Demand Forecast


Company Background

Founded in 1998, Chinese Pharmaceuticals (HK) Company Limited is a small, family-owned trading company. Their primarily business is in Chinese herbs, which has inheriting the wisdom of the traditional Chinese medicines. In 2007, they established  first research and development office at science park. Subsequently, Chinese Pharm transform from a trading entity to a developer and marketer of its own brand. Its first well-known and signature product was Noto37, which use for maintain the heart disease. The company's sales in Hong Kong are accounted for 95%, remaining 5% sales are in Macau and US, which for the Chinese community. 


Case Background

Chinese Pharm signature product── Noto37. Nodoginseng is a key ingredient for Noto37, an agricultural product grown in western China. However, the availability of Nodoginseng is fluctuates over the course of the year, as it would affected by unpredictable natural events. Also, there are seasonality of sales of Noto37, 50% sales above the average were generated during the peak season (Winter months and Chinese New Year). As Chinese Pharm has leased warehouse in Hong Kong, they would require additional space during the peak periods, and it would be ideal if the warehouse just maintain at most 3 months supply of stock to avoid excess stock situation. For the reordering point, it would be triggered when the stock level shrinks to 1.5 month’s supply. Other than that, the actual demand and supply are also fluctuated due to different factors. For instance,  the drought condition in Yunnan, retailer company own intermittent promotions, biweekly deliveries of different quantities to almost 1000 retail outlets. Hence, it is crucial for China Pharm to do the demand forecasting for Noto37 to prevent stock out or over-stock situation.



Design of Supply Chain (for Noto37)
Design of Supply Chain (For Noto37)

Overall, we have concluded the problems that Chinese Pharm need to deal with, including:
  • Long lead time
  • Seasonal demand
  • Limited warehouse space
  • Frequent promotional sales
  • Regular monthly reordering interval to their retail customers
  • Equipment-intensive manufacturing process
Chinese Pharm is undoubtedly need a more accurate prediction to prevent the out-of-stock situation. 





Three common types of forecasting model
1. Simple Exponential Smoothing
2. Adoptive Smoothing Model
3. Full Exponential Smoothing Model






Historical Sales Data (Jul 2009 - Jun 2012)



Using Full exponential model to predict sales from Jul 2012 - Jun 2013




Advantages of each type of forecasting model

1. Simple exponential smoothing model

This model is simplistic, which is not sophisticated as it just requires fewer variables and simple formula when comparing to another two methods. Therefore, it is easy for the beginners to learn and apply. Besides, its prediction relatively accurate which  takes the difference between prediction and actual value into consideration. 



2. Adoptive smoothing model

This method allows us to make some change in the historical data, and the following forecasting data will change automatically. It uses beta to control the decay of the impact of shifts in trends, then we can set a larger or smaller beta based on we need short-term or lone-term prediction. 




3. Full exponential smoothing model

This method can generate the most accurate forecast, as the newly added parameter gamma is contributed to measure the effect brought by seasonal factor, by adding the data from the same month last year to consideration. It also explains the differences between forecast and what actually occurred, which makes the prediction more accurate. Moreover, this method can calculates the differences between the observed value and the actual value, so that it is easy for us to find out the pattern of the data. 



Q: Is vertically integrated manufacturing or other presently outsourced operations vital to the company? Why?

Outsourced operations is more vital.


There are many tasks that would take lots of time to finish such as quality check or sourcing the raw material, it almost takes a week to finish. Outsourcing the operations can help the company focus on their core business and also saving their time. Some of the task can rely on experts to finish by outsourced operations, for example expert will be more familar with whether the condition of raw material is good or not. The company can pay attention to the demand forecast to decide the amount of inventory, and develop the sales plan. As company’s product affected by the market trend and season. The company in the industry requires high labor and capital resources to lower the cost as it eliminates the cost of hiring and managing the human force and the cost and risk investing the equipment. For vertically integrated manufaturing, although adopting the vertically integrated manufaturing can have more control over quality and avoid the supply disruption, it is not efficient as it would reduce the company flexibility and make the company lose their focus on their core capabilities. The main ingredients for their main products so far cannot be grown at a satisfactory rate locally. Otherwise, they would have to move their main focus to China which costs a lot of money and risks. Therefore, outsourcing the supply of the Notoginseng can ensure the quality and reliability of continuous supply for raw materials


Reflection

Through this lecture, we have gained a deeper understanding on how significant of adopting different forecasting model in a company, especially for the retailing company, as the demand and supply may fluctuates overtimes due to various dynamic factors. When implementing the demand forecasting, we must consider different factors. Predicting the future demand by using the past sales data is normal practise for most of the company. Without a doubt, we should put more effort on doing the forecasting. If there is no efficient demand forecasting method, the company may occur the excess stock or out-of-stock situation frequently, which is definitely unfavorables for the company development. Although the prediction may not be completely accurate, it have already been highly reduce the uncertainty. Beside, we have discover that there are various types of forecasting methods and models, it is vital for us to identify the most suitable forecasting model for the company, which highly depending on the company business attributes. 



Reference:
Chinese Pharmaceuticals (HK) Company Limited official website:

Chinese Pharm Linkedin:








 

留言