Why CUs Can't Afford to Wait on Data Analytics (And How to Get Started)

By Aaron Passman03.13.2017

The hardest part of data analytics isn’t the technology—it’s the cultural shift within the organization.

That’s the word from Brewster Knowlton, founder and principal consultant at the Knowlton Group, a business intelligence firm that works with credit unions. Knowlton moderated a panel at the Credit Union Analytics Summit here on how credit unions can begin their “data analytics journey,” and one participant was quick to shoot down any hesitation CUs might have about entering that field.

“To start, you really just start—that sounds simplistic and silly, but I say that all the time now,” advised Clay Yearsley, SVP of data analytics at Texas Trust CU. “You need to commit a resource. Somebody’s got to be in charge of this. It really is too big to be somebody’s part-time job.”

In mid-2015, Yearsley was given the challenge of tackling what he called “this data thing,” and he told attendees that he began by doing a data gap analysis, examining what information and systems Texas Trust had, how it accessed information, whether data was tied to third parties, inventorying data gaps and prioritizing those gaps.

“The question is really the base unit for data analytics,” he explained. “You have to start with a question. I built an excel spreadsheet tracking all of the questions we came up with and then used that to prioritize the impact and effort required to answer those questions.”

And, he added, credit unions already have plenty of data on-hand that they can mine.

“Start using the data you already have,” said Yearsley. “That ACH info is just a goldmine. Start using that. If you’ve got somebody who uses you as their PFI and you’ve got their ACH and card info, you’ve got everything about them.”

Naveen Jain, VP of digital analytics at First Tech FCU, offered lessons from his own experience leading First Tech’s data analytics efforts, which began with focusing on costs and increasing operational efficiencies. Part of the challenge, he said, was finding ways to reduce the workload of some analysts, who were spending as much as 90% of their time bringing together data from multiple sources. Phases two and three, he said, were focused on determining how to use the insights gained from that data, and building a data warehouse and data governance.

“Quick Wins”

Because of the cost and effort involved with launching data analytics, panelists agreed that “quick wins” go a long way toward demonstrating the necessity of data analytics and getting buy-in from the rest of the organization. And, many added, there’s no shame in going after the low-hanging fruit.

“I went to the boss and asked ‘What’s a painful process you do every month?’ and then I automated it,” recalled Yearsley. “That’s a quick win!” He suggested picking out important internal reports and sending out commentary with them, rather than just spreadsheets with data, to better understand what the information means and how it compares to three, six or 12 months prior. “Send it out to more than just the executive management team, because you’ll start getting feedback from other places,” he advised.

But, added Jain, time is of the essence.

“Quick wins have to be in weeks, not quarters,” he said. “If it takes a quarter, it’s not a quick win.”

Harsh Tiwari, SVP and chief data officer at CUNA Mutual Group, noted that financial services firms invest in data analytics as much as any other industry, but individual institutions often don’t use those capabilities to the extent that they could.

“If you do nothing else,” he said, “anchor your investment on how you can eliminate manual work (through automation).”

Harsh Tiwari, SVP and chief data officer at CUNA Mutual Group, and Clay Yearsley, SVP of data analtyics at Texas Trust CU, spoke during the 2017 Credit Union Analytics Summit.

Understanding the problems that an institution wants to solve with data analytics is important, he said, “but equally important is being able to technically figure out what pieces of data are helpful in getting the problem solved. Where might it exist and how do I get to it efficiently?”

Yearsley advised setting up a goal to find a way to impact all levels of the organization through data analytics and finding talent across the credit union that can serve as an ad hoc team.

“You don’t have to spend multi-million dollars or a million dollars,” he said. “You can start in the low six figures and get that going. If [cost] is the thing that’s holding you back, spend $5,000 on a couple of licenses from [a tech vendor], hook it up to the spreadsheets you’ve already got and you can do some pretty amazing analysis. It doesn’t take a lot, it just takes a little bit of commitment.”

This article appeared at www.cujournal.com and is reprinted wit