In this era of Big Data, it seems like financial services companies know everything there is to know about our lives as consumers - where we live, what we do for a living, how much we make, how much we have saved, what we buy and what we might want to purchase in the future.

But the reality is there are huge segments of the population in the U.S. and globally about which these companies know very little. Sometimes that's because people have left very few data "breadcrumbs" offering clues about themselves - they're unbanked or underbanked and lack a credit history. In other cases, consumers have left a trail, but it's not accessible to the company or agency that needs it to asses someone's worthiness for a credit card, a cell phone plan or an apartment.

And in still other instances, companies just don't know what they don't know. They haven't put systems in place to really get to know or collect data about groups of consumers who don't look like the people who populate the firms' own front offices - people who aren't white, aren't straight, don't live in a big city, who may lack college degrees or may have recently immigrated to their current country of residence.

What Do We Know About Jane?

To illustrate the importance of perspective in the sector, Adrienne Harris, a former special assistant for economic policy in the Obama administration, used the example of a fictional woman named "Jane" who makes about $60,000 a year. Harris described Jane's life starting from when she gets her biweekly paycheck - with step one being a trip to a check cashing business.

"In 2018, 55 million adult Americans were unbanked or underbanked, or about 22% of households - in the age of fintech and mobile payments, almost a quarter of the population is underbanked," said Harris, who is a member of the board of directors of the Financial Health Network (formerly the Center for Financial Services Innovation). "Jane goes to check casher and pays 3% or about $51. If she does that all year, it will cost her $1300 to have immediate access to money she's already earned."

Jane's month also included a trip to a payday loan business and having to draw late fees on a number of her utility bills because her paycheck doesn't stretch far enough to pay them when they are due. More than three quarters of renters make these same trade-offs or strategic payments every month, Harris said.

"It is damn hard to be middle class or what we call middle class in the U.S.; in fact, it is almost impossible to be middle class and to make ends meet," she noted. "It's not about people not knowing how to manage their money or having the right tools, it's just hard."

And despite knowing all of that data about people like Jane, Harris said there are still wide gaps in financial services companies' knowledge - and the gaps become even wider if Jane is a member of one or more minority or traditionally underserved groups.

"What if Jane is single? What if Jane is black? What if she's an immigrant? What if she's in a same-sex relationship? What if she's black, gay, single, an immigrant and caring for family in the country she came from and has a little one on the way?" Harris asked. "What do we know about Jane? Now we know nothing - we don't know anything about Jane."

Fintech as an industry has made improving and democratizing consumer financial health a key aim, but in order to do that, the sector needs to commit to learning more about women and how they spend their money, Harris said. "Women are the financial decision-makers across all sorts of demographic combinations. It's not just women and how they compare to men. We know they make less, we know they invest less, but my challenge for all of you is: What else you got?"

"Women are the financial decision-makers across all sorts of demographic combinations. It's not just women and how they compare to men. We know they make less, we know they invest less, but my challenge for all of you is: What else you got?"

- Adrienne Harris

Often firms train algorithms based on historical datasets, which may not have included minority groups.

"If minority groups were not given credit in the past, [the algorithm] might replicate that," said Kartik Hosanagar, Wharton professor of operations, information and decisions, in an interview with Knowledge@Wharton. Hosanagar is also the author of the recent book, A Human's Guide to Machine Intelligence: How Algorithms Are Shaping Our Lives and How We Can Stay in Control. "There are ways to address that in algorithm design. When they train their algorithms on biased datasets, they can intentionally inject some 'noise' into their models. Adding this noise would mean exploring a little bit because the algorithm has not seen enough data from minority groups."

In order to deploy algorithms successfully, Hosanagar suggested companies set up methods to formally audit the technology before it is deployed, "especially in socially consequential settings like recruiting." The audit process must be done by a team other than the one that developed the algorithm, he added.

"The audit process is important because it will help ensure that somebody has looked at things beyond, say, the prediction accuracy of the model, looking at things like privacy, bias and fairness," Hosanagar noted.

'Tremendous Potential'

"[Cryptocurrency] is not going to empower people in and of itself; it's the support ecosystem that's going to do that. If we give more transparent information to more actors in the ecosystem, can we create accountability and create a feedback loop?"

- Sheila Warren

Governments and businesses should also be wary of viewing blockchain or cryptocurrency as a "silver bullet" that will solve financial inclusion problems, Warren said. "It's not going to solve human-led problems. It's not going to solve people trying to cheat other people," she noted. "It's not going to empower people in and of itself; it's the support ecosystem that's going to do that. If we give more transparent information to more actors in the ecosystem, can we create accountability and create a feedback loop?"

Disrupting from Within

Warren said that part of creating an inclusive and empowering ecosystem is making sure that the sector itself is diverse. "One of the mistakes of [Silicon] Valley 2.0 was the lack of inclusion and diverse voices," she said. "We're trying to avoid that in this space; we're trying to do better."

Part of that is building internal teams that are themselves inclusive and also "including the community you're working with from the jump," she noted. "It's not hard, but it has to be done mindfully."

Her comments were echoed by Cassie Divine, vice president and QuickBooks online platform leader at financial software company Intuit, and Kathy Tsitovich, director of partnerships in the company's consumer group. To change the story of Jane, "it will take a ton of disruption and invention," Divine noted.

Although Intuit has a commitment to diversity, it's an evolving effort and at the moment, Divine noted that she and Tsitovich are often "still outnumbered." "How do you bring in that perspective when there are not enough voices and amplification to do it all?" Divine asked. One of the steps her company has taken is to try to give use cases to its product teams that include underrepresented groups or "a market that looks nothing like where we are. Stories that really get people to understand what it's like to be different."

