Introduction
Sport Obermeyer, Ltd is an American sportswear manufacturer based in Aspen, Colorado. Founded by a creative German emigrant, it is a successful provider of various skiing apparels for all ages. Being successful for over 40 years, Obermeyer had thorough experience in providing high quality clothing in the middle-higher price range. In order to keep its competitive strengths, outsourcing to the Far East started around the early nineties to lower production costs. The product portfolio was differentiated into several self-made customer categories, keeping excellent price/value ratio and functionality well in focus.
As competition kept rising rapidly, Obermeyer put even more emphasis on its costing strategy, which on the other hand brought tradeoffs. Production in Hong Kong and recently also in China required significant minimum orders, which at the end of the season could have resulted in excess, unsold products. On the other hand, underestimating demand was similarly painful: not being able to fulfill customer requests resulted in lost profits.
All of Obermeyer’s production were coordinated by Obersport
Ltd, a joint venture between Obermeyer and his Chinese business partner. This company organized subcontractors from mainland China and Hong Kong to produce Obermeyer clothing. Recently, a new plant was built in China to utilize its significantly lower production costs compared to Hong Kong. The tradeoff, however was that minimum order quantities were higher in China. (Hammond & Raman, 1994)
Production Overview
Obermeyer’s management had to pass its order decision significantly before actual season start. Very often, they possessed only limited information on trends and current fashion, which resulted in, sometimes, faulty production forecasts. In 1992, CEO Wally Obermeyer decided to base the company’s orders for the upcoming season on a board consensus. (Hammond & Raman, 1994)
Product Order Cycle
The activities and operations of the new products for the next season start well in advance at Sport Obermeyer. About 18 months before the retail-selling period between September and January, Obermeyer starts with the design of new skiwear. Finished products are shipped from China and Hong Kong to the USA for distribution to retail- and larger department stores.
Design activities
Obermeyer’s design team starts with the development of first design concepts by getting their inspiration from outdoor wear shows in Europe and a major trade show for ski equipment in Las Vegas. Based on these concepts first prototypes are produced. The redesign and improvements of the prototypes lead to the finalized design.
Production planning
After the design phase, Obermeyer has to determine the production quantity of its new products. The production order quantity is divided in two parts, the first half is ordered shortly after the design phase, the second half after the next trade show in Las Vegas (when about 80% of retailers’ initial orders have been received). The order quantities are based on forecasts of retailer demand. These forecasts are made by a team of experts at Obermeyer, and are approved by the management. During the time after the second production order, there are additional orders from retailers, which, later on during the retail selling period, go over in replenishment orders.
Procurement
Obersport, as described above, is the main manufacturer of Sport Obermeyer. Obersport is also responsible for the procurement of raw materials, like fabrics and further required components. After Obermeyer has placed the first production order, Obersport has to determine and order the required amount of fabrics and other components for production. Additionally, Obersport places then the dyeing and printing instructions to its sub suppliers. The same process is run after the second production order of Obermeyer. All different supplier and subcontractor relations of Obersport and its production are shown in Figure 1 below.
Production
The production of prototypes, based on the design concepts of Obermeyer, is the first step of production at Obersport for the new design. After finalizing the design, the sample production follows which uses actual already actual fabrics, in contrast to the prototypes. Cutting and sewing of the first production order starts after the ordered fabrics and components have been arrived at Obersport. This full-scale production lasts until the retailing period starts. Before shipping, Obersport also monitors the production, and carries out quality control for their finished products, as well for the semi-finished products of subcontractors.

Shipment and distribution
The transportation by ship of the garments from Obersport’s Hong Kong warehouse to Seattle, Washington starts three months before the selling period. From Seattle, goods are trucked to Obermeyer’s distribution center in Denver, Colorado. Goods produced shortly before the selling period and those produced in China are air shipped directly to Denver in order to ensure timely delivery to retailers, since they request full delivery before start of the season.
Shortly before the selling period, Obermeyer ships orders to retailers via small package carriers. In the middle of the selling period further requested replenishment orders are also shipped to the retailers. At the end of the selling period, replenishment orders of retailers are offered at discount. Remaining garments after the season are normally sold for prices below manufacturing costs to markets in South America, in order to liquidate inventory.

