There were two chief driving issues behind our analysis of this Sport Obermeyer instance: the measuring and apprehension of demand from unsure and disparate prognosiss. and the allotment of production between mills in Hong Kong and Mainland China ( Lo Village. Guangdong ) . The chief challenges confronting the company were long lead times. small to no feedback from the market before the first production determination ( the first existent demand signal is at the Las Vegas trade show in March ) and inaccurate prognosiss along with the lost net incomes that can ensue.
The first portion of our analysis involved deducing an order policy from the prognosiss provided in the sample job. We solved this job utilizing simplifying premises and so loosen uping some of those premises. Our initial premise was that there was no minimal order measure. We decided that hazard would be minimized by bring forthing the smallest allowable sum during the first production run due to the deficiency of information. Therefore. we calibrated our order measure expression to sum to 10. 000 units.
We wanted to utilize a expression that took into history the mean prognosis every bit good as the standard divergence – in other words. we wanted to account for both the expected demand and the uncertainness. We began with the expression Q = Average Forecast – 2* Standard Deviation of the prognosiss. since twice the standard divergence was said to come close the standard divergence of the existent gross revenues. Since this figure did non sum to 10. 000. we multiplied the standard divergence by a grading factor. k. and solved for order measure 10. 000 units across all designs. We found k = 1. 0607. which gives a measure of 10. 000 with no minimal order measure.
Following. we had to modify this order policy because designs Stephanie. Isis. and Teri had initial orders below the minimal order ( for Hong Kong ) of 600. Since we could non compare productively analyses without utilizing the sweeping monetary values ( which we were explicitly told to disregard in the job ) . we “rounded up or down” depending on whether the order measure without the lower limit was 2/3rds of the lower limit or non and adjusted our K value so that the measures summed to 10. 000 units. Therefore. we removed those three and calibrated once more. and found that K = . 9675. So we are exposed to somewhat higher hazard due to the other designs holding to be ordered in larger measures. but no design is being ordered in a measure above 75 % of the mean prognosis or higher than the lowest commission estimation. and so hazard is still comparatively low. We trust our commission to be at least reasonably accurate in their anticipations. particularly since Wally took stairss to garner many separate estimations from a panel experts in a mode that prevented groupthink. The appendix shows the ensuing orders.
Finally. when sing telling from China alternatively of Hong Kong. the higher minimal order of 1200 units becomes a major obstruction. As the appendix shows. Gail can non now be ordered. and Daphne and Entice must be increased to 1200 to run into the minimal order. When factoring these new restraints in. we find thousand = 0. 9345. The prognosis is now much more hazardous ; for Entice. the order measure is now greater than the minimal estimation. in add-on to being really near to the mean prognosis. All orders that are non constrained are now significantly closer than one standard divergence from the mean. increasing the opportunities that we will over order. We are forced to see the uncertainness less in our anticipations and are more controlled merely by the figure of units that we think we need.
Although we did non loosen up the premise that all monetary values were the same. loosen uping this premise would hold made this job into a complex optimisation job and the optimum minimal order measures would hold been somewhat different. We compared our consequences with the quantitative analysis of this instance done by an Industrial Engineering professor at Georgia Tech utilizing constrained optimisation and lagrangian multipliers and found that our consequences were rather similar.
The hazard for these orders can be measured utilizing statistics based on the expected standard divergence of demand. Since the expected criterion divergence is twice the standard divergence of the prognosis samples. we can happen the proportion of the clip that the existent value of demand will be within a certain scope. Because of our K value. our orders are all around 1 standard divergence below the mean. significance that approximately 12 % of the existent observations will fall below these orders. In that instance. Sport Obermeyer would lose money based on the excess units. If demand falls less than 600 units above the initial order. we have a similar job in the 2nd ordination phase – Obermeyer can non efficaciously set its order size to suit the reevaluated demand. Based on the cost of each point. which we ignored in our analysis. we could cipher the expected over-ordering from each point and happen the ideal degree. similar to the outside analysis that we found.
The 2nd chief issue to see is whether to beginning in Hong Kong or China. Sourcing in Hong Kong provides Wally more flexibleness because its minimal order measure is 600 units compared to 1. 200 units in China. The ability to order fewer units allows Wally to order units in a less hazardous mode. In either instance Wally orders combinations of merchandises so that he minimizes the opportunity he’ll order excessively much of any one merchandise. Lower minimal order measures allow Wally to order lower sums of more merchandises. Since he believes he will sell many of each merchandise. adequate to cover the minimal order cost. Wally can take down his order amount’s distance from its expected value ( cut downing the opportunity he’ll order excessively much ) .
This difference in minimal batch sizes can non be attributed to mechanical advantages because the mills in the two countries use similar equipment ( as described in the instance ) . The workers in Hong Kong are better trained than those in China. and in peculiar they are better cross-trained. This most likely explains why a typical production line in Hong Kong can hold fewer workers and therefore produces fewer units per twenty-four hours. However. the mean Hong Kong worker is more productive ( more units produced per twenty-four hours ) but the mean Chinese production line produces more units because it contains more workers. This suggests why the lower limit batch size is greater in the mainland mill.
