Big Data and Online Scoring: Fintech and Beyond

The world is certainly excited about the concept of big data and advanced analytics and it’s not just because of the data are big but because the potential for impact is big. 

Read our full research “Money of the Future”. Download PDF (20MB)

Big data is like a teenage sex: everyone talks about it, nobody really knows how to do it, everyone else is doing it, so everyone claims they are doing it. Admittedly, “gig economy” and “big data” are two phrases that are about as buzzy as you can find in the modern business world. But that’s no reason to dismiss either one. Both consumer (B2C) and business-facing (B2B) companies are facing intensifying competition when it comes to customer insights, and the data-science field is expanding in order to help meet the challenge.

Not all data analysis qualifies as “big data,” of course, but solutions are multiplying, and 2016 may well be the year that more companies seriously ramping up their investments in the field instead of just talking about it.

According to Forbes, only 15% of Fortune 500 companies are currently using big data analytics—and indeed, big data isn’t appropriate to many small and mid-size companies—but one trade group reports that businesses are pouring considerable resources into it: Big data spending is projected to grow by a CAGR of 23% each year until 2019. The ability to develop an integrated, analytical view of customer activities and business operations will separate the wheat from the chaff more decisively and rapidly than we’ve seen so far.

Anne Wojcicki, 23andMe‘s CEO, wants to improve drug discovery (and health insurance?) with big data and tailor drugs to individuals so they are more effective. When 23andMe launched its first product—a mail-order test that analyzes DNA from a saliva sample — the company wants to use the same genetic data to make better drugs.

“I really want to prove that we can revolutionize drug discovery, in almost a Moneyball kind of moment,” Anne Wojcicki, CEO of 23andMe, Itroymanagement, told a crowd at Civic Hall at Fast Company’s Innovation Festival. “Having a human database can give us so much of an edge—we can do it so much faster and so much more inexpensively.” If big data can be used in the rest of our lives—such as marketing products to us at chain stores—why not harness it for healthcare as well?


“We don’t know much about health…It’s crazy to me that in this world of electronic medical records, Walmart has so much information about how we shop, but no one has that information about our health,”


Wojcicki said. “Why can’t my doctor say, ‘Wow, Anne, based on your lifestyle and behavior, you’re five years from being diabetic.’ But I can go to Target and they know exactly what I’m going to buy.”

After months of polling in the middle of the Republican pack, presidential candidate Ted Cruz recently leaped ahead to first place in Iowa and second place behind Donald Trump in national polls. The Texas senator may have pulled ahead of the gaggle by mining data on “tens of millions of” Facebook users, reports theGuardian.

The information was reportedly gathered by Cambridge Analytica, a data company working exclusively for Cruz’s campaign—and was obtained largely without the knowledge of Facebook users. Big data is helping to bring transparency to the darker corners of politics.

Big data might change the way you find a job. CarrerLabs uses big data to explore all aspects of a company, from maternity to morale, growth, and financial health.

In the past the talks and euphoria about “big data” were left to mathematicians and their models and the monetization could be seen only in the distant future. But in the past year we witnessed a huge number of new players in this sector, and the discussions moved into practice.

The growth of online lending services and rising interest of telecoms and internet giant created a demand for online scoring, which takes into account not only traditional data (a passport and banking history), but also data from social networks, smartphone manufacturers and mobile operators.

A wave of startup are using unconventional information to decide if customers are credit worthy. Over the last few years, a number of technology startups have created ways to underwrite loans that don’t rely on the traditional methods. To evaluate the factors that go into administering a loan—identity, fraud risk, credit worthiness—the startups look at mishmash of seemingly tangential information about applicants, such as activity on their social network profiles, how often they use their cell phones, and whether they frequently send text messages.

The startups argue that this constellation of data can better (as in faster, more cheaply, and more accurately) hone in whether the applicant can replay the loan by analyzing how they live, the decisions they make, and how they interact with others. If it all checks out, the cash is transferred to users within minutes of applying for the loan.

Loans are typically in the hundreds of dollars, although some startups would like to increase the size they gain users and better establish the creditworthiness of their borrowers. Regardless, there is a steep price for people who use the services: Interest rates can sometimes go as high as 30% on the transactions.