Production Periods
Figure 2 below gives an overview of Obermeyer’s current (early 1990s) production periods, including all steps as detailed in the previous chapter.

Supply Chain Redesign Needs
Since Sport Obermeyer is selling products with relatively short lifecycles, the company has to design new products with changes in style, color and fabric each year. In the last years, this process with all the necessary operations and activities from design to distribution developed and was getting more and more difficult to manage. One reason for the change was the growing sales volume of Obermeyer until the limit of production capacity of the manufacturer was reached. In order to fulfill all the retail demand, above all during the peaks of demand, Obermeyer had to set production orders much earlier based on more uncertain estimations and forecasts of the retailer demand. Additionally, the company was faced with increasing competition which made it necessary to reduce costs in production and to differentiate their products from competition by an increased variety in product features. This led to a more and more complex supply chain with supplier and subcontractors all over the world but with improved production and cost efficiency. Nonetheless, the increased complexity caused also longer lead times. Finally, the expansion of their portfolio expanded the customer target group as well. This caused the retailers to offer an earlier and longer selling period and consequently Obermeyer to deliver the products earlier.
Today, Obermeyer has to deal with long lead times, composing of design, procurement, production and distribution processes. However, these long lead times make it far more difficult to forecast the retailer demand with satisfying accuracy what is related with a higher risk in production and inventory. Furthermore, long lead times and complex supply networks cut flexibility of ordering and limit responsiveness. Therefore, the reduction of lead times and in general the necessity for shorter operational processes are the main needs for a redesign of the supply chain of Obermeyer. In addition to that, it is always important to minimize the costs generated in all processes along the supply chain in order to offer products at competitive prices.

Identification of Redesign Options
Analyzing the Obermeyer’s supply chain operations there are several options of redesigning to cover the needs for more flexibility and responsiveness as well as the reduction of costs.
Cost and time optimized distribution of production
Obersport is running two manufacturing plants - one in Hong Kong and one in China where the cutting and sewing steps are carried out with the finished products at the end. Both factories differ significantly concerning production costs, weekly output, worker qualification and special constraints. Therefore, it is important to define the right distribution of production rates between the two factories. It should be aligned to the current needs, e.g. flexibility in ordering, cost efficiency or production and inventory risks. The production distribution between these two plants is addressed in 5.3.3.
Increase the production capacity
Since the fashion skiwear business concentrates on a short selling period with an additional peak in sales shortly before the season start, there should be enough production capacity available. The increase in sales volume at Obermeyer in the last years has forced them to spread and prolong the production more and more resulting in longer lead times. By simply adding new production capacity, eventually by spreading the production among more manufacturers this effect could be reversed and make the lead times shorter.
Ordering in advance – pre-position
After the production order, the full-scale production can start after the raw materials and other components have been delivered. Since there is hardly any possibility to shorten delivery time of raw materials, it would be possible to order a certain amount of greige fabric already in advance and keep them in warehouse in order to become more flexible with the start of production. This would also distribute the overall production quantity over a longer period and reduce the required quantities during the peak of production.
Furthermore, greige fabrics could be prepared as far as the products are still undifferentiated enough to be able to use them all. (Hammond & Raman, 1994) Reserving capacity
Similar to ordering in advance is the reservation of production capacity an option to redesign the supply chain operations. Thereby raw materials and production capacity could be kept in an undifferentiated state as long as possible, thus allowing Obermeyer to determine the exact design and quantity later with a higher certainty. To compensate the manufacturer for the capacity reservation Obermeyer could assume risk of the supply of raw materials. (Fisher, Hammond, Raman, & Obermeyer, 1994) Shorten and simplify the production and procurement order process
The usage of new technologies offers the possibility to facilitate and shorten the ordering processes in production planning and procurement. Management information systems would provide helpful tools for Obermeyer to improve and adjust the planning.
Shorten shipment times
The time for shipment of finished goods from Obersport to Obermeyer could be reduced by using more airfreight. However, this also relates to higher transportation costs of additional $5. Therefore, the decision depends on the urgency, price elasticity of demand and priorities.
Request for earlier retail order
Obermeyer could try to achieve retailer orders earlier. This would provide important information for forecasting with less uncertainty. One possibility would be to involve main retailers in the stages of development and design and provide them with a preview for the new design.