Choosing to run in China is much cheaper. China labour costs are $ 0. 78 per unit and Hong Kong’s are $ 9. 69. which is more than a 12 fold difference. Labor costs are about $ 9 less per unit than in Hong Kong. Wally has to make up one’s mind if he wants to pay $ 9 more per unit to hold the ability to cut down hazard in lower lower limit orders. If Wally loses 8 % of the monetary value of a merchandise when he can’t sell it in the primary market. so he loses less than $ 9 for 5 of the merchandises. Alternatively of losing $ 13 or $ 12 on the most expensive merchandises. he now merely loses $ 4 or $ 3. However. waiting for the trade show to garner information creates a two measure telling procedure. This creates more of a barrier in China. If Wally under-orders on certain designs ( which he certainly will because it is portion of his telling scheme ) he can merely purchase more units during the 2nd rhythm in groups of 1200. If he revises his prognosis based on demand and decides he needs to order 500 more units of Entice. he is now faced with a pick: he can order 1200 units and have 700 likely unsold. or he can lose out on 500 units of possible gross revenues. Note that in Hong Kong. this hazard is much less desperate since he can order in 600 unit blocks.
Ordering in China significantly increases the hazard of telling excessively much of a merchandise but it besides reduces the loss due to that mistake. Wally should beginning those designs which have the highest expected demand to China because it allows him to take fullest advantage of the cheaper labour. He should be careful. though. to order at least 1200 less than he thinks he will necessitate. and so merely those designs with an expected demand of 2400 or more can be ordered in this manner. On the other manus. designs which are expected to hold lower demand give Wally a tough determination: he can order 0 units. as we recommended with Stephanie. and seek to estimate demand after the trade show to find whether the cheaper labour costs will do up for holding to order 1200 units ; he can order more than 1200 units in the first order rhythm from China. in a instance like Entice. and hope that his estimation is accurate. since he will non be able to travel back once more for 1200 units ; or he can order from Hong Kong with greater flexibleness. Wally must make up one’s mind where to beginning. Sourcing in China would cover losingss by diminishing to amount lost per unit and sourcing to Hong Kong covers losingss by diminishing the sum of units produced beyond demand in the primary markets.
In the short-run. Wally must see this trade-off between the lower labour cost of the mainland site and the lower lower limit batch size of the Hong Kong mill and make up one’s mind how to optimally beginning production to maximise profitableness. In the long-run. there are three elements that could diminish the lower limit batch size at the Chinese mill: worker accomplishment and cross-training. mill layout ( assembly line versus a different theoretical account of production ) . and the nature of the cutting equipment.
We feel that wide operational alterations can be really good to Sport Obermeyer. both in the manner that it orders its merchandises and besides about the concern theoretical account it uses to react to consumer demand. Firstly. Obermeyer should be able to cut down on its merchandise lead clip. which gives it more clip to analyse informations and react. by carrying natural stuffs alternatively of telling entirely based on finished merchandises. For illustration. slide fasteners. catchs. and insularity can be standardized across many designs. presumptively without important value lost to the client. and stored for the 2nd ordering rhythm and so applied to whichever design ends up being successful. A typical production rhythm takes about 9 months to complete- from the trade show in Las Vegas to the beginning of the ski season. and cutting down on 60 or 90 twenty-four hours lead times on slide fasteners might let Sport Obermeyer to custom-make their refilling retentions to do certain the popular points can be replenished in the shops.
The 2nd operational alteration Obermeyer could implement is to contract their merchandise line and alter their push-pull scheme. Currently. Sport Obermeyer is seeking to react to the demand pulls. by doing a broad assortment of windbreaker to pull every purchaser. They could cut a major part of their costs. though. if they offered merely one or two windbreaker for each of their four “genders. ” There are many grounds why this could be a successful scheme. Obermeyer’s chief rival. Columbia. marks lower-income skiers. and so the danger of popular Columbia designs stealing Obermeyer clients is low. In add-on. two of the genders. Rex and Klausie. are to a great extent impacted in their purchasing determinations by position. and fewer designs would do those designs more recognizable. increasing the position associated with them. Besides. it would do designs from the old season stand out more. and therefore increase the obsolescence of the merchandise. promoting new purchases.
Another gender. Biege. attentions purely about proficient public presentation. therefore a standardised windbreaker with the latest engineering. possibly offered in different colourss. would be plenty to maintain him purchasing Obermeyer. However. the biggest benefit to contracting the offerings would be to cut costs. All of the designs could be ordered in much larger measures from China alternatively of Hong Kong. taking advantage of the cheaper labour. Since demand for each gender would non hold to be estimated and split between many designs. it could be tracked over clip and a much more accurate prognosis could be made. Besides. the job of running out of popular manners for refilling would be mitigated. since a much larger part of the stored windbreaker would be of any given type.
D. Simchi-Levi. P. Kaminsky. E. Simchi-Levi. Designing and Pull offing the Supply Chain Concepts. Strategies. and Case Studies. McGraw-Hill. Boston. 2000