On Aug. 9, 2014, in Ferguson, Mo., a white police officer shot and killed an unarmed black teenager named Michael Brown. Several witnesses described the shooting—which wasn’t captured on video—as unprovoked. In the national furor over police violence that followed, one remedy found common support across much of the political spectrum: outfitting more cops with body-mounted cameras to deter misconduct and create a record of tragic encounters. When a grand jury decided that November not to charge the officer in Ferguson, the victim’s family pushed “to ensure that every police officer working the streets in this country wears a body camera.” The White House proposed $75 million in matching funds for state and local police to buy the devices.
A few months later, in January 2015, employees of Taser International, the maker of stun guns, gathered for a sales meeting at the company’s futuristic headquarters in Scottsdale, Ariz. They filled the ground floor and lined the catwalks that crisscross the three-story atrium, a space where a lightsaber duel wouldn’t seem out of place. Shades blocked out the desert sun, and in the darkness, low, long trumpet sounds blared—the famous Richard Strauss theme used in 2001: A Space Odyssey.
As the sound of the timpani—boom boom boom boom boom boom—kicked in, Rick Smith, Taser’s chief executive officer, appeared high up among the rafters, dangling in a harness on a thin wire. He descended slowly, brandishing a giant yellow triangle: the logo for Axon, a recently expanded line of Taser body cameras and related tools. “Woooooo!” he shouted as he landed, pumping his fist in the air. The staff erupted in cheers and whistles as he lifted the logo triumphantly over his head. “Axon! Axon!” the group chanted.
Smith, 46, has intense round eyes that never seem to blink and floppy dark hair that channels Joey from early episodes of Friends. He’s often seen around the office in jeans, a Taser-branded shirt, and Taser-branded sneakers, with a Taser-branded Bluetooth headset wrapped around his neck—a get-up that echoes how ubiquitous the company’s black-and-yellow sidearms have become in law enforcement. When I ask for statistics on market share, Chief Financial Officer Dan Behrendt says he’s “comfortable saying that almost 100 percent of the CEWs”—conducted electrical weapons—“in use in North America have been supplied by us.”
When he co-founded Taser in 1993, Smith says his goal was “making the bullet obsolete.” The war on drugs and the 1992 Rodney King riots in Los Angeles had led Americans to reconsider how lethal they wanted their police departments to be. Taser’s breakthrough product, a pistol-shaped device that uses electricity to incapacitate a target, made the company worth more than $1 billion by 2004. But with market saturation came personal-injury and wrongful-death lawsuits; for almost a decade, Taser’s stock languished. The “Ferguson incident,” as the company refers to it, gave it a chance to grow again. By the next summer, with its shares at a record high, Taser executives were telling investors that thanks to body cameras, “2015 is our Super Bowl.”
Cop cams are inextricably tied to Taser, by far the dominant supplier, and the company will likely shape whatever the devices evolve into. For Taser, the cameras are more than just a new product category. Founded at one national moment of police angst, the company is using another such moment to transform from a manufacturer into a technology company. From a business perspective, body cameras are low-margin hunks of plastic designed to get police departments using the real moneymaker: Evidence.com, which provides the software and cloud services for managing all the footage the devices generate. Taser markets these tools under the Axon brand. About 4.6 petabytes of video have been uploaded to the platform, an amount comparable to Netflix’s entire streaming catalog. All of it must be preserved to an evidentiary standard. The company can sell a weapon or camera once, but cloud services are billed year after year.
“Taser wants to be the Tesla or Apple of law enforcement,” says Hadi Partovi, a venture capitalist who sits on the board. There are early signs the effort is working. In the first quarter of 2016, for the first time, Taser’s bookings for future revenue from the cameras and cloud services, $52 million, surpassed revenue from weapons sales. Every time a controversial police killing occurs—most recently with the July 6 death of Philando Castile in Falcon Heights, Minn.—there’s more pressure to outfit cops with cameras. And those cameras generate footage that will wind through the court system for years.
