The following was produced by Tearsheet Studios. We worked with employment data platform Argyle to create a podcast series about the rising importance of employment data and how lenders, banks and fintechs are using this data to make financial products available to more people, solving some of the challenges with today’s financial services.
The nature of work is changing. People are taking on more gigs, driving for Uber, freelancing on projects. I talked with Argyle CEO and founder Shmulik Fishman about the opportunity to build an employment data ecosystem, connecting employers, payroll providers, and consumers. We discuss his entrepreneurial journey, the technical challenges ahead, and the role of consumer-permissioned data in the financial industry.
Fishman has been building Argyle for the last four years, designing a gateway to employment records. “Employment records sound like a fancy term,” he said. “So it’s helpful to break that down. We all have an employment record with our employer. It’s the information about how long you worked there, and what your base salary is, how many shifts you take, who your manager is, what your job title is — these are all the elements that make up your employment record. And the longer that you work at your employer, the larger your employment record is, and the more valuable it is.”
This move to gig work presents a challenge for financial services providers that use a traditional paycheck as part of their underwriting process. And into this vacuum, new data and technology providers are helping financial services firms access employment records where today’s workers generate income.
Shmulik Fishman: It’s the way that you will be able to obtain a loan or credit card or rent an apartment. These businesses are going to evaluate your employment to extend you credit, or to allow you to lease a vehicle or rent that apartment that you want. Employment records are the foundation in a lot of ways of Western society. And we’re helping to make sure that consumers have easy, secure, and consistent access to them.
Argyle’s Shmulik Fishman knows a thing or two about using technology to solve big problems. At different points in his career, he’s tackled ad technology and fleet management. At the heart of what he’s designed and built is the network.
Shmulik Fishman: I’ve been infected by technology for the last 12 years. I find the ability for computers to make things easier for people, just to unlock a lot of possibilities. And I love being part of that. I feel like we’re making the world a better place when we build good software. And I’ve had the privilege of working at three different tech firms. All seemingly quite different, but have a string throughout them about networks.
The first company I worked at was called Adap.tv. It was a network that allowed advertisers to place ads on millions of websites on the internet. It was at the beginning of the ad tech boom. And that company was sold to AOL. The second tech company I co founded. It was a fleet management business called Stratim. It was a very different type of network, one that is physical, a ton of dealerships, and a ton of mechanics in the United States to network. We were helping to make sure that there was a simple way to understand what vehicles needed to be at what mechanic shop or be at what airport at what time.
In a way, Argyle is its own network, bringing together employers, their workers and financial firms that want to access employment data.
Shmulik Fishman: It is a network of the 8 million plus employers or people who paid somebody. The IRS says that about 8 million business paid somebody else last year. And then all the consumers or all the employees in the United States that need to use their employment records in order to obtain products and services. And so my background is very much of taking complex networks with a lot of nodes in them and making them very simple for the different participants in the network to leverage and use.
The issue of dealing with employment data is real: for the individual applying for financial services and for the firms doing credit checks. With 8 million businesses, getting at the data is a challenge. Even handling the forms presents a big hurdle. Fishman encountered these challenges in his previous work and sensed a big opportunity.
Shmulik Fishman: At Stratim, we had a very novel problem that everybody experiences, whether you’re a business or an end consumer. We had to process applications for employment. And this form that was on the internet required you to fill out your first name and your last name and your driving history, your social security number, and upload your driver’s license. I think we’ve all experienced either managing forms like this, or having to fill out forms like this. If you rent an apartment, you filled out a form that is like this.
It’s not just users that struggle with these types of forms — it’s the businesses that use them, too. The manual data included in these forms meant a human needed to review it, creating workflow issues. And with low conversion rates on the user side, it created a scenario that Fishman calls “everyone’s losing”.
Shmulik Fishman: This form was very irritating for me. I understood how complicated and laborious filling out this form is to an individual that wants to drive cars. And it was also a form that was quite laborious to ingest. It contained manual data, meaning that it had to be reviewed. The conversion rates on the form are very low. We had to have rather sizeable admin teams to validate the information on this form, and so it was bad for all parties involved, or as we said, internally, ‘everyone’s losing right now’.
The initial version was limited to addressing the scaling problems with these forms. He admits that the vision for Argyle started narrower and expanded from there.
Shmulik Fishman: And I thought there was a better way. There should be a way to automate this form. We were thinking at the time rather small. How do you automate applications if you’re filling out a job application on Monster.com or Indeed? How do you automate those forms that we want to automate for ourselves?. But there were a lot of these forms that persisted. And what we discovered over time is that this form has infected every element of society, right? Renting an apartment, filling out a form for credit, getting a mortgage — it really is about form fill outs and about the data that comes from these forms. And so what started as a rather simple problem has become a business unto itself with the infrastructure of automating forms or getting rid of these forms.
