Data infrastructure optimization, availability & security software
Data integration & quality software
The Next Wave of technology & innovation

Expert Interview Series: Erki Kert of Big Data Scoring

Erki Kert is a former bank manager on his way to changing the world of credit scoring with Big Data Scoring. Here, Erki discusses how Big Data Scoring is using data to change the way lenders evaluate individuals applying for credit.

Tell us about Big Data Scoring. What do you do?

Big Data Scoring is helping lenders around the world issue more credit and avoid bad loan decisions.

Traditionally, banks have been making credit decisions largely based on information provided by credit bureaus. However, there are billions of people around the globe who don’t have a credit record (the so-called “thin file” people, e.g. young people, immigrants, emerging market population, etc). Serving the thin file segment has been a huge challenge for lenders, and the segment has been significantly under-banked.

At the same time, living in the 21st century, there’s plenty of information available about pretty much all of us in the online channels. Based on this data, the Big Data Scoring algorithms can predict the loan repayment behavior of every individual, making it possible for banks to seriously lend to this segment for the first time. For consumers, it means that credit becomes available for the ones that actually deserve it and would be limited to those who cannot afford it. Lending becomes more honest.

How has lending evolved since you started your career?

Back when we started, we had to start all meetings with a thorough introduction to big data and the data economy as such. By now, a majority of lenders already understand the need to use the big data in credit decisions, so we can skip the first half of the sales pitch and dive in to more detailed discussions regarding specific business uses.

What are some of the problems or gaps with traditional credit scoring that you’re hoping to solve?

There are two major issues with traditional credit scoring:

First, it relies almost entirely on historical credit information, which means that life is very difficult for first-time credit applicants. And keep in mind, all young people and most of the emerging middle class of emerging economies fits the description. It’s billions of people around the world.

Second, the traditional credit scores are local, i.e. attached to a specific country. In today’s world where people like to travel a lot, it creates a lot of friction in the system (e.g a German woman with a good credit score at home isn’t able to access financial products in the U.K.).

How can Big Data Scoring fill in these gaps?

The first gap is filled by not basing the credit assessment on past credit information, but on a lot wider picture of the individual. We gather all the data we can find about an individual and make a thorough assessment taking into account hundreds of different factors.

The second gap (traditional scores being internationally unavailable) is also solved by the type of data we use. Regardless of where you go or travel, the data in the online world always travels with you, and hence, the consumer has easier access to credit wherever he/she is.

What should lenders be looking at to get a more complete picture of a loan applicant’s financial history?

I would actually suggest going a few steps further than that. Looking at “applicant’s financial history” is something that banks are doing today, but this is often not enough to have a good understanding of the person (especially if there is little or no financial history available).

Also, looking at the financial history might tell something about the person’s ability to repay a loan, but not that much about their willingness to pay. Hence, what we suggest is to look at a lot more data than just financial history. We search the web, monitor customer online behavior, analyze e-mail and phone usage, investigate the geo-location, etc. There are thousands of additional data points out there that help us really understand who those people are that ask for a loan.

What are the benefits or results of using Big Data Scoring versus relying on traditional credit scoring?

The two main benefits are:

1. lenders are able to issue more loans (i.e. more people have access to credit), and
2. lenders are better able to avoid lending to people who are not able or willing to pay the loans back.

What are some of the other interesting ways you’ve found the banking/lending industry use Big Data outside of scoring?

More and more, data-driven decisions are guiding the entire client journey – starting from marketing and client acquisition up to product selection and credit underwriting. Having more client data helps to make better decisions in nearly every aspect of the customer journey.

What are the Big Data trends and/or headlines are you following right now? Why?

A relevant topic for us to always keep on the radar is how different countries are opening up their data (e.g. U.K.’s open data initiative), and at the same time, how regulators are responding with new rules and instructions. Obviously, we’re also keeping track of our competition and the very interesting developments of emerging markets.

As seen in the video “Simplifying Big Data Integration with a Modern Data Architecture,” Syncsort is focused on enabling Big Data analytics use cases – like Big Data Scoring – and facilitating regulatory compliance in connecting enterprise-wide data sources, including mainframe data, to Big Data platforms such as Hadoop and Spark. 

Related Posts