The officer who shot Castile wasn’t wearing a camera; Castile’s girlfriend used her phone to post live footage of the immediate aftermath to Facebook. The scene “looks terrible, and it’s hard not to conclude it is terrible,” Smith says. “We look at that and say if that [officer] had a camera on, we would at least have some idea if there was any justification for what happened.” Situations where witnesses have cameras but police don’t are “maybe the worst of both worlds,” he says. “You get all of the bad and none of the potential good, at least from the law enforcement perspective. You are missing the most critical elements to determine if it’s abusive or reasonable.” After the Castile footage went viral, demonstrators gathered for a march in Dallas, where a U.S. Army Reserve veteran ambushed and killed five law enforcement officers. “One incident somewhere can affect police officers around the world,” Smith says.
Dallas is an Axon customer: Last year the police there signed a $3.7 million contract to buy 1,000 Evidence.com subscriptions and cameras, some of which are already in use. “This will enhance our opportunity to document critical incidents; it will enhance officer safety; it will enhance courtroom testimony, which we think will turn into more convictions,” the city’s deputy police chief, Andrew Acord, told a local public radio station at the time.
Body cameras are too new for anyone to say how they’ll change policing. Early studies show that officers who wear them use force less frequently and face fewer citizen complaints—and that the footage may increase conviction rates and guilty pleas in prosecuting crimes. The results from one of the largest controlled studies in the U.S., at the Las Vegas Metropolitan Police Department, are due this fall.
Before setting out on patrol one day in March, Las Vegas police officer Sal Mascoli makes an inventory of his weapons. “Write down my badge number,” he says, in case something goes wrong. “P number five-one-one-five.” In a bracing New York accent, Mascoli points out where he hides a knife on his khaki uniform, where the rifles are stored in his SUV, and how to release his pistol. There’s a Taser stun gun strapped toward the back of his belt. Peeking above the collar of his crisp shirt is an Axon camera the size of a lipstick tube.
Mascoli is a cop’s cop, with short, close-cropped hair and a never-ending squint. He’s worked for Metro for two decades, after first serving in the U.S. Marines, and he’s trained hundreds of his colleagues in hand-to-hand combat. Out on patrol, he’s stern but affable. At a gas station, a lanky young man in hipster glasses flags Mascoli down. A driver just tried to run him over in a crosswalk, he says. Mascoli can’t do much about that, but he chats up the citizen—giving a young black man a good vibe about cops never hurts, he says later. The guy says he’s a comedian from Harlem looking for work. “I’m from Queens!” Mascoli says, as if his accent didn’t give him away. “I’m all Italian, all the time.” Mascoli hands out his phone number, recommends a comedy club with an open mic night, and rolls on.
Soon, a scraggly, middle-aged white man pushing a shopping cart catches his eye. Did he throw something over that fence? Mascoli pulls a U-turn and, almost imperceptibly, taps his chest. Two quiet beeps signal that he’s activated the Axon. He gets out of his SUV, pats down the suspect, and finds a long kitchen knife tucked in the man’s jeans. “You’re losing your knife today, you know that, right?” Mascoli says. He runs the man’s priors. “Eighty-four times you’ve been arrested, brother, and you ain’t learning,” he says. He sends the man on his way—no arrest.
Mascoli tosses the knife into his SUV, speaks into the camera to say the recording is ending, and taps the body cam. When his shift is over, he’ll dock his camera at the station to transmit the footage to Evidence.com. As he drives toward a burglary alarm, Mascoli says he doesn’t care for Taser weapons, but he’s made a bet on the business of cop cams. When the department issued him his Axon, he bought 25 shares of Taser stock. Mascoli likes wearing the camera, because it provides a balance to the citizen footage, from Facebook Live streaming to smartphone video, that increasingly monitors policing around the country. “I wanted my perception to be portrayed instead of someone else pulling out a camera phone and putting out a 10-second snippet,” he says. As his stubby fingers hunt and peck on the new dispatch computer installed in his police car, he groans, “I’m so bad at technology.” But with the camera, “I want to be on the early edge,” he says. “I know it’s coming.”