Reduce the variety of products features
As mentioned before, Obermeyer’s supply chain has become more and more complex in recent years and is responsible for the long lead times. The reasons were the need for cost minimizing and the possible variety of product features. A promising measure today could be to evaluate the benefits of such a variety in regard how far the customer values special features. Thus, it would be possible to reduce the amount and variety of components needed for the garments and the complexity and lead time in the supply chain. Additionally, Obermeyer could focus on the usage of as little as possible different raw materials or colors, also dependent on how far the customer appreciates this variety. (Fisher, Hammond, Raman, & Obermeyer, 1994)
Production Planning for 1993-1994
Wally Obermeyer needs actual numbers for the season starting a year later to make the decision how many to order from each apparel model. Collecting his decision makers, he obtained their individual suggestions for every product. These data are summarized in Table 1 below.
Parkas | Wholesale price | Mean | Std. dev | 2 x Std. dev |
Gail | $110 | 1,017 | 194 | 388 |
Isis | $99 | 1,042 | 323 | 646 |
Entice | $80 | 1,358 | 248 | 496 |
Assault | $90 | 2,525 | 340 | 680 |
Teri | $123 | 1,100 |
381 | 762 |
Electra | $173 | 2,150 | 404 | 808 |
Stephanie | $133 | 1,113 | 524 | 1,048 |
Seduced | $73 | 4,017 | 556 | 1,112 |
Anita | $93 | 3,296 | 1,047 | 2,094 |
Daphne | $148 | 2,383 | 697 | 1,394 |
Total | | ~20,000 | | |

Simple Approach
The available data from employee forecasts can be analyzed further. We can determine the optimal ordering quantities using the Newsvendor Model. In this model, the so called “overage” and “underage costs are introduced. These costs refer to the expenses a company faces by having respectively “one more” or “one less” unit product. Overproduction results in unnecessary spending, whereas underproduction causes lost profits. We are aware of Obermeyer’s overage and underage costs (respectively 8% and 24% of wholesale price); therefore, we can also determine their optimal service level (SL) using the formula underage cost (UC)underage cost+overage cost. The optimal order quantity can also be estimated using the inverse of a normal distribution with service level, mean and standard deviation (or its double) as inputs. Since the plants in Hong Kong and China have different minimum order quantities, it is important to keep in mind, that certain products cannot be produced in China, which has 1200 units as minimum (vs. Hong Kong at 600). Obermeyer also produces in two rounds, one before the Las Vegas show, and one afterwards, thus it is the anticipated total quantities’ halves that determine production location. Details can be seen in Table 2 below, where HK and Ch refer to Hong Kong and China respectively. At this point, differences in production costs are not considered.
Parkas | Price | Overage | Underage | SL | Opt.qty. | Half qty. | Order | Loc. |
Gail | $110 | 8.8 | 26.4 | 75% | 1,148 | 574 | 600 | HK |
Isis | $99 | 7.92 | 23.76 | 75% | 1,260 | 630 | 630 | HK |
Entice | $80 | 6.4 | 19.2 | 75% | 1,525 | 763 | 763 | HK |
Assault | $90 | 7.2 | 21.6 | 75% | 2,754 | 1,377 | 1,377 | Ch |
Teri | $123 | 9.84 | 29.52 | 75% | 1,357 | 678 | 678 | HK |
Electra | $173 | 13.84 | 41.52 | 75% | 2,422 | 1,211 | 1,211 | Ch |
Stephanie | $133 | 10.64 | 31.92 | 75% | 1,466 | 733 | 733 | HK |
Seduced | $73 | 5.84 | 17.52 | 75% | 4,392 | 2,196 | 2,196 | Ch |
Anita | $93 | 7.44 | 22.32 | 75% | 4,002 | 2,001 | 2,001 | Ch |
Daphne | $148 | 11.84 | 35.52 | 75% | 2,853 | 1,427 | 1,427 | Ch |
Total | | | | | 23,180 | 11,590 | 11,616 | |
Table 2 - Newsvendor model for production distribution

