Data analytics: What not to do

Register now

By now most credit unions have heard they need to do more with data collection and analysis, but many are not sure of how they should be using “big data.”

In 2017, CUNA Mutual Group launched AdvantEdge Analytics as a way of helping credit unions with the process. During the National Association of Federally-Insured Credit unions’ 51st annual conference in Seattle recently, two executives from the firm offered tips and tricks for credit unions to better navigate the data journey ahead of them.

A CUNA Mutual Group member survey in 2016 found 73 percent of credit unions see analytics as a way to significantly transform their business, Peterson noted.

“I am a little surprised and concerned that number is not 100 percent,” said Tim Peterson, president of CUNA Mutual’s AdvantEdge Analytics. “Using data can grow a credit union’s member base, grow its number of products and services, mitigate risk, and, most importantly, allow the credit union to give its members a great digital experience.”

Peterson said banks are spending $10 billion per year on analytics, but individual credit unions do not have to respond by themselves.

“Why can’t we co-create this ecosystem together? I think we can drive value creation over the next few years. We will be investing $250 million in data analytics over the next few years.”

According to Shazia Manus, chief strategy and business development officer, we are living in the “most extraordinary period in the history of mankind,” due to technology, machines and robotics changing the way we live and consume services.

“What is all the fuss about data? From approximately 2002 to today, global computing capacity has exploded, especially in global connectivity to the Internet. This is driving a digital-first era,” Manus said.

Consumer expectations in a digital-first era are about speed and convenience, as well as receiving a personalized experience, Manus said. She suggested CUs start by gathering data and information from their members: What do they apply for? Where else do they have loans? The second component is deciding what problem the CU wants to solve, a process known as model development. The third step is workflow integration, the fourth is deployment and adoption.

The pair also offered five pitfalls to avoid in the journey to better data analytics.

Pitfall 1: Innovation for innovation’s sake

“Innovation truly is imperative, but credit unions need to know what is hype, what is myth and what is reality,” Manus said. “Identify one business problem, which will lead to a strategy.”

Pitfall 2: Big/bad data vs. smart data

There is so much information flowing in, even with algorithms it is difficult to know what is a “signal” and what is just noise, Manus noted. “Data does not give meaning, it takes humans to be data translators. These do not have to be scientists or mathematicians, they are regular people who decide how to use the data to serve members and run the credit union. Technology is simply a means to an end.”

Pitfall 3: Failing to prep for a surge

Studies have found 90 percent of the data in the world today was created in the last 2 years alone. Manus said big data has three main characteristics: volume, velocity and variance. “In the world of big data, ‘why’ is less important, ‘what’ is more important. Right before a hurricane, people buy flashlights and strawberry Pop-Tarts. There is no why, but stores in hurricane regions know to put strawberry Pop-Tarts in the front of the aisle when a storm is predicted.”

Pitfall 4: Neglecting governance from Day One

Before using data to serve members, a CU must establish a data governance component, Manus advised. This includes taking care of data security and making sure the data is clean. “When someone opens an account, the member service rep needs to put the person’s cell phone in the correct field and get permission for the credit union to contact the new member, otherwise there are punitive penalties. With great power comes great responsibility. There is a great deal of data coming in for members, so we need to be respectful.”

Pitfall 5: Culture fail

An analytics culture starts at the top with the board and the CEO, said Manus. “Any technology is very exciting, but it is important to always make it about the members. You have to build and nurture a nimble, iterative culture.”

For reprint and licensing requests for this article, click here.
Analytics Data mining Data management Data strategy Big data Data and information management