Big data is a generic term for a collected set of data points that is so large and complex that it can't be processed into meaningful information using traditional database management tools or data processing applications. The collection and analysis of big data is of crucial importance in large scale science experiments like the Large Hadron Collider (which measures some 600 million atomic collisions per second). Big data was also key in mapping the human genome and is the norm for large-scale weather and climate observations. The only way to arrive at solutions or come to conclusions using it is through the effective compilation and analysis of disparate data.
In the private sector, Internet marketing companies like Amazon and eBay track activity and searches in order to present prospective customers with ideas of products they might like to purchase. We laugh at those suggestions at times, as they seem to stem from the large amount of research we do and nothing to do with our actual preferences. Crunching huge amounts of unrelated data can still deliver unrelated results, but it will eventually get better. Over time, predictive marketing efforts that use such data will get ever more accurate and greater accuracy should result in a more successful effort. Certainly, the predictive aspect that can come from this kind of data capture has other uses. For instance, consider if you could predict customer dissatisfaction and attrition. Your bank could act quickly to preserve customer relationships as you cross-sold products and services that drove greater customer satisfaction. Awesome!
It is a given that within every bank there is a large quantity of data that describes customers' characteristics, their activities and preferences. However, even data within the bank usually tends not to be collected in a unified format, and especially if acquisitions have occurred, so comparison is either difficult or exceedingly time consuming.
Most useful is finding middle ground that is a combination of internal data from the bank and external data. In fact, our own relationship profitability model does just that by leveraging bank data and external data like probabilities of default and loss given defaults specific to the location of a loan. In so doing, bank lenders in this example have the ability to know the profitability of the customer's relationship within the bank plus the risk parameters of the market in order to calculate a risk-adjusted price based on real information. This is the same thing the largest banks do when they automatically price using relationship profitability information (although their ability to measure risk, especially in smaller towns or rural areas, is far less robust than a community bank one could argue).
Consider as well the valuable information around a customer's relationships with other banks. If you had this information you could make a specific offer to the customer and perhaps consolidate the relationship by focusing on customer needs. If an institution had the data for every commercial property loan in its market footprint, knew which financial institution currently is holding the loan, the interest rate, structure and maturity date, this would be highly useful information for targeted marketing.
While not everything is public and available, lots of data is accessible, so here again the key is turning it into useful information. Here, banks are often turning to outside providers to help them not only collect the data in a more useful format, but also to analyze the results. Building a body of business intelligence that includes predictive analytics can also give your bank a true competitive advantage. Community banks may not have an in-house analytics team, but the good news is that research-driven organizations that have historically focused on retail and other sectors are beginning to focus on the banking industry, so changes are happening.
All of this adds up to more information about your customers, prospects and your market. That is good as it should help your bank stay competitive. Our advice is not to write off big data with the argument that your bank is too small, but rather use it where you can and where it makes sense. Great information is out there, you just have to decide it's important enough to use it.