PayPal co-founder Max Levchin is one of the leading voices in the movement. In 2013, he founded Affirm, a company whose technology combs through a wide range of sources to help decide if applicants should receive loans. For example, Affirm wants to know whether a potential customer has contributed software code to digital code libraries like GitHub.

What does that tell it? That the customer is who he or she says they are. Other metrics can help assess his or her ability to repay a loan. (Like other companies in the market, Affirm is tight-lipped about how those data points specifically factor into a credit decision, to safeguard its algorithms against snooping competitors.)

ZestFinance, another startup, is using big data to make credit decisions. Rather than using simple credit checks, the company looks at certain variables on a credit report along with how people use smartphones and social network. ZestFinance says that its use of bulk data collection delivers a 40% improvement in default rates over current industry scoring methods. The company didn’t provide its typical default rates.

While some companies are focusing on younger borrowers, others are also focusing on emerging markets. One such company, InVenture, told that it provides a much-needed service to people in several countries in Africa, where credit scores don’t exist and loans are rare. People looking for loans can install InVenture’s app a mobile device. The app then tracks how they use their phones. Sending too many text messages or a low battery may be a sign that an applicant is a credit risk, based on historical data it has collected and analyzed, the company’s CEO Shivani Siroya told.

In contrast, people who make more calls at night or use gambling sites may be more credit worthy. Those activities suggest price sensitivity (calling rates are lower at night) and, in the case of gambling, an ability to pay debts.

London-based Aire.io, the startup that is democratizing credit scores for people with thin or no credit files announced a $1.1 million round of funding in January 2015. The funding comes from existing investor SparkLabs Global Ventures, with additional contributions. Aire pulled in its initial Seed round of $300k in the summer of 2014. Aire’s recently launched Credit API enables lenders to perform a credit check on people who do not have financial history on file with traditional credit bureaus. Its predictive algorithms assign a score to consumers based on self-declared data and other reliable sources, such as LinkedIn.
Another mobile startup, Branch, uses a similar method. According to the company’s website, after users log into its mobile app, it can determine their creditworthiness based on “SMS logs, call logs, and contact lists.”

Like InVenture, Branch believes that how people interact with others may ultimately determine their likelihood to repay a loan.

ZestFinance, a financial technology company whose algorithms enable it to make loans to borrowers traditionally not served by banks, pulls data from alternative credit bureaus to find out, for instance, if she’s ever declared bankruptcy or what she pays in rent. Since 2010, the company has used unconventional signals to make lending decisions to borrowers who would traditionally go for payday loans.

This summer, it began expanding the markets to which it applies its algorithms, announcing a partnership with JD.com, one of China’s largest online retailers, and, of course, the Basix Loans for near-prime borrowers. Dominick Ruggiero, Fortress managing director, said: “ZestFinance is a leading financial technology company poised for significant growth as it applies machine learning to disrupt consumer lending and meet the needs of consumers.”

The startup uses machine learning to determine, based on many data sources, whether people are eligible for loans — and now other companies can take advantage of this information to loan money to people who might otherwise be out of consideration.

Among these is Kasisto, a spin-off venture of SRI International—the creator of Siri. The startup is testing a voice-recognition add-on for mobile banking apps that lets customers ask questions about their accounts. Users can ask Kasisto, “How much have I spent on fees?” or tell it, “I’m looking for a three-dollar transaction on my checking account,” and the system will return an answer.

It’s a voice-activated assistant that, unlike Siri, isn’t a generalist. Kasisto has deep knowledge of two things: the semantics of the financial services world and your own relationship with your bank. It then uses natural language processing and machine learning to serve up the information you’re looking for.

The company, which launched last June, has been in “friends and family” testing mode in the US and Asia, but its roster of clients already includes Spain-based banking group BBVA and Wells Fargo, among others. Kasisto CEO and co-founder Zor Gorelov explains his company has basically built two types of apps: an enterprise add-on that corporate clients—especially company CFOs and treasurers—will use internally, and a consumer-friendly app that will integrate into existing mobile banking apps. The team plans to roll the app out to consumers later this year.

