Cities without landmarks
Niagara Falls, U.S./Canada
Matt Crampton is pushing back against my charge that his company, Gigwalk, is contributing to class stratification in the United States. And to be perfectly fair, I’ve got to admit that Crampton doesn’t look much like the embodiment of Silicon Valley income inequality. Sitting across a table from me at a Berkeley cafe, sporting a long bushy red beard, the co-founder and CTO of Gigwalk looks more like a Brooklyn hipster on his way to Burning Man than a paradigm-busting pioneer on the new frontier of work.
“We are creating kinds of work that have never existed before,” says Crampton. “The average person who downloads our app is finding work that they just wouldn’t have had access to previously.”
He makes a provocative case. Gigwalk mobilizes temp labor in a way that is intriguingly different from the other occupants of the virtual employment market niche, companies like Mechanical Turk and Task Rabbit and oDesk and Fancy Hands. Working mostly with big corporate clients, Gigwalk believes it has figured out how to exploit the distributed coordination capabilities of smartphones and the Internet to facilitate work that would have been logistically or financially infeasible just a few years ago.
For example: Suppose PepsiCo wants to make sure that its promotional displays are properly set up at grocery stores around the country. And the company wants to do it now, before a big marketing campaign launches during a major sporting event. Hiring enough full-time staff to check out every store before the weekend arrives would be a huge and expensive endeavor. Gigwalk’s pitch is that it can instantly mobilize thousands of part-time workers willing to zip down to the local supermarket and snap a photo — or better yet, fix whatever might be wrong with the display.
That’s interesting. We’ve become used to, and increasingly alarmed by, how Silicon Valley innovation tends to slice and dice work in ways that aren’t necessarily good for workers. Technology is automating or helping to outsource and offshore jobs that used to be the bread and butter of the middle class. Innovation is generating fantastic wealth, but that prosperity is concentrating in ever fewer hands. Unions have been replaced by a vast and growing pool of contingent workers who command zero leverage over capital, are ineligible for employer benefits, and seemingly condemned to an ever more desperate scramble for whatever scraps of work are available.
But there are also examples of the Internet creating new platforms for commerce or creativity that didn’t previously exist, and that add to, rather than subtract from, our opportunities for economic empowerment. Etsy and eBay both spring to mind. Crampton is telling me that Gigwalk’s model makes possible an entire spectrum of jobs that previously were too much of a pain in the ass, or too logistically complex, to pull off. If true, that’s an encouraging development, a sign that clever Silicon Valley start-ups are figuring out ways to take positive advantage of an underemployed labor force.
“Gigwalk could be the eBay of work,” suggests MIT labor economist David Autor, half-jokingly. “Maybe they’ve figured out a way to find those jobs that were gathering dust in your grandmother’s attic.”
* * *
Maybe. But the question of whether Gigwalk is enabling new work isn’t the only metric by which to evaluate the company. Crampton also makes a big deal of the company’s secret sauce, a worker evaluation algorithm it calls “Moneyball” in a nod to Oakland A’s general manager Billy Bean’s famous preference for building teams around advanced statistics measuring player performance. Gigwalk collects data about every aspect of a “Gigwalker’s” work performance. The data points include how fast a Gigwalker responds to a smartphone app alert about a new job, how quickly and efficiently the job is completed, customer satisfaction ratings, and even personality characteristics. Is the worker “gregarious” — i.e., someone suitable for conducting surveys of live people — or is the worker the kind of person who hates talking to anyone, and is therefore better suited for straightforward data collection, taking pictures of menus, for example?
“Behind the scenes we are watching everybody,” says Crampton, “while they are going about doing their work. We are building these mathematical profiles on top of people, figuring out who is doing good jobs on a variety of gigs. We can figure out what kind of jobs you do well and start routing more complex, higher-paying jobs to you based on the skills we see inside our system. And then we can provide companies with workers with the specific kind of skill sets they need to get work done. ”
“If you are taking more than 30 minutes to respond back to a message from an employer,” says Crampton, “the likelihood of you being successful on gigs on Gigwalker drop off significantly.”
