Are banking providers building shiny new data capabilities but using them to answer yesterday’s questions?
That’s one of the big questions haunting data analytics experts in the banking industry who participated in a panel of both financial institutions and solutions providers to discuss trends in data analytics.
Panel participants said many data initiatives in banking fail because senior-level executives don’t involve themselves sufficiently in the organization’s transition. This reality has, in turn, impacted how much institutions are willing to invest in data capabilities. However, panelists say the answer isn’t to stop investing in data analytics, but rather to change expectations and to better educate people about the new capabilities.
“Many banks invested in cool new enterprise database technology to bring all their data together to get a complete profile of the consumer,” said Poornima Ramaswamy, AI & Analytics Practice VP at Cognizant. “But in the last 24 months, I’ve seen a lot of these programs stop.”
The reason for the pullback, she explained, is waste. While there is the possibility of finding “gold” in the volumes of data within a financial institution, Ramaswamy said there are “a lot of other ‘minerals’ as well — not all of them usable.” She said a sizeable investment in people, management, training, and culture is needed to find and use what gold there is.
Key Digital Imperatives for Gen Z and Millennials
As your financial institution prepares for the impending Great Wealth Transfer, ensuring that your digital strategy appeals to younger generations is essential.
Read More about Key Digital Imperatives for Gen Z and Millennials
A Day in the Life With Instant Payments
Financial institutions can leverage the FedNow® Service to build instant payment solutions that consumers and businesses are demanding. See real-life use cases in this video.
Like Gold, Relevant Data is Rare
Picking up on the precious metal theme, panel moderator Steve Cocheo, now Executive Editor of The Financial Brand, observed that banking executives often believe that financial institutions can mine the vast troves of consumer and transaction data they have on file and quickly yield priceless, actionable insights.
That is a false notion — the panelists all agreed on this point.
One participant, Ashish Bansal, found the “gold” analogy particularly appropriate. Bansal has not only served as senior director of data science at Capital One, but also had prior experience at a mining company.
“The amount of actual gold contained in raw ore is typically about 0.01%,” he said. “So if you mine 1,000 kilograms of gold ore, you yield about a gram of actual gold. I think that’s true for data as well. Having a lot of data does not mean it’s relevant.”
Bansal, who is now a senior artificial intelligence leader at Twitter, also pointed out that consumer behaviors change over time. If you have ten years of transaction data, only the most recent three have any real relevance, he said. “The seven years before that are useless, because society moves on.”
Equally blunt was Allison Sagraves, Chief Data Officer of M&T Bank.
“To me, this idea of a data ‘gold rush’ where everybody is going find all this hidden value isn’t realistic,” Sagraves said. In her view, extracting useful data and figuring out how best to use it is an iterative process.
“Gold is not the only valuable metal in the world. When I first started buying jewelry, I bought silver — it was cheaper and I could buy bigger pieces,” she explained. Sagraves suggested doing the same with data: start small, with “silver,” then work up to “gold.”
Sagraves described how M&T built a 360 degree view of its customers using older technology they were already familiar with.
“For the first time we were able to see relationships together that we weren’t able to see before,” Sagraves recalled. “That provided tremendous value to our business lines. I would call that step ‘silver’.”
Read More:
- The Art And Science Of Data Analytics In Financial Services
- 8 Steps To Make Data Analytics Part of Your Bank’s Core DNA
- Machine Learning, AI And The Future Of Data Analytics In Banking
Everyone Must Be Immersed in Data
“Everybody’s data capability needs to rise — business managers included. It cannot just be left to the data analytics function.”
— Ashish Bansal, former Senior Director of Data Science at Capital One
The panel agreed that leaving major data analytics initiatives just to the data pros was a big mistake. They advocated for shared responsibility, distributed across a broad swath of each institution’s senior leadership team.
“Data analytics professionals can become so absorbed in their own world that they lose sight of the fact that they are trying to solve a business problem,” Bansal observed.
Chris Wheat, Director of Business Research at JPMorgan Chase Institute, an in-house think tank, stated things a bit more directly: “There are only business questions. I’d be concerned about any company where the data function didn’t have a clear business purpose.”
That said, Wheat added that data analysts do sometimes see things in the data that might uncover latent business issues or hidden opportunities.
Sagraves says M&T Bank is still in the early stages of its data analytics journey, according to, and so data work is primarily driven by business use-cases. She and her team meet with people on the business side to discuss the kind of data that is now available and how it could help solve specific business problems. She acknowledges that more work and training is needed, but she’s all for “trying a few things and seeing what we learn.”
Concerning the need for more training, Bansal said the data capability of everybody needs to rise — business managers included. “It cannot just be left to the data analytics function,” he explained. “Analysts’ viewpoints are very different from those of a business leader.”
Bansal advised users of data to roll up their sleeves.
“I would urge all business leaders to be more directly involved in their data rather than relying on somebody to interpret it for them.”
Data Skills Must Match New Tools
Sagraves said she has been in the CDO role for three years. The first two she devoted to setting up the foundational requirements to be sure M&T, a regional bank, was dealing with quality data. Now she’s working on helping the bank make the transition to being data insight-driven, a process she described as a “significant culture shift.”
“The skills required to be data and insight driven are not necessarily the same skills we needed in the previous era,” said Sagraves.
Regulatory requirements drove much of financial institutions’ initial focus on data governance and data management, pointed out Ramaswamy — compliance, not creativity. That was followed by the introduction of data dashboards, which, she said, became “prettier” and easier to use over time, but failed to move the needle toward banks and credit unions becoming more data driven.
“We now recognize that it’s not even just about being data driven,” said Ramaswamy. “It’s about being a data-insight-driven organization — making informed decisions and taking the right action.”
Easier said than done.
Referring to banking companies in general, Sagraves observed, “We’re building new data capabilities, but are we still asking the old questions? If so, how do we start to change the conversation and ask different questions of our data? What new problems can we solve for our customers to make their experience better?”
Asking such questions is simple enough, said Sagraves. However, “getting the business units to think differently about data and to understand what the data makes possible — that’s a seismic change.”
Read More: Banking’s Digital Talent Crisis: Who Will Fill the Tech Void?
Finding the Talent to Get It Right is Tricky
Even giant JPMorgan Chase — 19 times larger than M&T Bank — faces tough challenges as it becomes more data driven.
“It’s hard to find people who have the technology background to think about what data might tell you and who also have business sense,” said Wheat.
Wheat said the megabank tends to grow such talent, rather than acquire it. The expertise, he said, is essential to avoid, on the one hand, “chasing some really interesting technology project that has no business case,” or, on the other hand, being unable to frame a true business need to get the right data solution.
Wheat adds this caveat for all financial institutions working to become data driven:
“When you’re a financial service provider, it’s really, really important to get it right when you’re dealing with customer data. Some of the cool and interesting things you might want to implement out of curiosity have to be balanced against safeguarding peoples’ financial lives.”