You are here

Rol Fessenden

Rol Fessenden

BA Mathematics
SUNY, Binghamton

MA Geology, SUNY, Binghamton

Director of Inventory Control
L.L. Bean

Have you or anyone in your family ever placed an order with L.L. Bean? You might be surprised at how difficult it is to ensure that we have your specific item, color, and size when you order it.

You may remember in 1988, on Sunday before the Republican primary in New Hampshire, that the pollsters predicted Bob Dole would win. Two days later, George Bush won instead, and the rest is history.

At L.L. Bean we have thousands of items, and tens of thousands of colors and sizes. Nine months in advance, we have to predict for each one, how many our customers will want. If professional pollsters can be wrong about the New Hampshire primary two days before it occurs, imagine how wrong we can be nine months in advance on 50,000 predictions.

I graduated in 1969 from the State University of New York at Binghamton with a BA in Mathematics, and in 1974 I received an MA in Geology. Before coming to Bean, I was a geologist, a scientific systems designer, and a small business manager. In all these occupations mathematics played a key role. I developed models for surface runoff, models of heat transfer in complex, high temperature environments, and statistical profiles of accounts receivable. I have used sophisticated statistical procedures to gain insight into glacial landforms.

My current interests are in customer testing and in mathematical modeling of the merchandising process. In mathematical terms, there are clear analogies between inventory management and previous areas of my expertise such as heat transfer. The insights I have as a result of previous work have proven invaluable at Bean. Mathematics has provided me a common framework for understanding concepts in many fields.

At Bean we attack the forecasting problem at every stage of the process. We analyze historical sales data, model forecast uncertainty to design contingency plans, and analyze catalogue displays to understand how some can be more successful or attractive than others.

We also survey customers to find out their preferences among the products displayed. On some products, the customer feedback is an accurate indicator of sales, but on others it is not. We have to understand when the questionnaire is accurate, and when it is not. Finally, we analyze the first orders received from newly mailed catalogues to refine our predictions.

Next time you order Blucher Moccasins in size 6 narrow, think of me.