Similar Risk Model
Given the capacity constraints wherein two production runs (one before the show and the other one after it) are required to meet the demand, it can be easily seen that the problem of underage cost no longer remains a big problem as the necessary adjustments could be made in the second production cycle. Thus, in such a case given the capacity constraints we would like to reduce the risk for each parka while satisfying the constraint that the total production of parkas shouldn’t exceed 10000.
Also assuming the demand for parkas to be normal any point x on a normal distribution curve can be represented as x = mean – k*(standard deviation). Given this, we can find the x value for each parka with the added constraint that the sum of parkas should add up to 10,000. Using the Goal seek function in excel we can find k which can then be used to find the production amount of each parka. Please see the results below in Table 3.
Parkas | Mean | Std.dev | 2xStd.dev | Half qty. | Actual prod. | Location |
Gail | 1,017 | 194 | 388 | 605 | 605 | HK |
Isis | 1,042 | 323 | 646 | 357 | 600 | HK |
Entice | 1,358 | 248 | 496 | 832 | 600 | HK |
Assault | 2,525 | 340 | 680 | 1,804 | 1,803 | Ch |
Teri | 1,100 | 381 | 762 | 292 | 600 | HK |
Electra | 2,150 | 404 | 808 | 1,293 | 1,292 | Ch |
Stephanie | 1,113 | 524 | 1,048 | 1 | 0 | N/A |
Seduced | 4,017 | 556 | 1,112 | 2,837 | 2,837 | Ch |
Anita | 3,296 | 1,047 | 2,094 | 1,075 | 1,200 | Ch |
Daphne | 2,383 | 697 | 1,394 | 904 | 1,200 | Ch |
Total | ~20,000 | | | 10,000 | 10,737 | |

Methodology
In order to compare the two models i.e. the simple model and similar risk model we simulated the demand. This was done by using the rand() function in Excel to generate a random number and using the fact that the demand follows a normal distribution a sample demand was obtained for each of the 10 parkas. We then wrote a macro in order to simulate various demand scenarios of each parka by running the simulation 300 times.
The output of the simulation was the cost savings of the similar risk model as compared to the simple model. The cost savings were found out in terms of overage/underage costs, production costs and finally total costs, which include both overage and underage costs. The results of each of these simulations are presented below.
Comparison of over-stocking and under-stocking cost
Based on the simulation the results obtained are shown below. As in most cases the similar risk model turns out to be better than the simple model with mean savings amounting to $24,974.3. It should however be noted that the standard deviation is also large: $16,351.4.
Mean | 24,974.3 |
Standard Error | 950.4055 |
Median | 28,971.79 |
Standard Deviation | 16,351.39 |
Sample Variance | 2.67E+08 |
Kurtosis | 0.811219 |
Skewness | -1.00917 |
Range | 92,451.65 |
Minimum | -36138.8 |
Maximum | 56,312.81 |
Table 4 - Cost savings