Meanwhile, MoneyStream, a new service from a Silicon Valley startup of the same name, links your bank account to a range of services that together deliver personal finance predictions in the form of a simple calendar. In other words, you can see how much money you can expect in your bank account month to month. While other apps, notably Level, already aim to help you plan your daily budget, MoneyStream claims it can help you dig deeper into the more complex fluctuations that can upend the best-intentioned plans.

Simply link your bank account to the service (the web app is the most robust landing page for now), and it identifies your sources of income, your recurring bills, your credit cards, and your loans. After scooping in all the data it can glean from your bank statement, you can add new bills, utilities, accounts and credit cards, and also manage your email notifications and alerts.

An algorithm analyzes this “stream,” and, using AI, projects how much money you’ll have in the bank for months to come, showing you the information in a straightforward calendar format. The system gets better with use as the stream gathers more historical data and users correct individual entries.

China’s proposal is like a credit score that could encompass your entire life, from work performance to Internet activity. China’s proposals for a “social credit system” don’t seem that radical when you read the dry, official plan posted by the government. “They’ve been working on the credit system for the financial industry for a while now,” says Rogier Creemers, a China expert at Oxford University.

“But, in recent years, the idea started growing that if you’re going to assess people’s financial status, you should equally be able to do that with other modes of trustworthiness.”

China is proposing to assess its citizens’ behavior over a totality of commercial and social activities, creating an uber-scoring system. When completed, the model could encompass everything from a person’s chat-room comments to their performance at work, while the score could be used to determine eligibility for jobs, mortgages, and social services.

The document talks about the “construction of credibility”—the ability to give and take away credits—across more than 30 areas of life, from energy saving to advertising. “It’s like Yelp reviews with the nanny state watching over your shoulder, plus finance, plus all of these other things,” says Creemers, who translated the plan.

The versatile internet powerhouse Alibaba group is now sitting on the goldmine of big data and is innovatively monetizing it through internet finance. This time Alibaba Group’s online payment system Sesame Credit applies the cutting edge big data-based credit rating system in partnership with Luxembourg’s Consulate General in Shanghai for a launch of a credit-based visa application service. The regular process of obtaining a Luxembourg Visa is quite complicated in China. Bank records have to be handed in as proof of assets and the process is time consuming. But now, Chinese traveler can easily obtain a Luxembourg visa through Alibaba’s tour service Alitrip with a certain level of Sesame Scores. For applicants with a 750 score or more, the Luxembourg Consulate in Shanghai will accept their personal Sesame Credit reports as a proof of financial capability. All original documents will still need to be provided to the Consulate at a later stage as Schengen rules require.

Credit Karma (US) as best example of big data’s monetization (now)

Everything from renting an apartment to getting a business loan often requires a credit check, which can make life difficult for the 15 million Americans who don’t have enough credit history to generate a score. These people are referred to by credit bureaus as “thin files,” and often they’re in the dark about what’s on their credit history and why it’s so empty.

In April 2014, Credit Karma, valued at $3.5 billion, began offering tools to help these people by giving them free credit information. “Most thin file consumers that sign up for Credit Karma will be able to review what is currently on their credit report, as well as learn about why they don’t have a credit score and how to build their credit history,” the company said. It also offered to direct them to specific loans or credit card options that could help them build their credit history.

Credit Karma announced it signed up its 50 millionth member, a major milestone that gives the online personal finance platform insight into more than one-fifth of America’s total household debt. Credit Karma’s 50 million-strong member base accounts for 22 percent of all Americans with a credit profile in the United States, who use the platform to help them manage over $3 trillion in household debt. The company has given away more than one billion free credit scores since launching its product in March 2008, without charging members a penny to use its service.

Founded to bring new levels of transparency to consumer finance beginning with free credit scores, Credit Karma’s product expanded quickly to become a financial technology leader: adding full credit reports, daily credit monitoring, financial account monitoring, full credit information from two of America’s major credit bureaus, as well as educational tools and content for its members.

In 2015, it became the first platform to offer people without enough credit history to generate a credit score full weekly access to their credit report; and then launched Direct Dispute™, a free tool that allows people to dispute credit report errors directly from one of their credit reports using Credit Karma.