Taser initially turned to cameras as a way to ward off lawsuits, both for the company and the police. “It was about defending our business,” Smith says. As police departments across the U.S. discovered Taser in the early 2000s and bought into the idea that cops armed with a stun gun might reach for their Glock less often, the company saw explosive growth: Sales increased sixfold from 2002 to 2004. That brought scrutiny and charges that the company’s products were more dangerous than advertised. In 2004, Amnesty International published a report tallying more than 70 people who’d died after being shocked, either during the event or shortly afterward. (The group’s 2012 report says that number has risen to more than 500.) Dozens of plaintiffs have sued over injuries and deaths. Taser has prevailed in most of the cases, but in 2009 it revised its product warnings to say exposure in the chest area could cause cardiac arrest. The Securities and Exchange Commission investigated Taser’s safety claims in 2005 and eventually closed the case without an enforcement action.
In 2006 the company started selling the Taser Cam, a $400 clip-on device that automatically begins recording when an officer engages the stun gun. Street cops were skeptical. Smith likens it to the early hesitancy about dashboard cameras in squad cars—many officers dislike feeling as if they’re under constant surveillance. Few departments adopted Taser Cams. For those that did, it didn’t help Taser’s image that the device captured only the final moments of a conflict. Smith says, “You ended up with a highlight reel of people getting Tased.”
The few agencies that used Taser Cams gave the company a glimpse of an unexpected problem. Even large departments with ample resources struggled to download and preserve the video clips. As Taser began to develop a standalone body camera, one that could record long encounters or an entire patrol shift, it realized it needed to create a system to manage a new scale of footage.
The company started to build a cloud-engineering team in Carpinteria, Calif., a beach town near Santa Barbara, in 2008. Taser announced its first body camera, which cost $1,800, and an early version of Evidence.com in 2009. Smith said the platform would “help our customers not only doing their job in the streets, but defending their actions and protecting themselves against false allegations and claims against them.” His brother, co-founder and then-Chairman Tom Smith, said video evidence could cut Taser’s own legal liabilities in half.
That would be true only if clients bought the package. The physical product was clunky, with three components—a small camera worn like a Bluetooth headset, a radio device the size of a pack of cigarettes, and a minicomputer the size of a VHS tape. “That thing was a lead zeppelin,” says Taser President Luke Larson, who was the camera’s product manager at the time. “It was dead on arrival.” Only 14 small agencies adopted them. And while the website supported the basic process of uploading and managing footage, it had the dated look of an enterprise product, more IBM than Apple. “We heard from our customers that our technology absolutely sucks,” Smith says. “It looked like airline reservation terminals from the ’80s.”
Taser’s board at the time included medical, government, and business figures. In 2010, Smith recruited Partovi, a fraternity brother from Harvard, to join the board. He’d worked at Microsoft and invested early in Facebook and Zappos. (Partovi later helped recruit Bret Taylor, Facebook’s former chief technology officer, to be a director.) Partovi says he flew to Carpinteria, interviewed the Evidence.com staff, and told Smith most of the team needed to go. He suggested starting over in Seattle, which Amazon.com and Microsoft had turned into the capital of cloud engineering, and acquiring Familiar, a startup in which he’d invested. It was developing an app for families to share photos privately, but the technology and team could adapt. Taser bought Familiar in October 2013 for about $3 million in cash and stock. It beat out Twitter and Dropbox, Partovi says.
The Familiar coders shut down the photo app immediately. No longer were they devoting their time to pictures of cute babies; they suddenly had to find out what police needed from Evidence.com. “I knew nothing about law enforcement, absolutely nothing,” says Marcus Womack, Familiar’s earnest, khaki-wearing former CEO, who now runs the Axon unit. “Rick [Smith] says, ‘OK, you guys are going to spend 90 days in the field. You’ve got to do ride-alongs, you’ve got to talk to every person in the agency and command staff, and you’ve got to learn our customers’ pains.’”
