Forecasting in LL Bean's Supply Chain

The assignment content discusses the role of forecasting in the supply chain of a mail order firm like LL Bean. The firm uses the JDA Demand Tool to develop and maintain demand forecasts, which helps reduce excess inventories, improve item availability, and lower costs. The article also explains the difference between static and adaptive forecasting methods, with static methods being less effective when there are significant changes in model parameters over time.

1. What role does forecasting play in the supply chain of a mail order firm such as LL Bean?
Historically, the mail order firm LL Bean has followed the made-to-stock philosophy and
satisfies customer orders from the available inventory. In the past, this has often put the firm
at risk of overstocking.
At present, the JDA Demand Tool is used to develop and maintain demand forecasts at LL
Bean. For items which are well-established in terms of their robust demand, the forecasting
system uses past trends to forecast future demand. There is often seasonality that is observed
and thus enough information is available to forecast how much quantity of the item will be
required and at what point in time. Appropriate adjustments to the system parameters are
done as warranted by any in-season trends in the sales of items. For items which lack
sufficient demand information, the forecasting system allows for computation of the ratio of
actual demand to forecast demand (the A/F ratio) based on whatever limited past data is
available for similar competing items, and computes each such item’s profitability as well as
the costs of overstocking and the costs of understocking, and finally determines the
appropriate quantity of the item that must be made available at a point in time.
The firm’s demand forecasting system is able to get real-time information from all points of
sales as its items are purchased by customers and move from the shelves. This real-time
information on the demand for their items allows them make better item-line forecasts.
With a much better demand forecasting system than before, the firm has been able to reduce
excess inventories for its seasonal items, improve the availability of those items which are
sold at all times during the year, and lower its costs of warehousing of its items.
2. How do static and adaptive forecasting methods differ?
A static forecasting method, as the name suggests, would not see any change in model
parameters, such as trends and seasonality, with any new demand information. Thus, in a
static forecasting model, once the parameters are established, they remain unchanged (or
static) regardless of what the new demand observation happens to be. The parameters need no
adjustment and continue being used for all forecasts going forward. Due to the static nature of
the estimated values of the said parameters, a static forecasting method works well only when
insignificant variances are expected to occur in the model parameters over time. In contrast,
an adaptive forecasting method, as the name suggests, would ensure that the model
parameters, such as trends and seasonality, adapt to new demand information in an
appropriate manner. Thus, any model parameter established earlier would be adjusted to
incorporate the effects of new demand observation. Thus, if demand for a product were to be
impacted in a significant manner because of advent of any disruptive technology, the adaptive
forecasting method would have an immediate appropriate response.
References:
 James A. Cooke | From the Quarter 4 2011 issue. (n.d.). L.L. Bean's smarter stocking
strategy. Retrieved October 01, 2017, from
http://www.supplychainquarterly.com/topics/Strategy/201104llbean/

 LL Bean Customer Success: Partnering to Improve Forecast Accuracy | JDA
Software. (n.d.). Retrieved October 01, 2017, from https://jda.com/knowledge-
center/collateral/ll-bean-customer-testimonial-video
 Chopra, S., & Meindl, P. (2013). Supply chain management: strategy, planning, and
operation. Boston: Pearson.