It continues to expand its completely free offerings including its Credit Score Simulator, credit monitoring and friendly, personalized information to help each person understand and make the most of their individual situation.

Kenneth Lin’s (CEO of Credit Karma) parents emigrated from China when he was 4 and worked in the kitchens and at the blackjack tables of Las Vegas casinos to help put him through Boston University. Their sacrifices, he says, made him want to do something big: “Lots of people look at us and say we’re disrupting the way credit scores work. What we’re building is much larger than that.”

In 2003 Congress gave Americans the right to request their credit bureau reports for free once a year, but before Credit Karma they generally had to pay to see their credit scores. Credit Karma makes money–how much Lin won’t say–by recommending loans and other products consumers qualify for based on their credit histories and then streamlining the application process.

With credit card issuers such as Discover now providing free scores, too, Lin is staying a step ahead by offering consumers a quick way to dispute credit report errors (so far 650,000 have used it) and new tools to understand and raise their scores. He’s also developing a service to alert members when their scores have risen enough that they should be able to demand lower rates and save some bucks.

In December 2015 Credit Karma has made its first acquisition. The company has acquired the makers of the mobile application Snowball, with plans to leverage the team’s expertise in mobile notifications. Terms of the deal were not disclosed. That app grew to 250,000 downloads, but didn’t really take off the way they hoped.

This summer, the team pivoted to building a “priority inbox” for all of Android’s notifications. Though Snowball’s next version never topped half a million installs, the process of building the app itself was something of a technical feat, as the team figured out how to take over the entire pull-down notifications interface on Android, as well as the full notification swipe itself. It’s this kind of expertise that Credit Karma is now interested in bringing onto its team as it rolls out further enhancements to its own set of mobile applications in the near future, explains Credit Karma CTO Ryan Graciano.

Credit Karma, meanwhile, raised $175 million in summer 2015 at a $3.5 billion valuation. Credit Karma has exploded in popularity because of one core feature: it gives you two free real credit scores, and tells you the exact factors that go into them.

This means that instead of just saying your credit score is 700, it tells you how you scored on things like payment history, age of credit history, credit card utilization (how much of your credit limit you spend each month), and number credit inquiries (how man people are checking your credit). Credit Karma also tells you how much each of these six factors it looks at affects your score, and how you rank next to your peers.
In short, Credit Karma is valuable because it provides a detailed way to understand where you are with your credit score, and the ways you can improve.

From Philippines to unbanked world: journey of Lenddo and Ayannah

The Philippine Long Distance and Telephone Co (PLDT) is planning to invest around $100 million in its Big Data business, meant to help companies and government agencies develop more effective products, services, and programs for their target markets, a top company executive said. “The overall budget is roughly $100 million,” said PLDT chief strategy officer Winston Damarillo, noting that the telco giant has previously invested $30 million in developing its Big Data platform, during the last three years.

With demand for data solutions to profile borrower risk in emerging markets, Lenddo has closed a new round of funding in October 2015. The round was led by AT Capital and Life.SREDA with additional participation from existing investors: Omidyar Network, Blumberg Capital and Golden Gate Ventures.

Lenddo added that Co-Founder Richard Eldridge is taking on the role of CEO as part of the financing. Commenting on the funding, Eldridge stated that “the new investment will allow Lenddo to fast track growth into new markets, expand Lenddo’s presence in its existing markets and continue to develop new innovative products”.

As online based lending continues to rapidly grow, with several marketplace lenders such as Lending Club, OnDeck Capital, SoFi and Kabbage reaching billion dollar valuations, demand has been triggered for new ways of analyzing loan risk. Leading this trend are advancements in big data which firms are using to analyze non-traditional credit scoring information. Data points include reviewing borrower’s social graphs, the university they attended, their job title and what car they’re driving.

Specifically in regions that lack agencies to provide traditional credit scores, demand is high for non-traditional big data risk profiling solutions. As a result, these types of data solutions have been an important driver of increasing financial inclusion in regions with high percentages of under-banked population.