Tsitovich said the company also tries to be mindful of this when putting together teams to work on projects. "Product teams have been more male-dominated because they're mostly engineers," she said. "But if you create mission-based teams and pull in people from marketing or finance, those areas tend to have more women leaders and you get different perspectives."

She noted that aligning teams around a mission also has the added effect of putting more of a focus on problem solving rather than "just building something."

"Product teams have been more male-dominated because they're mostly engineers. But if you create mission-based teams and pull in people from marketing or finance, those areas tend to have more women leaders and you get different perspectives."

- Kathy Tsitovich

Case Studies

When Nicky Goulimis moved to the U.S. from the U.K. several years ago, she unwittingly found herself in a position similar to the fictional Jane.

"I tried to get a credit card and was rejected. I tried to get a mobile phone and was shifted to a prepaid plan. When I tried to get an apartment, they wanted a six-month deposit," she recalled. "I had a great credit history in the U.K., but I had no financial identity in the U.S."

It's familiar situation for many of the 250 million immigrants living around the world. The largest number of them, about 20%, live in the U.S. "Why is this challenge of financial identity still prevalent? It sounds very basic: How can I be understood from one country to another?" Goulimis said.

Her experiences inspired Goulimis to co-found a startup, Nova Credit, that aims to solve the problem. Nova acts as a "cross-border credit bureau" aggregating data from credit bureaus globally in order to provide financial institutions with a standardized "credit passport" that shows a person's credit history in all the countries where he or she has lived.

"We're unlocking a new population for lenders, for property managers here in U.S. who are looking to grow," Goulimis noted. "We enable consumers to access products they would be unable to get otherwise and give international credit companies the ability to help their customers and monetize data."

During the 35-day government shutdown that ended early this year, U.S. Commerce Secretary Wilbur Ross was criticized for suggesting that federal workers who were short on cash should go to the bank and take out a personal loan. While that is easier said than done, Ennie Lim, co-founder and CEO of HoneyBee, said it's all too common for many Americans to fill in their financial gaps through payday loans.

"The payday industry is a $90 billion industry; some of the most vulnerable families will borrow $5,000 and repay $4,200," Lim said. "There are actually more payday lenders than McDonald's locations."

"There should be a feedback loop from the user back to the firm. It's like being able to say, 'I don't like this ad,' and the system adapts and stops showing you similar ads."

- Kartik Hosanagar

Her startup, Honeybee, aims to offer an alternative. The company partners with employers to offer short-term, no-interest loans to employees, who use the cash reserves from paid vacation days as collateral and are able to use the loans to build a credit history. "I built credit with a T.J. Maxx rewards card when I moved to the U.S.," Lim said. "Seven in 10 Americans don't have access to reportable credit."

The Data Problem

Even if companies are able to learn more about the Janes of the world and better tailor financial products for them, there is also the question of: Who owns Jane's data? And if it's Jane, how can she be empowered to use it wisely?

"Most people have multiple financial accounts, and if you want to get a holistic picture of someone's financial life you need access to all of those accounts," said Jeff Cain, head of the Envestnet Yodlee Incubator. The company offers account aggregation services to see multiple bank and credit accounts, along with e-mail, travel rewards and other services, on one screen.

"Historically there has been tension between folks like Yodlee, who provide data, and banks," Cain continued. "Yodlee's position is that it is the consumer's data, it's your data, and if you want to make it available to third parties you should be able to be in a framework to keep it safe."

"I can actively control permission data in a way that it actually could be considered a unit of currency, a unit of value, that's tradeable."

- Jane Barratt

While there is growing debate in Washington and elsewhere around who owns consumer data and what new regulations should be put in place, a lot of the talk tends to focus on the extremes, Goulimis said. She noted that in the future, there need to be policies put in place that allow for the creation of comprehensive credit reports, but also one that "empowers consumers and breaks up monopolies."

In order to better empower - and educate - customers on the use of their data, Hosanagar said firms need to emphasize transparency and user control. "Transparency starts with letting the user know what kinds of data are being used to make decisions," he noted. "For example, what were the different variables used in making a mortgage application? You also need transparency with regard to the models, meaning explanations of not just data used, but how they are used to arrive at a decision."

Often consumers have very little control about how their data is used, and that can cause us to become overly passive, Hosanagar added. "There should be a feedback loop from the user back to the firm. It's like being able to say, 'I don't like this ad,' and the system adapts and stops showing you similar ads."

In order for this type of enhanced relationship between consumers and companies to work, the firms must prioritize building user interfaces that are intuitive and easy to understand. When people unsubscribe from online services, they are usually provided with a lot of information designed to help convey the value of the service and convince them to change their mind. "Companies rarely give the same thought to empowering users about data," Hosanagar said. "Instead, they mostly focus on regulatory compliance and proceed."

Companies have historically been opaque about what data they collect and how they use it, but open APIs and other changes have made it possible for humans to "get the benefit of their own data," Barratt said.

"We're moving to a place where people will be able to monetize their own data," noted Barratt, whose company recently published a white paper on the topic. "I can actively control permission data in a way that it actually could be considered a unit of currency, a unit of value, that's tradeable."

Barratt noted that fintech and traditional banks have an opportunity today to use their platforms to empower people to understand their personal data and improve their financial outcomes.

"A lot of talk about financial literacy puts the onus back on the human while completely ignoring systemic issues like wage stagnation, taxes, all the things that go into people's financial life," Barratt said. "We need to think of using data less around helping people understand money and think more through the lens of advocacy: How can I advocate as a professional for better outcomes for people of the world?"