Womack’s first time in a police car was in Mesa, Ariz. “We were doing 120 on the way to some robbery,” he says. “It was invigorating.” He visited 16 law enforcement agencies, from the Vermont State Police to the Los Angeles Police Department, and got a sense of the procedural headaches they faced in their jobs; he saw that Taser needed to create something simple and reliable. Womack’s team prettified the Evidence.com user interface; created an app for officers to tag footage on the go; and expanded the cloud infrastructure to support large agencies. All this cost millions, but sales of the package were unremarkable. Meanwhile, by 2013 stun-gun revenue had climbed to a high of $127 million. At Taser’s next annual staff meeting, Smith says, the No. 1 question from employees was, “When are we going to shut down this video unit?”
Then came the summer of 2014: Ferguson. After that and other killings, President Obama convened a task force on the future of policing, which recommended the use of body cameras. The task force cited a study that found officers in Rialto, Calif., who wore body cameras were half as likely to use force. In June 2015, South Carolina funded cameras in a bill named after Walter Scott, a black man whose shooting by a white police officer was captured on a cell phone. “I just thank God for it,” Scott’s mother, Judy, said at the bill signing. “It’s not just going to help the citizens, but the police officers and everyone everywhere. We thank God for this first step.” Later that summer, an activist group called Campaign Zero unveiled a slate of 10 policy reforms to reduce violence, including protecting the right of bystanders to film police interactions and supporting widespread adoption of body cameras.
Las Vegas, where a series in the Las Vegas Review-Journal had documented above-average rates of police shootings, had been considering body cameras since 2012, when it agreed with the U.S. Department of Justice to work on reforms. Bill Sousa, a professor at the University of Nevada-Las Vegas, began working on a study, still underway, with cops randomly assigned either to a group wearing cameras or to a control group without them. “We had a hell of a time getting 400 volunteers,” Sousa says. Officers were paranoid about supervisors’ ability to review footage to “monitor officer performance.” They feared their bosses would look for petty violations by cops they didn’t like. Six months passed. Then the department narrowed supervisor access to a few situations, such as when a policeman uses force, and enough officers volunteered for the study to begin. Some, like Mascoli, now embrace the camera. Others have turned their recorders back in, upset about being reprimanded when they forget to activate the device.
Dan Zehnder, a no-nonsense lieutenant, runs the department’s body camera program. Metro buys Taser’s Axon devices, but there are many smaller competitors—Zehnder says he stopped counting after identifying 35 different companies. The devices aren’t what really distinguish one vendor from another. “Someone said one time, in jest, that all of these cameras are made within 30 kilometers of each other in China,” he says. It’s the way vendors manage all the digital evidence that really sets them apart.
When the Las Vegas police department set out to buy cameras, its request for proposal listed 46 criteria. Each aligned with Taser’s existing product, from having unique audio tones that signal that a camera has been activated to providing a “secure cloud-based solution, prefer use of Evidence.com or direct equivalent.” Taser beat four other vendors for the contract. It charged $800,000 over five years for the first 200 cameras, later upgrades, and licenses to Evidence.com. The Las Vegas department has since added almost 1,200 cameras and multiyear Evidence.com accounts, costing an additional $3.2 million.
The rush to get body cameras on cops is so forceful that many law enforcement agencies are buying products without knowing what they need, according to a December survey from two professional groups, the Major Cities Chiefs Association and the Major County Sheriffs’ Association. Taser has been happy to tell them. Because of its weapons business, it has relationships with 17,000 of the 18,000 law enforcement agencies in the U.S., and it says it’s won 32 of the 39 camera contracts awarded in major city departments. Axon brought in $35.5 million in revenue in 2015. Vievu, a body camera startup acquired last year by the private Safariland Group, widely believed to be Taser’s next-largest competitor, has $10 million in annual revenue, according to confidential data provided to a job candidate in May. Safariland says the information is outdated.
