Page 365 - Orlicky's Material Requirements Planning
P. 365

344                                                 PART 3      Managing with the MRP System


        SKU/customer level. Simplification by looking at the overall context (e.g., region, brand,
        channel, technology type, etc.) of the data that reflect major drivers of change in the mar-
        ketplace and aligning the application of roughly right assumptions rather than precisely
        wrong tinkering are fundamental before trying to model even a few options.
             Finally, any strategy that calls for fewer and faster innovations and shortened prod-
        uct life cycles will mean that more items will fall into the planning horizon, even though
        they have not been invented yet, let alone have an assigned product code!
             The level of detail involved in planning should change as the time horizon changes
        (Figure 20-9). In the short term, businesses need accurate transactional data on supply,
        sales, stocks, and so on and will be focusing resources on collaboration with customers
        and the deployment of advanced supply-chain optimization tools. In the medium to long
        term, the roughly right, not precisely wrong level of detail is required. Aggregating data
        using brand families, average revenues by technology type, and so on with a focus on the
        development of scenarios using business planning/simulation tools becomes important.
        This can be achieved only if finance has a strong leadership role in implementation.


        UNCERTAINTY VERSUS A SINGLE SET OF NUMBERS
        Although there has always been a recognition that forecasts are either wrong or lucky, the
        early S&OP movement followed a premise that with immense pressure and focus the
        quest for a single set of numbers could remove uncertainty. In the early years, that pres-
        sure and focus often created some benefit by putting a basic rigor in place in sales and
        marketing and raising the profile of the need for improved forecast accuracy for opera-
        tions. For many, though, the operational ownership and quest for a single set of numbers



           FIGURE 20-9
           Roughly right
           versus exactly
           wrong.








                                                                      “Roughly Right”

                                                                      “Not Precisely Wrong”
                                                   Detail


                                  History  Short                                2–3 Years
                                          Term
   360   361   362   363   364   365   366   367   368   369   370