Decoding the DNA of Customer Profitability
This month we will feature ground-breaking work in finding attributes that predict customer profitability. This project was done for a Supply Velocity Client (products shown below) with the support of a Ph.D Supply Chain Management study at the University of Missouri St. Louis.
The Ph.D program has been a great way to learn new skills to help you, our Clients. This case-study will demonstrate these skills.
We provided both a powerpoint overview and the detailed report that was produced by this study.
If you are feeling analytical, dig into the report. It was written for our Client’s management team. Therefore, the statistical jargon is kept to a minimum.
If you want an overview, read below and click on the link to a powerpoint summary slides.
Our goal was to use advanced statistics to predict profitable customers so our Client could more cost-effectively use their direct-mail campaigns to grow profit. They print and mail out hundreds of thousands of catalogs each year to existing customers and prospects from mailing lists that they rent. This company distributes outdoor site furnishings, such as those pictured above. Customers include governments, property managers and franchisors.
The management team wanted to test 7 variables. This was a large study with over 50,000 customers in their database. We discovered that 5 of the 7 variables were statistically significant in predicting profitable customers. Some of these variables were somewhat obvious, but a few were big surprises and are impacting their marketing tactics. They are changing the prospects and customers they target through direct mail and assign sales representatives.
- Sunshine % of the state where the customer is located did not affect customer profitability. They were sure that sunnier states yielded greater sales and profit. This is not true.
- Coastal State customers were significantly more profitable than inland customers… a great finding for mailing out catalogs.
- Customers that ordered on the web were more profitable than customers that called in orders. They believed that customers that called in, and spoke with a customer service rep, purchased more and more profitable products. The customer service reps are trained to suggestive-sell. But apparently the power of the web for making it easy for customers to browse and buy is greater. Think of how they can use this information to convert more customers to web-purchasers, increase sales, reduce costs and become more profitable!
(Disclaimer: my Ph.D Professors would want me to state that all findings for a single variable were based on all other variables held constant.)
This model was then tested and proved robust. The test is described in the paper.