In an era of APIs and on-demand everything, the employment authentication and verification process is surprisingly antiquated. It’s still full of manual processes and old school thinking.
Shmulik Fishman: And what we’ve started to see, as more and more clients have come and started to use Argyle, is that there are a limited set of options on how to evaluate an individual. Here’s a set that we think of commonly today: you can use a paper; you can ask people to upload their pay stubs; upload their W2s; you can use an external verification provider. That provider is going to call, email or maybe fax the current employer. Those are the options that are available today to evaluate somebody before you issue them the product or service that you’re trying to issue them.
Still, there’s a bias towards extending credit to the most scoreable applicants. If you work in big tech, you’re going to get approved for a loan. That bigger addressable market has changed the way Fishman views his company’s mission. He’s taking aim at the credit score.
Shmulik Fishman: I think that those options don’t give a fair shot to most Americans. Surprise, surprise — if you work at Google, you’re going to get a mortgage for a house. Surprise, surprise, if you work at Lockheed Martin, you’re going to be able to buy a car. But for the average American, for somebody that is making $50,000, $60,000, $70,000 a year, maybe less, it is hard for them to get access to products and services. And it’s because credit decisioning is reliant on a credit score. And these credit scores are blackbox numbers. Why I have my number and why somebody else has their number is mysterious to all of us. I think that we want to live in a world — Argyle is working towards a world — where businesses are far less reliant on a single number on a credit score, are far less reliant on paper, are far less reliant on phone calls, and more reliant on looking at an individual’s employment record, the data set inside of it, and being able to make their own determination about whether to issue a consumer a vehicle or rent them an apartment, and not have to rely on these single box systems.
So if we’re able to move beyond the box: what does the world look like 10 years from now, when we’re not reliant on credit scores anymore?
Better access to employment data should give a fair shot to more Americans. By moving beyond the credit score, people will be judged more on their real time income, extending and accentuating the static, rearview mirror look into someone’s past payment history.
Shmulik Fishman: I think we’d live in a world that is far more inclusive, and one where far more individuals have access to products and services. So the ecosystem gets much larger. It’s also a world where risk is reduced quite substantially. Because if a business has transparency into somebody’s work, and is able to have a personal relationship with that consumer and tailor products and services, payback, payback schedules, provide flexibility and those schedules — that’s a more flexible world, a more cohesive world, a world where I think we all want to live in.
Fintech has focused for years at getting access to the bank account, to analyze transaction history. With employment data, though, things are changing. We’re moving from a vision of open banking to open finance — where more data sets are layered in to financial decisioning.
Shmulik Fishman: Fintech is a word that the industry uses a lot to describe many things. I’m not into labels. But people do seem to think that Argyle is part of the fintech space. That’s fine. I think that we are an infrastructure company that provides access to employment records. Put whatever label you’d like onto that that is most helpful. Our industry has spent a lot of time making it easy to look at somebody’s bank information or their finance information. How much money do you have at a specific institution? How do you spend your money? How do you save your money? These are important questions. But as we’ve really focused on that, the industry as a whole has focused a lot less on how people make their money. And that’s the trend that I’d like to see and that Argyle is pushing forward with — how somebody makes their money is critically important to evaluating that individual’s ability to pay for products and services. And I’d like to see the industry make more investment on that side of the ledger.
There’s a lot of focus on financial institutions. That means plugging into banks or investment accounts. The data infrastructure players have primarily focused on assets and spending — that’s the main departure for Argyle and where it wants to innovate.
Shmulik Fishman: We provide infrastructure that plugs into places where people work, whether that is a gig platform, like Uber, a payroll system like ADP, or a big box retailer, like Walmart. Those are places where people make money. And there’s far less access into those systems. It is hard today, for either a business or an individual, to get access to employment information from entities like those. Today, it’s fairly easy to get access into Chase and Wells Fargo. There are many providers that do that and do that quite well. There’s another set, right? There’s the set of income: how do you make your money? And that is the part where I think a lot more innovation needs to take place.
When we get into discussions around data sharing, it brings up questions around who owns the data and who’s empowered to share it. I asked Shmulik about his philosophy around data privacy and user content.
Shmulik Fishman: One, privacy and security should be the fundamental bedrock of how financial services work. Making sure that a user’s information is private, and is not leaked or hacked, means that the consumer is not the core product. All of us need to think deeply about security, encryption, and access rights. It’s one of the core tenets that Argyle builds off of — it’s really the reason why we encrypt everything end to end, everything inside of our rails and our plumbing is encrypted.
It’s also the reason why we take a lot of care to give consumers that interact with our products the ability to both opt out of using our system and to delete the access to a business using their data. We just think that those are core tenets and core rights of a consumer.
I also think that while data privacy is critically important to the products that our industry builds, data ownership is just as important. And data is owned by the person that makes it. When I go to work every day, I make money. If I that’s the thing I do as a worker, that means it’s my data. It’s the same tenet that would be held if I wanted to log into my bank account. That’s my money.
