The Decision Maker

Why Analytics Requires Standard Data Sets

Posted by Nate Wentzlaff on Mar 30, 2015 1:43:00 PM

With systems designed to capture a robust amount of data on every transaction that occurs throughout the credit union, standard data sets are required to form strong analytics.


Credit union leaders must develop a data-driven vision.  The ideals cast in this vision give clear direction to employees.  That being said, without standard data sets to monitor business processes, credit unions will fall into a worse condition than before establishing an analytics program.  Employees will bring data to meetings but will have different versions of the truth.  Standard data sets are essential for a solid analytics foundation.

Business Processes

Before establishing an analytical data model (ADM), credit unions must engage in activities to support its mission and business processes.  Analyzing the current state of its business processes is essential.  Any business can be broken down into processes and sub-processes.   In order to establish standard data sets, a credit union’s processes must be mapped, modeled and monitored.

Data Population

With advances in data management technology, business processes throughout the credit union can be monitored in detail (at the transaction level).  Whether it is originating a mortgage, opening a checking account or using a mobile device to complete a transaction, credit unions have the data on every transaction.  Many credit unions consider this data as a liability and ignore its true value.  It is important for credit unions to recognize their data as an asset and use it to improve business processes.

Data Sets

As each business process is analyzed in detail, business analysts must work with managers to establish standard data sets that will be used to analyze each business process.  Data sets consist of transactions (rows) and attributes (columns).  Clearly defining and communicating each dataset to all employees involved with a business process is crucial for establishing an effective analytics program.

Data Set Management

The word manage is an active verb.  Unfortunately, many managers see their roles as reactionary.  In order to effectively manage data that is populated from the processes under their watch, managers must proactively monitor and maintain standard data sets.  This is a difficult task because credit union managers are stuck in departmental thinking.  They are under the impression that they should manage people.  This is a grave mistake.  People cannot be managed; they must be lead.  Managers must focus on managing data and leading employees to their full potential through analytics.  Identifying opportunities for front-line employees (and therefore the entire organization) is the role of a manager.

Continuous Improvement

Data is the raw material of the credit union industryIn order to fully reap the value of data, credit unions must adjust their standard data sets carefully with approval from a data management steering committee.  Changes to how data is analyzed and shared have dramatic consequences.   Altering the definition of one data attribute can have long-term effects.



Data Set Evolution

No credit union will ever be perfect; however, in order to succeed and build a prosperous future for their members, credit unions must focus on improving business processes.  Analyzing the standard data sets of the credit union will allow managers to adjust the business processes accordingly.  Results from these changes can be monitored and used for future decisions.  For example, if a loan origination data set contained an attribute of approved vs. denied based on business logic (e.g. credit score), changes in the logic that define the attribute will result in changes to the final decision.  Decisions have dramatic impacts on the future of the credit union.  The potential of standard data sets is tremendous.  Now, it is up to credit union leaders to establish the data-driven culture and navigate toward a promising future.  This can be achieved through building and publishing standard data sets.

Stay Up To Date - Subscribe To Our Blog

Topics: Data Integration, Analytic Data Model, Data Standards