Consider this situation: Company X has many different successful product lines in a particular industry. It is introducing a new product which will take 8-10 months to get from pulp to shelf.
What factors determine its initial supply chain? What happens if the product line is a huge success/failure? What are the fundamental factors that it should incorporate in its supply chain to absorb, to a decent extent, such extreme positions?
In this paper, the authors focus on how to build an apt forecast and inventory management for P&G for a hypothetical new cosmetic line-up. They then test their model on an historical data of existing product lines and come up with fairly successful and interesting results.
The primary thing for inventory management for such new end-customer based products is the existence of a good informational flow within the DDSN (Demand-driven Supply Network). There should be a very good process in place for generation and transmission of real-time demand info between the company, its external suppliers and customers (both upstream and downstream) and also within the company itself.
The next important thing is to have a good forecast. But how do you have a forecast for a product lacking historical sales data? According to the insightful paper,[i] you need to use analogous product sales’ data, iterate for each significant change (this may however lead to problems discussed in the last paragraph) and make appropriate dynamic changes to minimize cost. Although the product can be in high demand or low demand depending on audience reception, the fact is that this demand is tractable for new products with low supply. You could easily invest aggressively in marketing, promotion and expenses to get the optimal number of wall stock units to get the demand curve to come closer to resembling the following (in its introduction stage):
Even if the product turns out to be a terrible idea, you have the optimal number of products on your wall to get rid of (depending on how quickly your forecasts track demand changes which, in turn, depends on information sharing in your DDSN). If the product is successful though, you can simply build on the robust supply chain you’ve built and scale it up. Although because of the long lead times it seems almost inevitable that all new products would need the initial leap of faith, if you’re not tracking your demand and adjusting production appropriately in the 6th or 8th week, there are some serious problems with your forecasting or DDSN.
But, then, if you create a hyper responsive system (to get to good demand forecasts in 2nd or 3rd week instead of 6th), how would you deal with the associated bullwhip effect? And since we’re at the bullwhip effect and all in the same SCM course, when do we get together to play the beer distribution game?
[i] Cheung, Christine. “A Short-range Forecasting and Inventory Strategy for New Product Launches”. Web. Accessed 9/25/2012. <http://dspace.mit.edu/bitstream/handle/1721.1/34844/63199379.pdf?sequence=1>
[ii] Image Source: http://www.quickmba.com/marketing/product/lifecycle/
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.