Sarah Holder is a staff writer at CityLab covering local policy, housing, labor, and technology.
A startup called Firefly puts sensor-equipped advertising screens on top of Uber and Lyft vehicles. Now they do more than marketing: They collect data.
Atop most New York City taxis are advertising screens—flashing LEDs hawking Lion King tickets and beer and McDonald’s. Along with the vehicles’ yellow hue, the rooftop ads once set traditional cabs apart from the Ubers and Lyfts with whom they share city streets. Rather than waste all that monetizable airspace, though, a venture-backed startup called Firefly started installing portable advertising billboards on the roofs of ride-hailing vehicles in five major cities across the country.
Last week, Firefly announced that its mini-billboards will start offering an even more valuable commodity: data. The company’s new higher-resolution screens can sense temperature and quantify pollution and track the acceleration and braking patterns of their drivers.
Hundreds of Firefly advertising screens are out in the wild today, says Michael Hudes, Firefly’s chief revenue officer. He hopes the company will double that number in the next two quarters and swap them out with the higher-tech models, supported by $30 million from a Series A funding round led by Google, and $21.5 million in seed funding before that.
For ride-hail drivers, many of whom have taken to the streets to protest low wages and may face further cuts as Uber and Lyft struggle to turn a profit, the sell is simple: Firefly representatives will install the billboards on your roof at their garage for free, and in exchange for shilling TD Bank and running Smokey the Bear PSAs, you’ll earn an average of $300 a month, depending on how much you drive over a 40-hour minimum. It’s ideal for Uber, Lyft, and taxi drivers, because they’re typically the ones clocking so much time on the road, but technically any car owner can participate. “All we ask them to do is commit to driving and commit to driving in the markets that are important,” said Hudes.
“I look at it as side money—gas money or snack money,” said Edy Ouchi, an L.A.-based Uber and Lyft driver. He’s been using the screens for a year and a half. He says he makes up to $500 a month, and drives between 40 and 80 hours.
Firefly claims that its ads are responsible for 430 million impressions a month, with over 650,000 hours of content played over 40,000 square miles of coverage. Ten percent of unsold ad inventory is passed back to nonprofits and local government. (Amber Alert gets space, for example.)
Firefly is just one of several startups trying to harness the advertising potential introduced by an army of drivers doing laps around a city full-time. Wrapify and Carvertise will cover your entire car in advertisements; Play Octopus offers in-car games for ride-hailing passengers, interspersed with ads. Marketing abhors a vacuum, as Darren Anderson notes in his recent CityLab history of urban advertising signage: “Anywhere we spend idle time is covered: subway-car walls, bus shelters, the view above urinals, the backs of toilet doors.” Why shouldn’t our cars get in on the action?
“The street is as valuable as the desktop, as the mobile phone,” said Hudes. “We spend a good amount of time navigating the world.”
But turning rooftop car billboards into environmental data-grabbing sensors goes outside the realm of advertising. For Firefly, the move could represent a way to smooth relationships with host cities, some of which have been strained. Los Angeles District 3 city councilmember Bob Blumenfield, for example, has been working to ban the digital screens, calling them a distracting safety hazard and arguing that they’re illegal under California law. His legislation passed the transportation committee, a spokesperson said, but hasn’t yet been heard in front of city council.
Firefly could conceivably boost its standing by offering cities reams of data on traffic, weather, and other urban variables. “At this point, we’re not monetizing it,” Hudes said of its sensor info. “But at scale, if you can imagine a day in the not-too-distant future when there are 10,000 Fireflies on the road … When you look at that density in major cities, we do have the potential to collect significant amounts of data. The question is what to do with that data.” (After this story was published, a Firefly representative said the company was “not questioning what to do with that data.”)
In its pre-data-gathering iterations, Firefly screens used GPS data to roll out advertisements at specific locations. Now the company will be able to pair that with accelerometer data to identify where drivers are stopping and starting. “One of our ideas is to pass that data back to the city so they could, for example, understand traffic patterns better, how to calibrate lights, and where to possibly put a stop sign,” said Hudes.
Later iterations could assist other government entities, such as law enforcement. Hudes says Firefly is sensitive to privacy concerns, and for now is not installing cameras in its screens, for example. But he said they’ve been “contemplating” and “testing lightly” a feature where the screens pick up noise off the street with built-in microphones. (Noise sensors are not part of last week’s roll-out. A Firefly representative stressed to CityLab after publication that Hudes has nothing to do with the product team where these decisions would be made, and that all his mentions of such a feature are hypothetical and are not part of the company’s roadmap.)