The cutting geek edge of baseball statistic analysis is known as sabermetrics. What Crampton is describing might be called “labormetrics.” Your data profile is a résumé you can’t escape.
Think of the way Netflix gathers data on your viewing preferences to create a model of what you might like to watch in the future. Gigwalk gathers data on what kind of worker you are, and rewards you accordingly. “Good” workers become eligible for more complex, higher-paying jobs. “Bad” workers get stuck at the low end, where they can do the least damage in what Crampton calls the “low cost of failure jobs.”
Crampton boasts: “We’re ending up knowing more about workers than most employers do because we are capturing so many statistics around all these different data points.”
You couldn’t ask for a better practical distillation of the ascendant ideology of Silicon Valley meritocracy. Under the reign of labormetrics, the productive shall be rewarded and the weak shall flounder. It all adds up to a spooky video-game exercise. Gigwalk’s workers are playing the labor market equivalent of “Ender’s Game.” Hit the right buttons and you move up a level. Hit the wrong buttons and you never get off the bottom rung. So don’t dawdle answering that email or uploading that photo! Your performance is being measured, all the time.
Aside from the sheer creepiness of your smartphone morphing into a panopticonic prison guard, one can well ask: Are we sure that this kind of meritocracy is healthy for an economy in the long run?
Job polarization over the last few decades is a real phenomenon, something that David Autor has been documenting for years in his research. A labormetric meritocracy might help define those polarities under a stark new glare. If you reward the good and punish the bad, based purely on the numbers, what happens to the middle? Is there even a middle?
* * *
Gigwalk was launched in July 2011 by three Yahoo veterans who worked in the domain of local search, but were becoming increasingly dissatisfied with the quality of the data they had access to.
“What we realized was just how fast local search data can get out of date,” says Crampton. “Ten percent of restaurants go out of business every month. Yahoo would get data from these companies in Vietnam that were transcribing information from phone books and shipping it to us on CD-ROM, and that’s how our local search results would get populated.
“So we started wondering whether, since there were so many people around the world with smartphones, we could create a platform that would incentivize people to go out and collect that data for us.”
Without the smartphone, there would be no Gigwalk.
“The smartphone is what enables us to operate at the scale that we do,” says Crampton. “There is simply no other way to do it. ”
One of Gigwalk’s big early customers was Microsoft, which wanted to improve restaurant listings. But the logistics involved were intimidating. “Microsoft wanted to go in and survey 65 cities worth of restaurants,” says Crampton. “There was no way that they could have done that without a system like Gigwalk.”
Companies like Procter & Gamble and other retail giants, says Crampton, want to know what is happening in the aisles of CVS and Safeway in real time. If sales drop in a certain store or region, they’d like to know if there is a practical explanation. Are products being shelved correctly? Again, prior to the rise of the smartphone generation, there just wasn’t any easy way to mobilize thousands of bodies to walk into those stores.
What we are seeing right now, suggests Crampton, is only the beginning. Now that we can mobilize large numbers of geographically distributed workers in real time, retailers will figure out other ways to exploit that potential. And he makes no apologies for the “Moneyball” number crunching. To pull off more complicated jobs, Gigwalk needs to be able to distinguish the capable workers from the incompetent. It needs to be able to mobilize exactly the kind of workers an employer needs to meet specific challenges.
Gigwalk won’t release detailed numbers, but claims revenue has grown 300 percent in its most recent quarter compared to the corresponding quarter the previous year. Crampton says Gigwalk has more work than it can handle right now, even with “340,000 workers inside our platform.” The company needs more workers, says Crampton. In fact, Gigwalk’s invitation to talk with Crampton is part of a press push timed to publicize the debut of its new Android app on Aug. 15. Crampton is hoping that moving outside of the iPhone ecosystem will double the available pool of workers.
That would be great for Gigwalk and its investors. But that doesn’t mean necessarily that Gigwalk is a great way for workers to get rich. Crampton says employees “average $10-15 an hour” and that the most productive workers are earning $1,200 to $1,500 a month. But that’s before taxes. And as is typical of just about everybody in the “contingent labor” workforce, Gigwalkers are independent contractors who must cover their own health insurance.
