The last several years have brought an explosion of both data and the raw processing power to analyze it. Big data tools and cloud computing mean banks and credit unions can do a lot more with analytics, but only if they become truly data-driven organizations with an analytics culture.
Without that commitment and investment, there is potential for poor decision making, as well as good, said Glen Sewell, Global Practice Director – Analytics, for Datawatch Angoss, during a webinar hosted by The Financial Brand. It’s important to note, he said, that along with the many benefits of increased reliance on data, also come new risks and challenges.
Banks and credit unions are looking for three things in their data strategy, said Sewell:
- Fast access to data
- The ability to make better business decisions
- Putting data at the center of everything they do.
There is a good reason why financial institutions are pursuing these goals: “Analytics is the foundation on which everything else is built,” observed Jim Marous, Co-Publisher of The Financial Brand and Owner and Publisher of The Digital Banking Report. “If you don’t get analytics right, you’re not going to get the customer journey right, open APIs right, or advanced innovation right.”
Right now, said Marous, who participated in the webinar, a major gap among financial institutions between knowing the importance of customer data and success in applying that data. “Banks and credit unions must go from using data analytics to produce great internal reports that give then better understanding of the consumer, to applying those analytics so the consumer knows we know about them.”
Improved use of data can improve the customer journey and deliver a real-time personalized experience, Sewell added. In addition, the benefits can result in cost efficiencies, reduced time to market, increased loyalty, and increased revenues for financial institutions.
Understand the 5 Barriers to Better Use of Data
Sewell outlined reasons why the current data path being followed by many banks and credit unions is not the correct one. “It’s important to understand the barriers up front and learn how to deal with them,” he said, “in order to design a successful plan to monetize data.”
1. Swamped by hype and misinformation. The data analytics landscape is very crowded right now on the vendor side, according to Sewell. While that provides many choices, it can also be confusing and overwhelming. Data analytics needs to be easy to use by many parts of a financial institution. “Instead, what users are faced with are highly problematic toolsets that do not address the variety of personas across the organization,” said Sewell.
“We have this Wild West approach to data right now with no standardization within the institution”
— Glen Sewell, Datawatch Angoss
2. Insufficient data governance. “We have this Wild West approach to data right now, with different individuals taking their own approaches to extracting value,” Sewell stated. There’s no standardization across the data ecosystem within many financial institutions, which points to the need for improved governance. Without that, institutions often end up with a “data science project or a PowerPoint presentation.”
3. Lagging efficiency. For all the advancements in technology, data preparation is still taking more than 80% of a data analyst’s time, Sewell noted. There is economic impact from poor decision making if your data is not good, of course, but the process needs to be more efficient.
4. Need for increased focus on data protection. The emergence of regulations such as Europe’s GDPR (General Data Protection Regulation) has put the spotlight on data protection. Going forward, financial institutions must “treat data as an intrinsic asset and not just a happy byproduct from business processes,” said Michael Rowley, Director of Global Product Marketing for Datawatch Angoss, who also participated in the webinar.
5. Coping with data explosion. Banks and credit unions have seen an exponential explosion in the types of data they have to deal with, according to Sewell, but are falling behind as they absorb it at a linear rate.
How to Overcome Data Barriers
To help banks and credit unions overcome the barriers to effective use of data, Sewell and Rowley offered several suggestions for different ways to think about analytics.
Don’t view analytics models and algorithms as the end goal. What banks and credit unions should care about is the ability to act on the data. In addition, they should insist on consistency of outcomes from their data. This requires producing analytics that everyone can understand and apply. “The data ecosystem needs a platform to provide confidence and consistency,” Rowley advised, “so that users can create consistent repeatable results.”
Take the long-term view. A data-driven culture must infiltrate everything a financial institution does. Sewell posed this example: “If we’re building a new application, we need to think about how and at what point are we going to capture the data and what is the downstream consumption going to look like?”
Build a marketplace. To cope with the big increase in types of data and data volumes, an internal data marketplace would enable bank and credit union users to quickly find the right data assets for the questions they’re trying to answer.
Keep a close eye on governance, security and privacy. “Those three dimensions will require a larger share of financial institutions’ time and energy to manage and control. One individual should not be responsible for that,” Sewell counseled. To successfully become data driven, the entire enterprise must be responsible for those three dimensions.
Be sure understanding comes along with increased access. Because data scientists are in short supply, said Sewell, some financial institutions want to have more employees move from being business analysts to “citizen data scientists” — individuals who have some understanding of data analytics or basic statistics. As a wider range of employees have access to data, banks and credit unions must be sure the employees have sufficient experience and training to properly interpret the data.