Associate editor Jon Camhi cites an SAP survey that illustrates how banks lag behind other industries, especially retail, in customer analytics. The survey reveals that:
- Only 46% of banks can analyze external data about customers
- Only 32% can analyze social media activity
- Data volume and analytics complexity are the most common challenges cited by respondents
- Cloud and predictive analytics technologies will be "extremely valuable" to around 60% of respondent's strategies in the next 24 months
Some of the experts Camhi spoke with said unstructured data is overwhelming for legacy systems in banks, and even internal data management protocols are often in conflict. The first step for banks today is to get their database structures normalized and standardized, before they introduce new data sources into the mix.
Great Western is a great example of reaching across an organization to create data quality standards. Camhi quotes Ron Van Zanten, the bank's vice president of data quality:
The bank created a data committee with members from different teams across the organization, Van Zanten shares. The committee, which reports up into Great Western's business intelligence operations council, created standard definitions that teams across the organization now use for different tiers, pricing, and terms on accounts. Those definitions are standardized across Great Western's various systems.
Once that process was complete, Great Western began phasing in demographic data from credit tracker Experian to help identify profitable customers.
Fifth Third had issues that any business (and any banking customer) can relate to -- its acquisition of several bank chains created a mishmash of products, rates, and perk programs that various customers were grandfathered into.
Over 2012 and 2013, the bank simplified its product portfolio and began offering relationship-based discounts. It also cleaned up its customer database and partnered with a third-party analytics provider.
One interesting conundrum for Fifth Third and many other banks is the potential for interest rate hikes in the next year or two. Predicting the hikes and their effects is extremely difficult. Camhi writes:
Banks simply don't have relevant data to analyze and understand how money will move... Data from the past few years doesn't show any rise in interest rates. And data from 2004 to 2007, the last time interest rose, isn't usable because customers didn't have modern digital tools such as mobile banking back then.
To address this, Nomis Solutions, the analytics provider working with Fifth Third, patented a methodology that analyzes current consumer behavior and tweaks predictive models in real-time. This way, if a trend begins to appear, the bank can forecast its effects and react quickly.
Members -- do you think these kinds of early-warning models can help banks deal with uncertainty? Do you wish your bank did a better job of customer analytics? Share your thoughts below.
— Michael Steinhart, , Executive Editor, AllAnalytics.com