Bullwhip Effect: What is it and how companies try to avoid it?

The main concept of supply chain management is simple: customers buy products from you; you try to keep track about demand of your product in the market; and you manage inventory by producing enough items to cover your customers' demand; producing means you order enough materials from your suppliers.

It sounds simple, but still it makes forecasting demand as a difficult job. Wrong forecasting can lead to out of stock or excessive inventory. The problem is about coordination. Suppliers, manufactures, sales division and customers have their own estimation about demand. It is not match between each other. Further, each group makes decision based on others. This situation refers to as Bullwhip Effect.

Prof. Hau Lee from Stanford illustrated this Bullwhip Effect by story when Volvo found itself with extra inventories of green cars. In effort to minimize inventory, Sales Department made special discount offers, so demand increased. Production department, unaware about the promotion, saw pattern of increased demand and made decision to produce more green cars.

Coordination is not only the problem. Delays on manufacturing and procurement also contribute to bullwhip effect. To explain about it, there is simulation game called “The Near Beer Game” that I found from internet. I recommend trying this simulation game to get better understanding about bullwhip effect.

http://forio.com/simulate/mbean/near-beer-game/run/ 

After I tried this simulation game, It obviously clear that even with no breakdown in communication and knowing exact information about demand, I still get Bullwhip effect from procurement and manufacturing delays. The result that I have got is either out of stock or excessive inventory over period of time.

So how does company try to reduce the Bullwhip Effect?

Article from www.chon.com gives 5 steps to reduce the Bullwhip effect.

Step 1: Improve better information and communication along the supply chain in effort to forecast demand.

Step 2: Reduce or eliminate delays in supply chain. Cutting order and delivery time will decrease fluctuation on inventories.

Step 3: Use your point of sale to analyze customer preference and behavior.

Step 4: Reduce order batching and create smaller order increment to avoid demand fluctuations.

Step 5: Maintain stable price for products. As we already know, when price goes down, the demand will increase and create bullwhip effect.

Right now, companies all around the world   try to optimize their supply chain management by using sophisticated software. For examples: Seven Eleven Japan has its cashiers record the sex and estimated age of each customer so the company can predict its customers' behavior. These data have been used for its data mining process and give result for a convenient way to set out its shelves. Another example is Coca Cola with its Minutes Maid and Simply Orange brands. For Coca Cola, Producing orange juice in consistent quality requires complicated process. Coca Cola uses computerized system called Black Book. This system uses algorithm to produce information how to blend batches of orange fruits to replicate certain taste and consistency. Another function of Black Book is also to incorporate external factors such as weather patterns, expected crop yields and cost pressures to help company plans about supplies for 15 months ahead. 

From these examples we can see how serious companies with its supply chain management. They understand that out of stock inventories will means their customers will buy their competitor products and excessive inventories will become a burden in their balance sheet. So it can't be more or less. Wrong formula will make them loose business. Isn't it becoming a bit like rocket science?What is your opinion?

Reference:

http://www.economist.com/node/949105?Story_id=949105

http://forio.com/resources/article/bullwhips-and-beer/

http://smallbusiness.chron.com/reduce-bullwhip-effect-3908.html

http://www.businessweek.com/articles/2013-01-31/coke-engineers-its-orange-juice-with-an-algorithm