Due to the complexities of creating and maintaining big data risk analysis solutions, it has led to demand for third party products which can be licensed by online lenders. Among these vendors is Lenddo, who provides P2P lenders, banks and card issuers with an API based solution that can be integrated to risk profile their clients. According to Lenddo, the firm’s solution analyzes over 12,000 data points of customers seeking loans. Headquartered in Manila, the Philippines, Lenddo’s services are used by firms in 10 countries.

With its plethora of e-payment services and large customer base – more than one milion repeat customers, Philippine fintech-company Ayannah is well-positioned to leverage the enormous payment data it is accumulating to provide value-added digital financial services such as identity verification, credit scoring and lead generation. “Payment services can be very idiosyncratic; each country has their own payment infrastructure and payment problems. In each of those countries, there is probably [something]  like Ayannah that is focussed on solving those payment problems.

What we can bring to the table is analytics,” says Mikko Perez, Founder and CEO of Ayannah. Its first Big Data initiative is called Project PIGLET. Ayannah is utilising “recently declassified Big Data analytics and machine learning technologies used by the CIA and NSA” to profile the large unbanked population in the Philippines and develop the very first risk-based credit scoring system in the Philippines.

“Take, for example, a Filipino maid in Singapore. Her employer pays her cash and so she goes to Lucky Plaza (a shopping mall) to remit money twice a month. She has been doing this for 10 years religiously, but she does not have a bank account. So if you are using traditional credit scoring rules, she may not even be eligible for a credit card. But since we can measure how conscientious she is because we see the frequency and consistency of her remittance transactions, we could assign her a higher score as a possible candidate for additional credit or access to financial services like insurance or savings products,” explains Perez.

Currently, Ayannah possesses over a half a billion payment records to profile customers for financial services such as SME loans, insurance eligibility, car and mortgage loans. These profiles are then created by cross-referencing customers’ payment data with their top-up data, remittance data and social media data. “We want to target insurance companies, property developers or anyone looking for new customers who do not have access to this broad segment of the emerging middle class that may not have a complete credit history or hard assets to qualify for a loan. The next billion customers basically,” says Perez.

Ayannah’s other Big Data initiative – Project COMPASS – will combine offline and online analytics to build an omni-channel predictive and prescriptive analytics network to increase traffic and conversion for retailers. According to Perez, the initial trials in Singapore and the Philippines have been very encouraging.  Through monitoring and analysing the in-store behaviour of offline customers, Ayannah was able to gain actionable insights that help retailers stock better and engage their customers to increase conversions and sales velocity.

This new data source of offline and in-store customer behavior is even interesting for the Internet Unicorns such as Google and Facebook as they both race to understand the behavior of the next billion consumers in emerging markets who continue to shop in brick-and-mortar retail stores,” explains Perez.

“We will announce the launch of retail analytics innovation centres in Singapore and the Philippines and partnership with some very large publicly-listed retailers as well as leading Silicon Valley data science companies. We believe that the data analytics service we will deploy soon will be relevant not just in Southeast Asia but also in the rest of the world,” he adds.

It is also working with other startups to roll out a scholarship programme by next year so that it can hire more software developers and data scientists from the Philippines’ leading universities.
People are repositories of rich information and contextual data, Shivani Siroya, CEO and founder of InVenture told. And yet in most economies a single number – the credit score – remains the blunt instrument by which we evaluate who can receive loans, and who is likely to pay them back.


The result is that across the globe 2.5 billion people have no chance of receiving loans that can fund new businesses and lift entire economies out of poverty.


Her company puts InVenture’s science into practice with its app Mkopo Rahisi, which launched in Kenya and has already loaned $1.5m (£975,000) to businesses and individuals. With this app, InVenture considers people in the context of their financial transactions, but also data like their social media accounts and web searches, to build a rich profile of their personality.

From this it works out the terms of loans and delivers them instantly to mobile phones. “We service all our loans digitally through services like SMS, Facebook and WhatsApp,” Siroya said. “It’s a financial identity that looks more like a person and less like a score […] It’s an entirely new way of thinking about financial products.”

