At the outset we need to distinguish between “maximize”, “minimize” and “optimize”.
“Optimize” deals with “the best possible compromise between opposing tendencies”, whereas maximize and minimize merely deals with seeking the most or least of a tendency. (It may, of course require optimization to achieve this most or least of a tendency.)
Therefore, when is optimization not really optimization?
Achieving the selected target level of customer service (availability) is not optimization. It is merely investment accuracy. Optimally setting that target service level is optimization.
Merely setting the stock levels to achieve some pre-determined service level can hardly be called optimization. In this instance, what is being optimized? Probably not much! It is assumed that the service level chosen has been chosen through a process of optimization (but, in our experience, this is rarely the case. It is more commonly arrived at through more intuitive processes, i.e. experience, judgement, emotions, competitor actions, reaction to disaster, etc.).
Furthermore, true optimization requires the freedom to optimize. Only then does an optimization discussion become meaningful. The greater the freedom to optimize the more meaningful the discussion becomes.
Optimization programs should therefore also:
Create the freedom to optimize (the subject of a White Paper).
Measure this freedom.
Select the “best possible compromise” among the alternatives within this freedom.
Optimization is about mutually exclusive co-relationships and choices amongst the paired or linked co-relating variables. In the inventory context it relates to the linked variables of the (non-linear) cost of stock investment on the one hand, and (non-linear) product availability benefits on the other hand.
It is the best possible compromise between inventory holding costs and availability benefits that is sought. The task of examining these trade-offs and then finally selecting the optimum compromise between these variables is the true nature of optimization. This will require critical inputs about market behavior, supply chain risk, capital costs, and intelligence – a 4-dimensional task, in fact.
In this context, optimization can be expected to address bottom-line results, i.e. marketing and financial results that are the outcomes of these choices. We can now try to evaluate the alternative inventory performance optimization tools out there. To be an optimization tool they must at least be able to compute and trade-off the non-linear benefits / penalties from changes in availability with the non-linear holding cost of inventories.
Other than that offered by our ICS tool, we are not yet aware that any such optimization tools are available in the market today. The topic of optimization freedom is further discussed in our White Paper titled: “Financial Gridlock in Inventory Performance”.
Commenti