And I think that businesses need to come to terms with who owns the information in their system — that businesses can be custodians of a data set without being the owners of a data set. ADP is a custodian of payroll data — they are not the owner of payroll data. The owner is the person that is making the money. The owner of the data set is the worker. I think the same can be said for Starbucks. I think the same can be said for Uber. Uber is the custodian of how many trips I made last Tuesday — they’re not the owner of that information. So data privacy and data ownership are the core things that the industry is working through and they both go back to the consumer, the consumer controls that.
At the center of any discussion about sharing data is the consumer. The primacy of consumer permissioning has become a staple of any talk about open finance. The industry is providing more and more functionality to users to control their data. Educating them about how to use their permission needs to catch up.
Shmulik Fishman: Let’s take an example like Facebook. How aware is the consumer that the clicks and activities they make on Facebook — how aware is that consumer on how Facebook is monetizing that data? I would claim that they’re very unaware. Whenever somebody interacts with one of our products, we have them opt in to their specific data. Let’s say that somebody wants to share their Starbucks employment records with a lender. We have a screen that says, do you want Good Loans to have access to your Starbucks employment information? That is true opt in. That’s true consent. That is true awareness about data about that specific consumer being shared with a specific entity for a specific purpose. I think that that is good consent, good awareness.
That’s not what is happening in every system. And that’s not what’s happening broadly in financial services today. So I think that there are many different ways to provide good consent. Not every business provides it. I fundamentally believe, business aside, that consent is a good thing, not a bad thing. And that it’s important for products and services to be providing consent.
For healthy ecosystems to arise around employment data, consumer permissioning has to be pushed to the forefront. Getting access to our employment data isn’t just about sharing our personal information — it’s about control.
Shmulik Fishman: The vision is that a consumer is fully in control of their employment records. They get to decide who sees them, for how long, at what granularity, for what purpose, and they can turn on and off that capability when they want to. That’s a long term goal right. En masse, most employment data is sold to a credit bureau, which then decides without any consent from a consumer how that data is used, repackaged, and resold. I want to live in a world — and I think most consumers want to live in a world — where control over how people’s own data is used is with themselves. And I think that the industry should be spearheading that. That’s where the puck is going. So our vision for Argyle is one where we can spearhead this wave of ownership. I invite more businesses to participate in this. I invite more businesses to skate to where the puck is going with us.
It’s often said about free products that if you’re not paying for something valuable, you’re likely the product. Some businesses monetize your information in return for giving you valuable tools, software, and experiences. But there’s pressure afoot that that model will change as people become more selective with whom they share their deets.
Shmulik Fishman: There are many businesses that exist today that make a lot of money off of data not being in the hands of end consumers. And I do expect that a set of companies is going to want to make it harder, not easier for there to be consumer-permissioned employment data access. And as we get larger and provide more services to more businesses under our model, I imagine that there will be businesses that would seek to make that harder, not easier. So that’s the part of being an innovator. That’s the part of trying something new: there are people that are going to want to keep the status quo alive and not shift to this new model. I expect that to be a spirited debate. And I think that that’s the reason why we want to be very clear with the market and have clients on our roster that believe in this way of transferring data and examining a consumer’s employment records.
Fishman’s vision for Argyle is bold. There are a lot of businesses in the US — way more than there are banks — and making their employee’s data accessible is a heavy lift technologically. But get this right and it’s likely better for everyone involved: for consumers and financial services, as well as payroll providers and employers.
Shmulik Fishman: So some statistics, I think there’s about 1700 banks in the United States. And, you know, many banks store data the same way. There were 250,000 payroll service providers in the United States in 2020, says the IRS — all of them store data differently. There are no federal standards at all. The data normalization project for employment records is massive. It’s one that one requires standards to be made in the first place. And then it also requires the consistent maintenance of those standards against the people that are the custodians of that data, and how they might change how they’re storing data in real time.
It is a it is a Herculean feat to make sure that regardless if the data is stored at Target, Uber, Postmates, ADP, Paychex or Paycom or QuickBooks, that you get the same set of normalized fields, like job titles, start date, gross pay, net pay, commissions. In a way, it’s the only thing that Argyle really works on every day, beneath the marketing that we have and the use cases that we can be used for. What we’re really engaged in every day is a massive data normalization project, to make sure that regardless of how an individual platform is storing this information, it gets normalized in the cleanest way possible.
That is something that we can’t do alone, and we invite all payroll processors to participate in it with us. It’s the reason why we’ve created OEDS.org, which is our specification on employment data. We want more people to be members of this organization because we want the best standards possible. We think we have a pretty good framework, but we know it only gets better with more participation. So we want more payroll providers, more employers to participate in developing a standard that we all can abide by. It’s better for consumers. It’s better for businesses, and I think it’s also better for payroll processors and employers.