“This is a conversation we’ve been having with the San Francisco police department,” said Hudes. “In the event that there’s, regrettably, a gunshot, they can come to us and ask if there was a Firefly in that neighborhood.” Acoustic data cross-referenced with GPS coordinates might help police identify where the shot came from, and from what direction. The San Francisco police department says that “never had such a discussion with a Firefly representative about this hypothetical use of this technology.” They did meet with a representative “to discuss the possibility of featuring public safety messages on the display screens,” but have no formal agreement, and have “no plans to move forward with an agreement.”
Such an application bears a resemblance to an existing system called ShotSpotter, which uses acoustic sensors to detect gunshots and estimate their locations. But that system, which the company says is used in 100 cities, has been criticized by the ACLU for privacy concerns, and there’s little research to support the idea that it helps cut down on violence.
Hudes says any theoretical Firefly-mounted microphones would not pose a surveillance risk: “It’s not something where [we’d be] listening in on conversations,” he said. “That wouldn’t be the right application.”
But as with many data-collecting systems, the “right” application may be in the eye of the beholder, or in this case, the entity that gains access to the data. Equipping ride-hailing vehicles with data-gathering powers might be an example of starting with a solution—more sensors, everywhere—and searching for a problem for it to solve, says Rachel Thomas, the founding director of the Center for Applied Data Ethics at the University of San Francisco’s Data Institute, and the co-founder of the online coding school Fast.ai. “Companies aren’t necessarily taking into account what the actual needs and requirements of cities are when they push these products,” she said. “If you start with the problem of what’s the best way to solve traffic congestion, you’re going to get to another answer” than strapping a screen to an Uber.
Flooding streets with data-collecting sensors without first asking permission—or knowing if there’s a market for the data—is also right out of the “collect first, monetize later” playbook to which many companies subscribe, says Jathan Sadowski, a researcher at the University of Sydney’s School of Architecture, Design, and Planning. But data doesn’t have to be monetized to be valuable: “For the most part, the vast majority of data that is collected about us—by data brokers or by whoever—if it’s not turned into advertising dollars, it’s not turned into money for the most part,” said Sadowski. “It’s used to power other kinds of systems. … A lot of the value of data is the power it provides to people who hold it.”
Air quality sensors could be useful for community health officials, accelerometer data could be useful for street planners, and, down the line, microphone data could be useful for law enforcement. But those three entities are not created equal in the governance of a city, Sadowski says. “It’s not hard to image these kinds of slippery slope add-ons where more stuff gets added onto the vehicle-as-platform,” he said. “And then whose interests end up winning out in the end? At this juncture in society, it’s always the police.”
Beyond the privacy implications, there may be more kinks to work out before data-scrounging Firefly screens are widely adopted. For one thing, Uber drivers are technically not allowed to mount them, because the company prohibits “commercial branding” in vehicles used by their drivers. Ouchi said that when he went to get his car inspected by Uber, they told him he had to take his screen off or he’d be deactivated. (Uber did not respond to a request for comment, and Los Angeles Magazine notes that the company rarely enforces the branding bans.)
Lyft’s policy is a bit more liberal: A spokesperson for the company said that they leave third-party advertisements and signs to be regulated by local laws, and that “[d]rivers should refer to their city and state’s regulations when giving rides, as well as keeping in mind the impact they could have on riders.” Because advertising is prohibited at San Francisco airports, the San Francisco Chronicle reported that Firefly-equipped drivers currently shut off their screens in the vicinity of SFO.
Some riders, Ouchi says, don’t appreciate car-mounted advertisements. Once, a passenger canceled on him, accusing him of “double-dipping” after seeing the flashing screen. “He was like, ‘I don’t like cheaters,’” he said. Another driver shared a similar story on Reddit, claiming the display has cost him five-star ratings.
It’s possible that the naked commercialism of vehicle-mounted advertising feels like something of a rebuke to ride-hailing’s old “sharing economy” image. But with Uber and Lyft still fighting California’s attempt at reclassifying its drivers from independent contractors to employees, that image correction is probably overdue. At the end of the day, gig drivers look more and more like drivers. And now their cars may look more like… cabs.
CORRECTION: A previous version of this article included an incorrect number of Firefly billboards on the road, shared by the company. It has been updated.