Half the contracts in major cities were no-bid, Taser told investors in May, which has led to some criticism. The Wall Street Journal reported in April that Taser coaches police departments on how to avoid competitive bidding. Taser says it was “educating” customers as to why its products were unique enough to qualify for sole-source contracts. The company, which survived a decade of being accused of electrifying people to death, doesn’t appear concerned about a procurement issue. In a slideshow about company strategy shown at a May investor gathering, Taser said it would “embrace being the gorilla” (this slide was illustrated with a picture of the large primate) and “block out the noise.” (A stock photo showed a man with fingers jammed into his ears, as if singing to himself, “La la la la la.”)
Hardware—cameras and docking stations—made up less than 15 percent of Taser’s first contract with Las Vegas. “Our strategy is not to maximize the profits on the sale of the cameras but to get people into the ecosystem and on our service,” CFO Behrendt told investors last year. The remaining 85 percent of Taser’s fees were for Evidence.com subscriptions. The raw cloud storage to keep the videos is cheap. Taser’s bid document showed it would pass on about a third of subscription revenue to its cloud provider, Amazon Web Services. (It has since switched to Microsoft’s Azure.) The other two-thirds Taser would spend on custom applications, such as the website to manage access to the files, the mobile app, and tools for redacting parts of footage and creating audit trails to submit in court.
That seems as if it should create a fat margin, one that grows fatter as more customers get on board, but Taser’s financials show that the Axon unit is operating more like a tech startup, with blazing bookings but more and more losses over time. In the first quarter of 2016, the Axon’s operating losses were $10.2 million, more than twice what they were a year earlier. The weapons unit, meanwhile, produced a $15.4 million profit.
Most of Axon’s losses come from increased sales and engineering costs. To hire talent in hypercompetitive Seattle, Taser President Larson says the company will start offering Tesla Model 3 sedans to select new hires. It’s recruited staff from Amazon, Microsoft, and other tech companies. More than 80 people work in Axon’s Seattle office, a new space named the “Geekiest Office Space” by the local tech site GeekWire. It has the usual startup perks—ping-pong table and kegs—wrapped in a sci-fi pastiche, with a spaceship entry portal out of Star Wars and a retina scanner, though employees can badge in through an old-fashioned side door.
Since Ferguson, the push to put cameras on cops has focused on protecting civilians, but it’s likely that most footage will be used as evidence in prosecutions. “The mentality is, ‘We are out to catch a cop doing something wrong,’ ” says Las Vegas’s Zehnder. “My belief is that in three to five years, we’ll no longer be talking about the accountability piece—we’ll be talking about the impact of the videos in the court system, which will be the largest consumer of these videos.”
In 2015, Taser began offering free Evidence.com licenses to prosecutors, allowing them to access and share videos with police departments digitally. Taser says it started the program after some prosecutors in Florida said they feared being inundated by DVDs because so many interactions were filmed. The prosecutors’ service is run by Mike Cozart, whom Taser recruited from Amazon; he says he read The Complete Idiot’s Guide to the Criminal Justice System before taking the job. Prosecutors tell him they’ve been hit with a wave of new evidence that they struggle to manage. None of the Justice Department funding for body cameras went to prosecutors. “They, as a group, commiserate about that,” Cozart says.
Taser says it reaps a network effect from prosecutor customers because district attorneys often work on cases from multiple law enforcement agencies. The company argues that it’s helpful to get them all on one platform. Prosecutors in 20 major cities and counties are on Evidence.com, Taser says; Delaware has adopted it statewide.
Add-on products could lead to more revenue. The basic Axon camera must be activated manually, but departments can buy Axon Signal, which activates the device automatically in certain situations, such as when an officer flicks on the light bar on his car. For $10 per officer per month, another Taser service links Evidence.com files with existing dispatch and records software, so officers no longer need to individually tag files for retention or risk having an untagged file automatically deleted.
