“It didn’t take a genius to realize that if I couldn’t find any useful online information for finding parking in San Francisco, then drivers in Paris, Shanghai, and thousands of other cities around the globe were probably also struggling with the exact same problem,” explained Eugene Tsyrklevich, founder and CEO of London-based parking platform Parkopedia, in an interview with VentureBeat.
It was way back in 2007, during the Web 2.0 conference in San Francisco, that Tsyrklevich tried to find a place to park his rental car before his arrival at the event. But after Googling for reliable parking information for the Moscone Center, he came up blank. So he left home earlier than should otherwise have been necessary and circled the block several times until he finally found somewhere to park. “That was the spark for the idea — helping drivers find somewhere to park — which I launched in 2008,” Tsyrklevich explained.
Eugene Tsyrklevich is a computer engineer who holds a BSc and MSc in computer science from the University of California, San Diego. He initially learned how to code in high school and went on to fund his university education by working as a software engineer at a number of tech startups during the late ’90s dotcom boom, which had exploded while he was at university. He started his first business, a consulting company specializing in computer and network security, just after the dotcom revolution went belly up in 2000. “I had the opportunity to travel the world, get paid for legally breaking into corporate networks, and to train the U.S. military in hacking — all before I graduated from university.”
After graduating, Tsyrklevich continued consulting, and he went on to work in finance after moving to London in 2004.
Fast-forward 10 years from that frustrating day outside the Moscone Center, and Parkopedia is now one of the biggest parking data compendiums in the world. Headquartered in London, the company describes itself as a “Wikipedia for parking,” and — through a suite of apps — serves up information on availability and pricing of parking meters, parking spaces, parking lots, and even private driveways for hire.
Much has changed at Parkopedia since the original incarnation was unveiled back in 2008, as the entire connected landscape has shifted from desktops to the pocket, with many cars now permanently online too. Moving with the times as a fledgling startup is always difficult, but when you’re completely new to an industry, things are just that little bit harder.
“When I had the idea for Parkopedia, I knew nothing about the parking industry, maybe apart from how to feed a parking meter,” said Tsyrklevich. “The first version of the website, which I built myself, allowed drivers to find parking lots in San Diego and San Francisco. Within months, we expanded to London and Sydney, and we have carried on adding cities ever since.”
Though Parkopedia is a consumer-facing service in its own right, claiming around three million monthly users, it’s not a household name like Uber, Google Maps, or some other transport-focused services. But it has achieved scale by powering countless third-party services from major global brands. Today, Parkopedia says that it serves more than 6,000 cities across 75 countries.
In 2015, GM subsidiary OnStar launched AtYourService, a new “commerce and engagement” product that connects drivers with merchants during drives — serving up information and offers related to their route and destination, including hotels and places to eat. Parkopedia was one of the launch partners, used by OnStar to give subscribers in North America and Europe a “premium parking service” that includes dynamic space availability. This week, the duo renewed their deal, which will extend the partnership by another three years.
Back in 2011, Parkopedia added Garmin to its Roster of clients, powering the GPS tech giant’s navigation devices with such data as real-time parking space availability. This was followed a year later by a global deal to power BMW’s in-car parking information service.
Last July saw what was arguably Parkopedia’s pièce de résistance when the company inked a global deal with Apple to power parking information in Apple Maps. And shortly after that, Parkopedia entered an agreement with TomTom to provide data to its own automotive customers, followed by Audi China a few months later. Throw into the mix Ford, Jaguar, Land Rover, Peugeot, Toyota, and Volvo, and you start to get an idea of Parkopedia’s reach.
“With the continued growth of Parkopedia’s coverage, as well as explosion of connected devices and cars, Parkopedia grew rapidly and now works with most car manufacturers globally,” explained Tsyrklevich.
Show me the money
What’s perhaps most notable, in this age of unicorns and failed tech IPOs, is that Parkopedia has yet to take on a single dime of venture capital. Moreover, the company says that it’s profitable, making enough money through licensing its services to its business customers and funding everything from its cash flow.
“When we were starting out, we did consider taking on VC investment and pitched to a number of firms,” said Tsyrklevich. “However, we didn’t raise any VC investment and ended up bootstrapping to profitability.”
Whenever a major client is announced, says Tsyrklevich, his company gets a lot of interest from VCs, but he isn’t interested at the moment. “Being profitable and funding our growth and product portfolio development organically, we are currently not actively looking for money,” he said. “This may change in the future with potential acquisitions or entry into new markets.”
Parking in the digital age
If you’re the sort of person who just shows up at your destination, fingers crossed, hoping to find a parking space, then here’s a quick recap of what kinds of parking data is typically found in Parkopedia and similar services.
First up, there’s dynamic, real-time information on available spaces in barriered parking lots — this data is garnered directly from the companies that operate the parking facilities and is complemented by “static” data, which covers such variables as prices, opening hours, and location. But Parkopedia also manually checks much of this data. “On top of the extensive industry network that Parkopedia has built out, we physically re-verify all of the data by deploying field teams globally,” said Tsyrklevich. “This requires significant investment, but that’s the cost of delivering a highly accurate service that is expected of us by our premium brand customers.” In other words, Parkopedia can’t afford to take the data it receives at face value.
Additionally, Parkopedia serves up static data for on-street parking in 500 cities around the world, while also providing some real-time availability for on-street parking where sensors are installed. For example, in the City of Westminster, London, Parkopedia receives sensor data from Smart Parking, which provides information about unoccupied on-street parking bays in the city center.
This is an expensive and labor-intensive initiative to scale globally, however, and getting cities to invest in smart sensor technology is a hard sell. “Some cities, such as Los Angeles, London, Shenzhen, and Barcelona, have rolled out sensors in their cities, at least in core business district areas, but they are the exception, despite sensor technology being around since 2007,” continued Tsyrklevich.
