In 2009, the company launched an analytics project called Voyager, which uses Microsoft's SQL Server database alongside analysis tools from CA Technologies and SAP BusinessObjects to segment its credit union customers by variables such as product, profitability, and demographics. The first step was consolidating customer data from sales and marketing systems, as well as systems used by its credit unions, into a single data warehouse. Then business analysts explored the data with canned reports and iterative queries on the fly.
"They were able to complete in a couple hours using multiple queries an analysis that before would have taken one to two weeks to complete," Roy says. "It's game-shifting change."
Business analysts were startled to find that half of CUNA Mutual's $2.8 billion in revenue comes from 3 of its 12 customer segments. Now the company wants to build financial products to attract the other 9. To appeal to Generation Y consumers, for example, CUNA Mutual is developing more Web and mobile access to products it offers to its credit unions. The company also built software that automatically offers life or disability insurance to people after they take out a loan, over whatever channel the customers used to close the deal -- phone, website, or in-person.
The new thinking has CUNA Mutual moving away from doing 3 or 4 large marketing pushes per year to 12 smaller ones focused on producing specific results: selling a particular product to new customers in a given demographic, for example, or gaining more profit from an existing customer segment. "You run the marketing, test it and use what you learn to start new campaigns," Roy says.
New ways to save
While CUNA Mutual looks outward, at customer dynamics, Welch's, the grape juice and jelly cooperative, uses analytics to make internal operations, namely transportation, more efficient. During a 2007 upgrade of Oracle's enterprise resource planning suite, Welch's saw it needed newer tools to enable more flexible queries using multiple dimensions of data, says Kevin Kilcoyne, director of customer operations at the family farmer-owned co-op. The organization does manufacturing and marketing for the National Grape Cooperative Association.
Welch's wanted to collect every data element from each year's 40,000 orders and bills of lading, sweep it into a database and look for patterns to highlight where it could save money on transportation. Welch's auctions its transportation business to trucking companies every year.
At the time, Oracle's reporting tools couldn't perform the in-depth analysis Welch's wanted quickly enough, Kilcoyne says. To prepare for the auction, it took about 30 hours to cull a year's worth of data and then a few months for analysts to study it to decide how to formulate each bid request. Analysts consider routes, kinds of transportation available, and fuel pricing trends, along with carriers' limitations and performance statistics.
Because so much time was involved, Welch's was not always able to bid out all of its distribution routes; typically it bid out only about 60 percent of them each year. The balance went largely unanalyzed and, therefore, unoptimized, says Bill Coyne, director of strategic sourcing.