A little while ago, I was giving a talk on predictive analytics, and a young professional in the audience interrupted me: “What exactly is a data scientist, anyway?” he asked. Given the number of people who suddenly perked up to hear the answer, I realized he wasn’t the only one in the dark. If you’re wondering, too, allow me to shine a little light science on the topic. Let’s start with data science.
Data science is the process of extracting meaningful patterns from large sets of data. These days, data science has proven important to all businesses, including credit unions, because it employs reliable methods of analyzing data, discovering trends, and identifying business insights.
To be sure, many analysts are involved in similar activities, but a data scientist is, naturally, deeper into the data. In many ways, the data scientist spans the gap between IT and the business. For most analysts, the data has already been prepared for them, whereas data scientists discover data in the systems themselves, extracting it, transforming it, and programming connections. While working with the data is part of data scientists’ activities, they also take that exploratory analysis one step further by codifying and predicting outcomes of the information they are studying. They are constantly searching for opportunities to build repeatable, technical assets that we call predictive models.
Predictive models are where the greatest business value lies for the credit union. It’s one thing to use data to understand what’s been happening in the business, but it is quite another to use data to make accurate predictions of future outcomes. Once these models are created, the information can be delivered to the business areas so that they can take such actions as building marketing campaigns or creating specific business initiatives.
For example, you might want to predict which members are likely to churn from your portfolio and identify probable causes for leaving. Predictive models allow you to do that. What’s more, they can help you actively put the appropriate programs in place to keep those members. They also provide the means to measure the success of the programs.
Want to learn more? Read our whitepaper: Data and Analytics Toolkit: Practical Success Factors for Your Data Management Solution.
Chris Nicholas is Sr. Vice President of Product & Solution Development for AdvantEdge Analytics.