By the time Dave Tulloch signed up for the paycheck advance app B9 late last year, he was far behind on his bills. A recruiter who financially supports his family, Tulloch had been laid off twice during the pandemic and was in need of fast money as soon as he found another job. Like the many cash advance apps that target workers living paycheck to paycheck, B9 required no credit check and no interest payments: All Tulloch needed to do was create a B9 bank account for $9.99 a month and log in to his company’s payroll provider, Paychex, to provide real-time access to his information.
The app initially offered Tulloch an advance of $75, before quickly increasing that to $300. But when Tulloch’s company took a Monday holiday for New Year’s, his paycheck arrived late, which delayed his repayments. Tulloch says that B9 promptly dropped his cash advance limit from $300 back to $75. He was stunned—he needed the money to pay bills.
“It almost seems too good to be true when they get you the money,” he says. “Then they cut you down at the knees.”
When he complained to customer service, they told him that the limits are based on a variety of factors. These include earning and spending patterns, which generate a score and can change at any time.
B9 said it could not legally discuss specific customers but that it “takes every complaint it receives seriously” and makes “every reasonable effort to resolve customer complaints both efficiently and amicably.”
Tulloch is one of a growing number of US workers turning their personal data over to private companies in exchange for paycheck advances, fueling an industry potentially worth up to $12 billion, by some estimates. In 2020, $9.5 billion in wages were accessed early, according to the research firm Aite-Novarica Group, up from $6.3 billion in 2019. These early payouts can be habit-forming; a 2021 report from the Financial Health Network found that more than 70 percent of pay advance users took out consecutive advances.
What Tulloch didn’t know was that when he signed up for the app, a company called Argyle was retrieving the data that would be used to decide how much money to give him. It builds the technology that allows companies like B9 to extract a wealth of data from payroll accounts—up to 140 data points. These can include shifts worked, time off, earnings and promotions history, health care and retirement contributions, even reputational markers like on-time rate or a gig worker’s star rating and deactivation history. For every worker that uses its product, Argyle charges customers like B9 a fee, plus an additional monthly charge for continuous monitoring. This makes for a valuable data trove; it’s further upstream than banking data, providing a fuller picture of a worker’s earnings, deductions, and behavior. Some estimate that payroll data could be worth $10 billion. Argyle pegs it at 10 times higher.
Argyle is part of an emerging set of payroll data companies founded over the last four years to cash in on workers’ personal information. They build secure connections between payroll providers like Paychex and businesses that want to access that data, like B9. Argyle acts like a courier, shuttling data from one account to another, the same way banking data is transmitted to apps like Venmo. Its competitors include Atomic, Pinwheel, Truv, and Plaid (which builds those bank integrations but recently began releasing payroll products). The data that workers provide can be used to underwrite financial products like loans, mortgages, insurance policies, and buy-now-pay-later apps; simplify direct deposit switching; or verify income and employment for apartment and job applications.
These companies are quick to stress that they store workers’ data for little or no time and that all the information shared with third parties is provided voluntarily. Direct access to payroll accounts can lower risk to lenders, who can pay that forward in the form of lower interest rates. It provides alternatives for immigrants without credit scores and gig workers who struggle to qualify for credit due to variable earnings and multiple income streams.
A 2019 Harvard study of one earned wage access product found that the fee for a $200 salary advance was the cost equivalent of 16.7 percent of a traditional payday loan and 14.3 percent of a typical $35 overdraft fee. “In general, [payroll data sharing] has been good for credit inclusion for millions of people who were otherwise blocked from access to credit under the traditional scoring models,” says Rob Levy, vice president of research at the Financial Health Network, a nonprofit that invests in companies developing products for low-income customers.