Given the inherent challenge of garnering real-time data for on-street parking, Parkopedia has been investing in algorithms to predict when spaces may be available where there is no physical counting infrastructure. Back in September 2015, Parkopedia announced a partnership with Garmin, which became the first company to use Parkopedia’s predictive parking smarts, starting in a handful of cities across Germany. Rather than relying on sensors or other expensive hardware, it analyses historical data garnered from cars that already have Parkopedia integrated. It’s not perfect, but it claims around 80 to 90 percent accuracy.
It’s actually not too dissimilar, in concept at least, to a predictive parking feature that was recently introduced into Google Maps in a handful of U.S. markets. Google collects the data in much the same way it obtains data for its Popular Times feature in Google Search: by aggregating anonymized data from Android users who have “Location History” activated on their device. While Google’s parking feature is no doubt useful to have, it doesn’t quite hold the same sway as Parkopedia’s predictions, because Google doesn’t know for sure whether the parking is legitimate, and the app lacks other fundamental data points, such as pricing and hours of operation.
Parkopedia currently employs more than 60 people across its hubs in London, San Diego, and Beijing, and Tsyrklevich says the fastest-growing area for the company right now is its data science division. This burgeoning team includes PhD scientists with backgrounds in robotics, computer vision, and computer science who are working on “statistical models” — among other future-gazing projects — to garner accurate parking availability predictions.
Since the predictive parking feature’s initial rollout in Germany’s five largest cities 18 months ago, Parkopedia has signed more deals that have given it access to additional car activity data, and it’s now available in around one hundred cities across Europe and North America, according to Tsyrklevich. “We now process hundreds of millions of data points every single day to figure out parking space availability,” he said.
Parking data landscape
Parkopedia isn’t alone in the parking data / service provision realm. Other notable players include Chicago-based SpotHero, which has raised around $28 million in funding since its inception in 2011 and which connect garages, lots, and valets with car owners across the U.S. Then there’s fellow Chicago-based ParkWhiz, which has nabbed $36 million in venture capital funding since 2006, but again it only operates in the U.S.
Parkopedia has local competitors in most major markets, with some offering in-app bookings and payments, while others serve more as information and data providers. Though Parkopedia enables parking reservations for some off-street parking facilities, actual payments can only be made in-car (on-street and off-street), with Parkopedia opening to payments for the first time back in 2014 when it signed a deal with Volvo.
Where Parkopedia seeks to stand out from the crowd is by being a truly global parking service provider that covers nearly every facet of the parking spectrum. Perhaps its closest competitor, from a global standpoint at least, is ParkMe, which was acquired by real-time traffic data provider INRIX back in 2015.
“Global organizations such as Apple, BMW, Garmin, and so on have a strong preference to work with a single provider who can provide them with global parking coverage, consistently high data quality, and transactional capabilities,” said Tsyrklevich.
Another area in which Parkopedia is looking to stand out from the competition is through embracing the fledgling self-driving car industry.
The future, according to Parkopedia
With its in-house army of scientists, Parkopedia is in the early stages of building a system that uses the car itself as the sensor, rather than relying on in-ground sensors, something that would make garnering real-time parking data for streets easier to scale. “The cars act as ‘eyes’ on the road, reporting information to us extracted from their built-in sensors, such as cameras, for example,” said Tsyrklevich, giving a rough outline of what the fledgling technology entails.
But what’s perhaps most interesting at Parkopedia in the near-term is that it’s forging ahead with plans to infiltrate the fast-growing autonomous vehicle movement.
A day doesn’t go by without the burgeoning self-driving car industry hitting the headlines, and flying cars are now entering the fray too. But amidst all the hullabaloo of self-guiding cars traversing busy highways and local conurbations, it’s easy to forget that the rise of autonomous vehicles also presents an interesting opportunity for companies such as Parkopedia.
Back in January, the company announced a project it had been working on with Volkswagen’s research group that involves building parking technology and services for autonomous vehicles. Part of the work included creating high-definition autonomous driving (HAD) maps of multi-storey parking facilities across Germany. The maps enable vehicles to position themselves while driving indoors “with a precision of one centimeter” without relying on GPS connectivity, according to Parkopedia, and while navigating in fully automated mode. “The self-parking car of the future will need excellent data to find the right parking spot, transaction capabilities to book and/or pay for it, and indoor navigation capabilities,” said Parkopedia’s COO, Dr. Hans Puvogel, at the time.
The road to self-driving cars becoming a reality around the world will be peppered with incremental advances — baby steps, if you will — whereby more advanced features will be introduced in stages. For example, carmakers such as Honda have been working on automated valet parking (AVP) systems for a few years already, letting drivers drop their car off and having strategically positioned cameras take over to enable the car to park itself — and return when the driver wants it.
Many cars are already equipped with features like cameras and sensors to help drivers park with precision. And with city congestion and limited parking expected to worsen, the smart parking assistance market is predicted to rise to $5.25 billion by 2021, up from $2.13 billion in 2015, according to a recent report.
Put simply, parking is playing a pivotal part in the race to bring automated vehicles to market, and Parkopedia wants to be at the forefront of the movement.
Juxtaposing the Parkopedia of 2008 with the company today reveals as much about the broader technological shift in society as it does about Parkopedia’s evolution. “If someone told me 10 years ago that Parkopedia will be running on a supercomputer in your pocket, employ a team of PhD data scientists crunching hundreds of million of data points every day in a cloud, and would be powering autonomous vehicles, I would not have believed it,” said Tsyrklevich.