think car sharing is competitive inside cities because ride-hailing is more convenient -- you don't have to think about parking, or taking pictures of the car and so on.
Returning to the insurance question: how often do incidents happen during rentals -- is that something you can share?
About one percent of rentals have 'something' reported, and the whole thing is managing this one percent -- which is below the industry average. If this number increases the model is still going to work, but the insurance cost will go up. With our model of super-convenience and proximity, what we want is repeat customers: the more repeat customers you have, the more respectful they are compared to one-off users.
And is your repeat customer rate where you'd like it to be?
It's improving. When I launched in France, growth just happened -- cars would pop up without any control over how we did that. Now, as a more mature company, in the UK we decided to launch only in London to really focus on getting a critical mass. If you have cars here and there across the country, people cannot repeat.
Drivy, by the numbers
Talking of expansion, Drivy has raised quite a lot of venture capital -- the last round was 31m euros: are you profitable now, or is more VC needed?
We didn't start with a 'freemium' offering: to give you a sense of the economics, we have a business model with every rental giving us a margin of about 10 euros, so we could stop raising funds and make the company profitable. But there's a huge potential market in the switch from ownership to on-demand, and we think we can be among the four or five apps that move billions of people, on the specific use case I've described, so all our investors want us to go faster. If we keep on expanding at this pace, we will raise a new round to invest more in the team and the tech for geographic expansion and so on.
What were the main challenges in building the business, beyond finding an insurance partner: was it hiring the right software designers and engineers, support people...?
The main issue is on the supply side, to get a good level of quality and engagement of car-owners on the platform. It's fairly easy to have people sign up, take pictures of their car, set a price and so on -- to make a listing. But the tricky part was, in the beginning we got tons of listings, tons of cars, but if you asked 'can I rent your car?', the answer was 'yes' only 20 percent of the time. So the experience was pretty disappointing for customers: when you're looking for a place to stay on vacation, you're willing to put a bit of effort in, but for a car it's like 'just give me a car, otherwise I'll go and rent one'.
On our side, we need to make the business profitable for car owners, to make them really engage, and not have inactive owners on the platform. It doesn't seem like rocket science, but this management of the marketplace is really key. That's why, in London, we're offering new owners a guaranteed £250 a month for listing a car for at least 17 days a month -- it's a way of showing we're a 'no bullshit' operation. Also, by giving this guarantee we can require owners not to set a price that is absurd, and also to put our [otherwise optional] Drivy Open technology in the car from day one.
How does Drivy Open work?
It lets you hire a car when it suits you, without having to arrange to meet the owner. First we send someone to install a box in the car containing a GPS, GSM connectivity, an accelerometer -- an IoT device. It's a 45-minute installation. Afterwards, the car owner needs to update their calendar showing when it's available, and renters need to have their licence and identity verified. Then we show them the car's location on the app, they come to the car, do an inspection and unlock the car with their smartphone. Our software handles the 'digital handshake' -- if you've ticked all the boxes, then you can access the car. Via the box we can read the fuel level and the mileage, so we can charge the right amount.
The box is really just a bridging technology: very soon, you'll just be installing the Drivy app on the car's OS. When you install, it will ask for authorisation -- just like a smartphone app can ask for access to the camera. We're starting to get into conversations with car manufacturers and dealers, because we want to promote Drivy as a way to finance a car. Then it'll be nothing more than ticking a box -- 'I want to sign up for Drivy' -- and putting it on the marketplace. Today, having to convince people to put a physical device in a car...that's why we offer the £250 guarantee, to persuade them that it's serious and worth the hassle.
If we take a step back and look at car sharing, it never really took off big-time -- I mean, it's been existing for 15 years, with nice growth rates of around 20 percent a year. But if you compare it to ride-hailing, to what happened with Uber and Lyft and all those guys, they went from zero to billions in just a few years. It didn't happen with car sharing because building a critical mass of cars is costly and it's still a bit complex. But I think with what's coming with connected cars, this is going to be the game-changer. You'll be able to turn any car into a share car -- that's what all car manufacturers say now: 'we are not manufacturers, we are mobility providers'. We're in a good position now, with nice volumes -- 50,000 cars -- but if we keep our leadership position, and thinking about what's coming in the automotive sector, it can do 10x very quickly.
Obviously when cars and the IoT are mentioned, worries about cybersecurity usually come up. Has Drivy had any problems in this area?
Actually we've never had anything, and I think it's much more difficult to steal one of our cars than a standard car. But it's obviously an issue, especially as car manufacturers open up their APIs -- and obviously when it's autonomous driving, the consequences of hacking are much worse than just opening the doors, for example.
How do you see the 'mobility as a service' space -- and Drivy's place in it -- developing over the next few years?
I think the evolution will come along with the evolution of cars. So with autonomous, or even semi-autonomous cars, the convenience and experience of driving is going to improve, as is the market. Basically cars will become robots and will specialise, because it's all on-demand, so you'll have mobile living rooms, or mobile offices. It's hard to predict exactly what all this will look like, but I think it will be focused around having robots serve you for a certain amount of time. It will be different from just getting you from A to B.