Some worker advocates, however, have raised concerns about the granularity of the data workers are forking over, perhaps without even realizing it. During a 2019 interview, Argyle’s founders said that payroll data carries “not only financial insights” but “a holistic view of a worker’s identity, including typical hours, work trajectory, reputation, and more.” In response, the worker organizing platform Coworker raised concerns that the technology could be used as “a tool for the future extreme vetting of workers, especially low-wage, immigrant, and other BIPOC workers.” Nicole Moore, president of Rideshare Drivers United, says companies can “market [access to] that data out to bad lenders who offer us soul-crushing loans.”
If a worker’s fortunes go south, their creditors, plugged into their payroll software, will be the first to know. An Argyle blog post aimed at its customers says, “Our webhooks [automated messages] notify you instantly and automatically of any changes to your user's income and employment data on an ongoing basis. So if a customer is laid off, gets a raise, or starts a new job, you'll know about it right away—empowering you to make better underwriting, lending, and sales-related decisions.”
Ready cash has always been a problem for low-wage workers—so much so that Uber began offering “Instant Pay” in 2016 to gig workers whose variable incomes often left them needing money in a pinch. Cash advance apps have become more popular during the pandemic, and since they claim not to be offering loans—and are therefore not regulated by lending laws—they cannot charge interest. Many apps partner with workers’ employers, some of whom cover the fees. Others are direct-to-consumer, relying on membership fees or suggested tips, which they often spin as helping others in the community. The National Consumer Law Center has argued that these fees and tips are a disguised form of interest and that these products should be regulated as credit. Last month, the Consumer Financial Protection Bureau signaled that it would reconsider whether certain advance pay products should be regulated. In December, it launched an inquiry into buy-now-pay-later practices.
Marisabel Torres, director of California policy at the Center for Responsible Lending, worries that this type of data collection could make people vulnerable to being targeted for other credit products they cannot afford. “You enter into this contract with financing to buy a shirt online,” she says, referring to buy-now-pay-later products, which, like paycheck advances, have skirted lending laws. “Now you have a relationship with this lender, who is collecting data about you. There’s a concern that it could be used to market predatory products or other types of credit.”
Walmart associate Faith Brown uses an employer-sponsored pay advance app named Even when she’s short on cash. When a dental bill came due last month, she used it to get a $400 advance. Walmart covers the membership fee. “Being strapped for cash can be stressful, so it's nice that the apps aren't just another big hassle,” she says. When she exceeds her limit, sometimes she turns to Earnin or Dave, which both offer small cash advances in exchange for tips—Dave’s default suggestion is 10 percent. She occasionally borrows herself into a pay advance cycle, with one early cashout fueling the need for the next. “When you really need the advance, sometimes it just happens,” she says. “The best way to get out of the cycle is to just borrow less and less each pay period until you can go without it.”
Payroll data firms can access other companies’ employee data without their permission. Once they are granted access by a worker or a cooperating payroll provider, they can build integrations directly with payroll systems using application programming interfaces (APIs). This allows them to programmatically extract data that’s useful to their customers. If a payroll provider doesn’t have an API (although most do), some companies revert to screen scraping, an automated data capture method that entails logging on as the account holder and using a computer script to copy their information. Some payroll data companies are loath to discuss screen scraping, perhaps because it can be less secure. (Argyle and Truv denied screen scraping; Atomic and Plaid acknowledged doing it as a last resort; and Pinwheel repeatedly sidestepped questions about the practice.) Alex Johnson, director of fintech research at Cornerstone Advisors, notes that even though payroll data companies don’t need employer permission, it doesn’t necessarily mean sharing login credentials wouldn’t violate a worker’s employment contract.
However they get the data, companies are in a race to reach as many workers as possible. Full coverage would mean access to 250,000 payroll systems in the US alone. The more coverage they can offer, the more enticing their product will be to lenders and other customers. This in turn benefits the payroll data companies, which earn a fee for each employee who links their account. While some say they cover the majority of US workers, none have approached full coverage yet.
