Customer data mining built up by Northern Tool |
 |
By Hannah Smalltree, News Writer
12 Jul 2006 | SearchDataManagement.com |
|
|
It's a cross-channel world out there for retailers like Northern Tool and Equipment Co.
The Burnsville, Minn.-based tool retailer knew it could grow sales with a better understanding of customer behavior across its retail, online and direct mail channels -- a practice often called customer data mining or customer intelligence. The marketing department wanted to better segment customers, bolster internal records with third-party data, make better promotional decisions and predict outcomes based on past events. But, to implement advanced analytics and marketing automation, Northern Tool needed to bring its marketing database in house, away from the third-party provider where it had been outsourced. The new database also needed to integrate data from other internal and external sources. IT had a simple request, recalled Todd Wermerson, vice president of IT.
"It cannot be an IT project," Wermerson said. "So, marketing laid out what they had to have and we guided them on what technology we felt would best accomplish those needs."
Marketing and IT worked together for over a year to define requirements, pick the technology tools and manage the migration and integration of more than 10 years of marketing and transactional history. Co-project managers from marketing and IT led the project, Wermerson said. IT specifically allocated people from its group, dedicating 100% of their time to the effort. Marketing owned the process of defining requirements -- no small feat, according to Wade Mattson, director of marketing at Northern Tool. The 25-year-old company wanted to interact with customers in a whole new way.
Historically a catalog company, Northern Tool now has Web sites serving the U.S., U.K. and Canada, international sales channels, a growing chain of retail stores and over 10 million business and consumer customers. The amount of customer data is growing exponentially, and it didn't make sense to have some of the most valuable data in an external, outsourced database, Mattson said. Overall, the marketing department wanted to better segment customers to make better promotional decisions. But with a project this big, setting priorities was important.
Direct mail is still a major channel for the company, so marketing wanted to select better mailing lists, especially for the annual mailing of a 552-page catalog to 4 million customers in August, Mattson said. The first priority was making more informed mailing list selections and developing response models to predict the outcomes, Mattson said. Longer-term, the company wanted to bolster its customer intelligence, by analyzing behavior across channels, integrating external data sources and eventually using predictive analytics. In the first phase of the project, the team added external data sources of demographic information. Eventually, the company plans to add other data sources, like weather conditions, to better target marketing, Mattson explained. Using location-specific information in this way is a part of a new trend toward location intelligence. For example, a customer in a region experiencing storms might get a well-timed offer for generators or emergency equipment.
With short and long-term goals from marketing in hand, IT reviewed marketing automation vendors and narrowed the list to two providers. Both marketing and IT participated in the evaluation and collaborated on the final decision.
Based on the project requirements and Northern Tool's past experience with Cary, N.C.-based business intelligence vendor SAS Institute, the group selected Base SAS, Enterprise Guide and Marketing Automation. The team liked the interface and the related BI products they could add on in the future, Mattson said. IT also liked the ease of integrating the tool into their infrastructure.
IT designed and created a new IBM DB2 database and installed and tested the SAS tools. The team migrated all of the customer data back from the third-party provider and wrote custom code to integrate other data sources, including 10 years of history from a JD Edwards order processing system and external demographic data. For data quality, the company is relying on a combination of customer service representatives verifying information via phone and batch cleansing from a third-party provider, mostly focused on postal-service-friendly mailing addresses, Mattson said.
Over the last six months, the project team tested the database, trying basic queries and campaign selection activities against it, and training marketing people on the interface. There were no major surprises and the project stayed on schedule, said Wermerson. After more than a year of preparation, the new tools are officially going live this month.
Northern Tool anticipates that better customer segmentation will quickly result in more effective direct mailings, Mattson said. He expect that sales from this channel will grow just under 5% in the first year, a notable gain from a technology implementation.
"We are a relatively mature catalog company and incremental gains of that sort are few and far between," Mattson said.
|
|
|
|
