Data-driven Bluestem makes faster, better decisions with new BI system
Bluestem Brands, Inc. is driven by data. The company owns catalog and online marketers Fingerhut, Gettington.com and PayCheck Direct. Their sweet spot is reaching budget-driven consumers with goods they can purchase via a number of payment options, including credit and payroll deductions. Extending credit to over 95% of their customers, Bluestem is as much a finance company as it is a consumer products retailer.
“We use tools and data to better understand our customers so that we can match them to the right products, and offer them a way to pay that meets their needs. explained Bluestem’s Director of Business Intelligence Jayme McBride.
Bluestem collects data from and about its customers constantly. Shopping history, payment preferences, credit history – all are grist for Bluestem’s huge data mill. Having started long ago using metrics to help in decision-making, Bluestem’s departments each built their own data warehouses. Eventually, the IT department built an umbrella to connect these systems but it lacked the scalability required for business growth, was costly to modify and was prone to data quality issues.
“Besides being fragile, the system was complex and required a significant amount of time for an analyst to extract the data and roll it up into a readable report,” said McBride.
The existing system was inefficient and contained redundant data that led to confusion across business units. Bluestem wanted the ability to have more frequent updates and more integrated data so that all the departments, from Credit to Merchandising to Marketing, could better communicate and make smarter decisions.
Putting a road map in place
Coherent Solutions was one of two suppliers working with Bluestem at the time that McBride was brought on board to lead the creation of a new BI system. McBride quickly determined that they needed just one supplier to handle all the outsourced pieces. He chose Coherent Solutions based on the quality and work ethic of the company’s work force, explaining:
“Coherent’s people – both onshore and off – are bright, engaged and ambitious. They are eager to solve problems. We knew we could rely on them to work through anything.”
Bluestem’s Coherent team consists of three onshore people, including an account manager, a technical project manager and a business analyst, as well as nine offshore developers and testers. Chief Technical Officer, Max Belov, also got involved when he helped to architect the new system.
Bluestem chose to go with Microsoft’s SQL Server for the warehouse and added MicroStrategy’s BI software to allow business users to perform visual analytics and build their own reports, including charts, graphs, and dashboards, without having to be so dependent on IT.
“My team is now able to devote more of their time to new development and functionality, and the business is able to spend more time analyzing and less time extracting data” said McBride. “We aren’t just keeping the lights on.”
Better data, better business decisions
Other benefits of Bluestem’s new BI system are:
- More and faster data updates – prior to the start of the business day as opposed to mid-day.
- Historical tracking of data changes over time so that business users can identify trends in customer behavior
- The ability to easily drill down into the data to look for drivers such as cross-selling, customer demographics, credit levels, etc.
- The ability to more easily assess the impacts of change to the overall environment.
McBride’s team also built a metadata repository that could be leveraged by all departments to build a common understanding of where their data comes from, how it’s represented in the warehouse and what the data means. A suite of data models and data mappings, including business definitions, were reviewed and vetted by representatives from multiple departments.
“We haven’t built out the entire system, but at about 75% of the way there we are seeing faster, more accurate results. For instance, we used to measure the effectiveness of a marketing campaign by gathering response rates and manually tying these to sales numbers gathered from raw data. The process was slow and full of potential for error. It’s all automated now,” said McBride.