It’s part-time work, acknowledges Crampton. But it often comes on top of other part-time work. He says that 60 percent of Gigwalkers identify themselves as already having another job in addition to Gigwalk gigs. And “it’s part-time work that has been made accessible to hundreds of thousands of people where it might only have been accessible to a select few before.”
But depending on the region you live in, Gigwalk work can’t be assumed to be steady work. Gigwalk put me in touch with Nikki Goodson, a Gigwalker who lives in Kansas City. A stay-at-home mom, Nikki told me that in September 2011, when she first signed on, there were very few gigs available.
Even now, she said, “it goes in waves.” When Microsoft Bing was looking for restaurant help, a ton of local gigs showed up on her Gigwalk app. At other junctures, gigs in Kansas City were so scarce that Goodson drove all the way to Omaha, Neb., for a gig. After accounting for gas, she estimates she cleared $50 that day. “Maybe it wasn’t the most intelligent business decision,” she conceded.
I asked Autor what could be said about the macroeconomic effects of the rise of hyper-temp outfits like Gigwalk.
Not much, he said. In the terms of the overall economy, the Turks and Task Rabbits and Gigwalks still account for only a tiny slice of labor activity. Very few people, he speculated, were actually making a decent living from such services.
But Autor wasn’t buying my thesis that Gigwalk might be yet another Silicon Valley contributor to increasing class stratification. His own research has made clear that technological innovation has contributed to job polarization and squeezed the middle class, but Gigwalk, he noted, isn’t outsourcing jobs to the Philippines or automating them out of existence. Instead, he said, it is providing at least a modicum of new opportunity. In effect, Autor said, Gigwalk was flipping the outsourcing model on its head. Instead of exploiting Chinese or Indian labor, Gigwalk had discovered a way to mine the huge pool of labor in the United States currently twiddling its thumbs.
The critical question, says Autor, is whether a tech company “is displacing work or creating work.” Other Internet economy temp labor clearinghouses — such as oDesk or Mechanical Turk — he noted, have pushed basic data entry jobs out of the developed world. There’s no way that even the most cash-strapped high school senior in the U.S. can compete with the wages workers in the Philippines or Central Europe are willing to accept. But taking pictures of menu listings can’t be done by Bangalore call centers.
“Things that create demand for activities that can’t be really mechanized or offshored are generally a positive thing for the economy,” said Autor.
Moreover, according to Crampton, my thesis that the new efficiencies of work distribution pioneered by Internet economy start-ups — all hail labormetrics! — woul inevitably force us all to scramble harder for less “assumes that there is a finite amount of work available,” says Crampton.
“But one of the things that we get excited about at Gigwalk is that we are enabling working that never existed before. We are actually increasing the total amount of work that is available to everybody to go and do. The total number of man-hours of work that is available in the U.S. economy is going up because Gigwalk is available.”
“We’re making a huge difference in people’s lives,” asserted Crampton. “You might be living in Lousiana, and you download our iPhone app, and all of a sudden, there’s jobs there that you can go do.”
It’s an attractive vision, once one wades through the hype. And there is certainly a sense in which a “mobile-centric workforce” clearinghouse like Gigwalk fits neatly the evolving customs of a generation that filters everything through its smartphones.
But there is an obvious price to pay: the labormetric price. Do we have any conception of what the long-term psychological impact of knowing that every aspect of your work performance is being captured and graded and crunched by algorithms will be? Sci-fi dystopias have built on a lot less.
The ultimate extension of a labormetric smartphone future is that how we use our smartphones will define who we are, what kinds of jobs we get, and how well we get paid. Who knows what will end up in that mix, or what correlations will emerge? There’s no slack in a labormetric future. You can’t hide the hole in your strike zone. Everything goes on your permanent record.
So you better start hustling, now.
Niagara Falls, U.S./Canada
Sydney Opera House, Sydney, Australia
Mount Rushmore, South Dakota, U.S.
Eiffel Tower, Paris, France
Colosseum, Rome, Italy
Taj Mahal, Agra, India
Siena Cathedral, Siena, Italy
Christ the Redeemer, Rio de Janeiro, Brazil
Arc de Triomphe, Paris, France
Lost City of Petra, Jordan