Of course, the footage must exist to be useful. Both officers involved in the shooting death of Alton Sterling, a black man in Baton Rouge, La., on July 5 were wearing body cameras, which the police department there says fell off during the altercation. Taser says a competitor supplied those devices, though the failure is a reminder that the widespread deployment of the technology is still quite new. “Many people see this as showcasing the limits of the cameras,” Smith says. “We look at that as no reason to give up.”
Body camera footage is already starting to enter the court system. Tim Fattig, the chief deputy district attorney in Clark County, which includes Las Vegas, estimates that about 20 percent to 25 percent of his cases have video evidence from the devices. Much of the footage is banal, but some can be helpful, he says. One recent trial involved a traffic stop that turned into a high-speed chase after a suspect fired a gun into the air and ambushed an officer. “It brought to life the officer’s perspective to the jury,” Fattig says. “They saw this all happened within a couple of minutes. It went from a minor traffic thing to life and death.” The defense challenged the footage, saying the suspect wasn’t identifiable in the recording and the gunshots weren’t audible, but the suspect was convicted and sentenced to as much as 157 years in prison.
Other Las Vegas cases show that body cameras aren’t the panacea that transparency advocates may be hoping for. The police department regularly releases body camera footage within three days of an officer-involved shooting, a practice praised by the American Civil Liberties Union. So far, none has caused a major uproar. But there’s one video, whose contents are almost certainly inflammatory, that the department hasn’t released. At 5 a.m. on Jan. 6, 2015, Officer Richard Scavone saw a woman named Amanda Ortiz drinking a cup of coffee in the parking lot of an area police say is popular for prostitution. According to a federal indictment, Scavone handcuffed Ortiz, then beat her, throwing her to the ground, striking her, and slamming her face into his patrol car several times.
The Justice Department charged Scavone with violating Ortiz’s civil rights and falsifying his report. He contests the charges. Although his camera captured the whole incident, Metro won’t release the footage, saying it’s part of an ongoing investigation. “We are not going to release evidence in this case,” Lieutenant Zehnder says. “Some people say, ‘We want to see it.’ You don’t get to. … It will come out in court.”
In Las Vegas alone, police have uploaded more than 16 terabytes of video to Evidence.com; that will likely pass 30 terabytes by yearend, Zehnder says. Combined with footage from other cities, it’s an unprecedented trove of interactions between police officers and the citizens they serve. What Taser does with all that data may shape the company’s—and policing’s—future.
“We’ve kind of leveraged a weapon to introduce a camera, and we’ve leveraged the camera to introduce Evidence.com,” Larson says. “This is really, really awesome, because no one else in the world is going to capture the data that we capture.” The company may explore license plate recognition, automated transcription of dialogue, and other feats of machine learning. “I mean, the possibilities are endless,” Larson says.
From where Zehnder sits, it’s obvious that big data is the next frontier for Taser and policing. “Some of it is very Orwellian and very scary and will rattle the cages of civil libertarians around the country, but it’s coming,” he says. Zehnder riffs on how facial-recognition technology might be deployed: An officer could patrol the Las Vegas Strip with a camera streaming to the cloud, “and there is real-time analysis, and then in my earpiece there is, ‘Hey, that guy you just passed 20 feet ago has an outstanding warrant.’ Wow.”
Smith says Taser plans to roll out live-streaming capabilities in 2017, and he expects facial recognition to become a reality someday so agencies can query police records or social networks in real time. “If we think about that situation in Minnesota, maybe that officer did have some preconceptions [about Castile], and maybe they were unhelpful preconceptions,” Smith says. If the officer had known that Castile didn’t have a violent police record, as has been reported, his death might have been averted. “The more we can help reduce that uncertainty, the better,” Smith says.
In Las Vegas, Professor Sousa is still crunching the data from the study on whether cop cams lead to better policing. Early results indicate some improvement, but Sousa warns against expecting too much from mere electronics. “The technology is being sold very strongly as a means to improve trust and enhance legitimacy, and I worry about tossing around those terms too loosely,” he says. While body cameras provide transparency and have come to represent the hopes of restoring faith in policing, Sousa says, “it is asking a lot of the technology to do that, especially in a really short period of time.”