Four Steps to Creating a Data-Centered Culture (But Don’t Forget Step Zero)


Running a financial institution in today’s market can feel like a never-ending game of catch up. As soon as a bank adapts to one threat, they’re facing a new one. Two of the main issues banks face today are understanding today’s consumer and what they want from a modern bank and guarding themselves against ever-increasing fraud threats. With the growth of alternative banking methods being offered to consumers and because cybercriminals are always diversifying their targets and using stealthier methods to commit identity theft and fraud, it seems impossible to keep ahead of the game.

But adaption and building strategies to address these threats is easier than you think. The answers are likely right at the banks’ fingertips . . . they just have to study the data and hear the story it tells. But to do that, the organization must be primed with a data-driven culture, an operating environment that seeks to leverage data whenever and wherever possible to enhance business efficiency and effectiveness.

There’s no question that data has changed the way every industry, including financial, does business. It’s increasingly clear that banks no longer can operate without incorporating data analytics into the business culture. The amount of data to be leveraged these days is endless. Between contactless payments, mobile channels, social and digital sources banks have a plethora of data to work with. However, with the emergence of disruptive Fintech and Banking models the lack of a data-driven culture in today’s financial institutions is a key component holding them back from competing with the online and tech-driven options available today.

We no longer have to go with our “gut” when there are volumes of data amassed by financial institutions that can be used to create tremendous value. Leveraging it more effectively can provide crisp analytics that deliver knowledge and drive better business decisions. A company-wide digital transformation is necessary to empower both the business and the customer.

But how do you get there? In my view, creating a data-centered culture where everyone knows what data is available, how to access it and is compelled to use it for business-driven purposes on a daily basis boils down to a four-step process. Well, it’s really five steps if you count Step Zero, but I’ll get to that in a minute. This four-step process is designed to help you harness and mine the data for answers to specific business questions and issues.

Step One: Where is the Data?

It’s an enterprise-level game of hide-and-seek. There are vast amounts of data available to every financial institution — credit portfolio, money supply, consumer credit, commercial bank assets and liabilities, industrial production and capacity utilization, flow of funds and more — but we must first be aware of where it’s all located.

Step Two: Cataloging (What Kinds of Data Do I Have?)

Identify what types of data exist within those sources. This is a data cataloging exercise designed to identify the data stored in marketing, finance and e-commerce systems as well as other sources like social media feeds. With a proper data catalog, “users can search for business entities, then find data sets related to them, so they can quickly perform analysis and derive insights.” In other words, cataloging makes data sources more easily discoverable and provides the metadata that describes exactly what data is stored, allowing for classification and labeling of all data, including sensitive information.

Step Three: Time to Tidy Up the Data!

This step, data cleansing, is a huge undertaking but a critical one. Data quality analysis is essential to using your data at all. This process helps you identify fallacies or holes in your information, improving data quality. As an example, maybe in the “column” of customer phone numbers you have three different data sets: some are just numbers, some have dashes and some use parenthesis around the area code. Or, maybe a fundamental entity such as “customer” is not as clear-cut as one might think. For example, a customer (according to your database systems) could be someone who used your bank once, or someone who received a payment from your bank or someone who paid through your bank but never used it again.When you clean your data, all outdated or incorrect information is gone—leaving you with the highest quality information.

Step Four: Theory testing

Now it’s time to bring in samples of your data, run it through an analysis and find out if it answers the questions/ business problem at hand. This step is supposed to tell you something which you didn’t know before. As example, the data can answer questions such as how and where funds are coming from within a specific account. Or “is ACH more popular than checks or wires?” “Are any of my consumers using alternative payments or bill pay?” “What does my customers’ risk scores look like when checking and savings data is combined with credit and loan data?” Those answers and many more are ready and waiting for you in the data.

Now, I’d like to address a problem with this four-step process I’ve just recommended. It’s important to point out here that many banks are getting stuck on step four, unable to really find the actionable insight the data is supposed to be revealing. There’s a reason for that: they skipped Step Zero.

Back to Step Zero: What’s the Question, Again?

This is the most important thing. I’ve not mentioned it until now to illustrate a point, that too many businesses jump into the data-driven process without taking a step BACK first. The very first thing that needs to be done before embarking on any data analysis is Step Zero: Identify the problem. This doesn’t require data analysts, this requires the people IN your organization who know exactly how the business is run, what is working and where the obvious pain points are. You need these businesspeople to tell you what questions need to be answered. Then, you can hire the data experts and go through steps one through four once you know what you want the data to solve.

When you lay this business mindset over the whole process, the ROI happens automatically. Step four starts belting out all kinds of stories and information that help you make better business decisions that are directly related to the business problem you identified in step zero. This output creates incredible value to the organization, which in turn proves to everyone the value of data-centered business decisions. Once you accomplish this, you’ll find you can go back to it again and again to solve any number of issues, such as finding ways to better detect fraud, better meet customer needs with new offerings and act quicker than your competition on multiple levels. A corporate data-centered culturebe it customer marketing, fraud protection or any variety of other topics, will forever be ingrained.



Adwait Joshi is a Chief Seer and an expert on big data analytics. His firm DataSeersprovides a big data appliance for banks that uses the concept of a data lake. FinanSeer is a big data appliance that ingests multiple data sources and creates powerful analytics that help drive reconciliation processes, BSA/AML Regulatory Compliance Monitoring, Complex Fraud Detection and a full 360 view on consumer and business data.