Comparison including Production Costs
Cost Item | HK | China |
Material | 30 | 30 |
Labour | 10 | 0.78 |
Transportation within China and China Overhead | 0 | 2 |
HK Quota , Obersport Profit , Overhead | 9.9 | 9.9 |
Cost FOB Obersport | 49.9 | 42.68 |
Agents
fee to Obersport (7%) | 3.493 | 2.9876 |
Freight | 1.4 | 1.4 |
Duty Insurance and Misc | 5.29 | 4.9 |
Total Cost | 60.083 | 51.9676 |

Production costs
In order to make the analysis more complete we decided to include the production facilities in China and Hong Kong into the simulation. This was done by including production costs as shown in the table above. Also, in order to decide to which region a parka should be shipped to we used a simple algorithm wherein:
If number of parkas < 200 then do not produce
Else if 200 ≤ number of parkas < 900 then produce in Hong Kong
Else produce in China
This was done taking into consideration the fact that the minimum lot size for production in China is 1200 and the minimum lot size for Hong Kong is 600.
Mean | 99,815.33 |
Standard Error | 1,559.784 |
Median | 94,048.48 |
Standard Deviation | 26,835.54 |
Sample Variance | 7.2E+08 |
Kurtosis | 0.847203 |
Skewness | 0.895377 |
Range | 14,9672.1 |
Minimum | 43,788.62 |
Maximum | 193,460.7 |

Production cost savings
Simulation of the cost savings in this case showed that the similar risk model always outperformed the simple model with significant cost savings involved with expected savings amounting to $99,815.3.
We concluded our simulation by finally combining the cost aspects just described above and found that on the whole the cost savings by using the similar risk model were significant compared to the simple model. The mean of the cost savings obtained via simulation was to the tune of $124,789.6 as shown below.
Mean | 124,789.6 |
Standard Error | 2,156.911 |
Median | 114,672.7 |
Standard Deviation | 37,108.9 |
Sample Variance | 1.38E+09 |
Kurtosis | 0.276041 |
Skewness | 0.792618 |
Range | 187,950.3 |
Minimum | 56,332.24 |
Maximum | 244,282.5 |
Table 6 - Cost savings

Figure 6 - Total cost savings

Sourcing Options and Conclusion:
While it is more expensive to produce in Hong Kong than in China, Hong Kong production takes less time. The Hong Kong manufacturer also provides higher quality, higher productivity, higher responsiveness, flexibility and quicker production times, but at a higher price.
In Guangdong China bigger batches can be produced cheaper, but production process requires larger order sizes and longer lead times. The worker’s productivity is still lower than in Hong Kong because the worker’s still require additional training. Additionally, products produced in China have to be shipped early because of the U.S. quota restrictions for the number of units from China. Consequently, if the production is not shipped before the other manufacturers, Obermeyer risks that its production might be sent back. Long-term Sourcing
The first lot should be a large batches order and be placed approximately 10 months before the selling season in China. This order will consist of the quantities produced from the data attained through the similar risk model and includes the models: Assault, Electra, Seduced, Anita and Daphne.
Short-term Sourcing
The second lot should be ordered approximately six months before the retail season. The quantities in this order should be decided upon the feedback after the Las Vegas fashion show because forecasts based on the Las Vegas show offer more precise forecast of trends and colors of the coming season. This order should be produced by Hong Kong manufacturers considering that they are faster, more efficient and more flexible, in particular, if the batches are smaller. However, for batches above 900 it is cheaper to order in China. The lead times can be reduced by air shipping. Although the price per parka would increase to $56.92, it would still be cheaper than the Hong Kong production.

Conclusion
Sport Obermeyer should make a majority of its long-term sourcing order in China. Taking into account that strict quota restrictions limit the number of units that can be imported from China into the United States, it is essential to ensure that the orders are brought into the country before the quota has been used up. Obermeyer’s short-term orders should be manufactured in Hong Kong to avoid the risk of rejected shipments.