“We do this in under one minute,” Siroya added. InVenture can understand who pays their bills on time, who uses capital wisely and where their sales come from. As a result they can ensure a better repayment rate in risky markets (they have reached 85 percent repayment in Kenya on 40,000 loans, and a 92 percent repeat rate). More crucially, they have found a way to build financial identities for those who otherwise would have none.

InVenture now plans to launch Mkopo Rahisi in more countries, building on existing markets in Kenya and Tanzania to two more countries by 2016. But the firm is also thinking bigger: with a more “vivid and accurate” picture of businesses and people in Kenya and other countries, an entire spectrum of financial services could become viable.

Next steps of online-scoring: debt collection, KYC for e-commerce and recruitment

Old-school financial institutions are typically slow-moving giants. And that’s a shame, because banks also tend to accumulate deep troves of data on their customers that goes mostly untapped. If you’re a consumer looking for an answer to a specific question about your finances, tough luck. Your usual recourse would probably involve a lot of digging through bank statements and bank website pages, or endless hours on the phone with a customer service rep.

But as of late, a swell of banking startups are seeking to change this. They take all that undifferentiated data tucked into your bank statements, and then, harnessing artificial intelligence, transform and organize it into helpful information that people can actually understand—and act on.

In developing these tools, Kasisto and MoneyStream join a multitude of other companies that are riding the cresting wave of artificial intelligence to surface information that’s useful for humans. Companies from Google to Facebook to Baidu are using AI to drive voice and image recognition. Smaller startups are using AI to rethink stodgy practices from job hunting to finding a restaurant that perfectly suits your budget and dietary needs. Now, both Kasisto and MoneyStream aim to inject personal money management with the same AI touch.

The credit score has become a ubiquitous way to assess someone’s credit-worthiness. It is used in a host of underwriting decisions and background checks, from mortgages to auto to business loans, apartment rentals and even employment candidacy. There is an opportunity for scoring to include more people and more financial transactions, and for scores to be used by consumers as well as providers for a host of credit and non-credit decisions.

That opportunity starts with designing with the customer in mind. It includes creating shared value in tracking, maintaining and building true financial health — not simply coaching customers to improve their scores at a few, discrete moments when credit checks are necessary.

[su_pullquote align=”right”]”60% of employers pull current and potential employees’ credit reports.”[/su_pullquote]Debt collection is entering the social media age—and sometimes outpacing laws meant to protect consumers. As landlines die and a simple screen tap can deny any unwanted phone call, debt collectors have started relying on new modes of communication to contact debtors, from text messages to social media sites. In the process, they may also be breaking consumer protection laws. While Federal Trade Commission regulations don’t explicitly refer to social media, in 2011, the FTC did address how text messaging can and can’t be used to lawfully collect debts. Regardless of the medium, however, it’s unlawful for the collector to harass the debtor or violate his or her privacy, by, for instance, communicating with relatives and acquaintances.

While debt collectors are prohibited from harassing the debtors they pursue on social media, there are no rules prohibiting the use of social media for locating or learning about debtors. This sort of data collection—for the purposes of debt collection and beyond—has only grown exponentially with new technology. We give up more information about ourselves, from the geolocation information we yield when we check in on Foursquare or add a location to an Instagram photo, to the information we let apps have when we install them on our smartphones.

Another typical situation – in the land of job hunting: you’ve managed to create a résumé that made recruiters drool. You’ve successfully run the obstacle course of video screenings and in-person interviews. Then, a recruiter tells you that the company will be doing a background check. No wild past—whew!—so you think you’re in the clear. But next you think about the thousands of dollars in student loans you have (not to mention that maxed out Nordstrom card). Uh-oh. So here’s the million-dollar question:


“Can HR recruiters actually see my credit score—and can they reject my application based on my financial situation?” 


A study by the Society for Human Resource Management (SHRM) showed that 60% of employers pull current and potential employees’ credit reports. Companies pull credit reports in order to determine how financially stable you are. That means whether you’re toting a Birkin or a backpack, employers want to make sure you won’t raid the company petty cash trying to pay off your iPhone bill.

Photos: getty, Company profiles.

Life. SREDA VC is a global fintech-focused Venture Capital fund with HQ in Singapore