Argyle has come under scrutiny for its alleged data collection methods. According to Vice, a group linked to Argyle called Workplace Unite was offering up to $500 for access to payroll credentials. A subsequent Krebs story found that a company behind this and other websites used to collect this data, Worker Research Alliances, appeared to be linked to Argyle’s API. As of this week, a CraigsList ad linked to WRA was still offering $250 to workers who use Namely, a payroll provider. Argyle denied paying for payroll credentials and shared an open letter it posted at the time. When WIRED asked if Argyle was affiliated with WRA, the company stopped responding. WRA did not respond to WIRED’s requests for comment.
Argyle CEO Shmulik Fishman says the company can coach lenders on factors like consistency of work and upward trajectory. “Is your job title changing in an upward direction every six months? These are signs of a good worker and one where you might want to take another look,” he says.
Reputation markers, however, can reflect bias. Shannon Liss-Riordan, an attorney who is suing Uber for its allegedly racially biased customer star ratings system, recently surveyed the drivers she represents. Of more than 4,000 respondents, 17.4 percent of white drivers said they’d been deactivated due to a low rating, versus 24.6 percent of Asian drivers, 24.1 percent of Black drivers, and 24.9 percent of those who marked their race as “Other.” Only 16.9 percent of Latinx drivers answered affirmatively, but the real number is likely higher because several drivers self-identified as races such as Hispanic under “Other.” “It's shocking to me that customer service data would be used for other purposes that could affect drivers' livelihoods, including access to loans or other benefits,” says Liss-Riordan. “That's a very dangerous precedent.”
Asked about the risk of perpetuating bias, Fishman says, “We are not in the discrimination business. And we're also, very importantly, not in the business of creating criteria for approval or rejection choices.”
To be sure, not every payroll data firm is as focused on reputational data. “We don't do that,” says Truv CEO Kirill Klokov. “I just don't find that helpful when you apply for a loan to know your star rating on Uber. The primary use case is you should be able to prove that, in the absence of a FICO score [for an immigrant] like me, I'm actually a person who will pay you back the loan. Or I actually worked at a company that I'm claiming I worked at.”
While consumers must consent to share their data, if they change their minds later, they may lose access to a product and have handed over their data anyway. And some workers in a financial pinch may feel they have little choice. Michael Gray, an Iowa pest control specialist, regularly uses a cash advance app called Earnin for advances up to $550. He agreed to have his GPS location monitored by Earnin to confirm he went to work. (Earnin doesn’t use payroll data.) Although he found it invasive, he complied. “They’ve kind of got you by the balls when they're dealing with your money and you're trying to scrape by.”
Despite borrowers’ uneasy relationship with pay advance products, the convenience can be hard to resist. “If I need $100 for a bill or my groceries or whatever, it's right there,” says Gray. “It's quick. It's a few clicks. So it's been pretty effective at keeping me in their ecosystem.” He adds, “I really want to be out.”
What consumer and worker advocates all seem to agree on is that the proliferation of these financial products is the symptom of a deeper problem: insufficient pay. Employer-sponsored earned wage access “basically allows you to get away with paying your workers as little as possible because you can prop up bad employment practices,” says David Seligman, executive director of Towards Justice, a nonprofit law firm that represents workers.
“The thing we most need is higher wages, better tax programs, more support for low-income families, and a child tax credit,” says Levy. “But short of that, the reality is that we have many people living paycheck to paycheck. They're going to occasionally need credit to make ends meet.”
Updated 3-23-22, 6:45pm EDT: An earlier version of this story said that buy-now-pay-later and paycheck advance products were not governed by lending laws. Regulators are examining whether or not they are subject to these laws.
- 📩 The latest on tech, science, and more: Get our newsletters!
- The infinite reach of Facebook's man in Washington
- Of course we're living in a simulation
- A big bet to kill the password for good
- How to block spam calls and text messages
- The end of infinite data storage can set you free
- 👁️ Explore AI like never before with our new database
- ✨ Optimize your home life with our Gear team’s best picks, from robot vacuums to affordable mattresses to smart speakers
