# L.L Bean Forecasting

Topics: Inventory, Forecasting, Leon Leonwood Bean Pages: 5 (1802 words) Published: September 16, 2012
1. Inventory decisions at L. L Bean use statistical processes on the frozen forecasts provided by the product managers. L. L Bean uses past forecast errors as a basis of measurement for future forecast errors. The decision for stock involves two processes. Firstly, the historical forecast errors are computed. This involves taking the ratio of actual demand to forecast demand. The frequency distribution of historical errors is then compiled across items, for new and never out items separately, to form a probability distribution. The probability distribution is then used to predict errors for the future. The second step involves calculating the contribution margin if the unit is demanded and the loss if the unit has to be calculated. This is done to calculate the critical fractile for the demand which can be calculated by Gain/ (Gain+ Loss). The critical fractile is used to calculate the optimal level of stock by taking losses and gains from adding an extra unit to stock. This gives inventory managers at L. L. Bean the optimal order size which is the critical fractile of the item’s probability distribution of demand. The same fractile is applied to the distribution of the forecast errors calculated in the first step. The resultant error ratio is multiplied with the forecast demand to give a specific number that L L Bean commits with its vendors. Taking the information given in the case into account the ratio would be calculated as follows: Gain= 30-15=15 Loss=15-10=5 Therefore the ratio equals= 5/(5+15) = 0.75. So given this case the company should keep the additional item of inventory, only if 0.75 is greater than the probability that the item won’t be needed. 2.The costs and revenues primarily used by L. L. Bean to make decisions include the cost of the item that L. L. Bean pays to its vendors, the selling price of the item and the liquidation cost. The selling price and cost are used to calculate the gain from selling each item. The cost minus the liquidation gives the loss per unit. These values are further used to calculate the critical fractile for demand and forecast errors. However, there are certain costs and revenues that L. L. Bean largely ignores when calculating the critical fractile. This could include the cost of overstocking before the product is liquated thus increasing the loss in that situation. In case there are stock outs L. L. Bean may experience lost demand and possible reductions in goodwill. Additionally when stock outs occur L. L. Bean may be able to cover up with backorders. Such backorders are likely to reduce the profit earned from selling the product. L. L. Bean should consider all these costs in deciding how many units of item to stock 3. Forecasting for a “new” item in L. L. Bean bears resemblance to the Obermeyer method. A group of people including Scott Sklar rank individual items in the catalogue and assign dollar sales value. Information required by Scott Sklar will include some level of sales and demand information which the inventory managers are likely to have. It is therefore that Scott Sklar includes four or five inventory managers when ranking different items from the catalogue. Additionally Scott Sklar assigns dollar value to each item in the catalogue. He must have the cost data before he can assign dollar values to each product. Scott Sklar then uses spreadsheets to compare individual item forecasts to forecasts for the entire catalogue. Therefore he must be provided with book forecast by product managers to change forecasts of individual items to comply with the book forecasts. There are certain aspects of forecasting that Scott Sklar largely ignores. There seems to be no interaction with customers when forecasts are being made. Forecasts could be made accurate if customers were included in ranking the items and applying dollar values to individual items in the catalogue. Primary market research could greatly assist Scott Sklar and his team to make forecasts more...