The fact that optimization programs base their
calculations on ‘hard data’, such as your usage history,
makes the approach particularly appealing. With this
kind of solid input the results must be right - right?
There are two problems with the ‘data only’ approach.
First, historical data, no matter how accurate and
‘clean’ it is, only tells us what ‘has been’. When it
comes to your inventory investment you are more
interested in what ‘could be’. Second, the data more
often reflects the behaviours of your team rather than
the actual demand for your inventory. Who among us can
say that our team members don’t take some ‘just in case’
items that distort the usage data?
As a result you are forced to make assumptions
(sometimes implicitly) about the characteristics of both
demand and supply for your parts. These assumptions are:
that what happened in the past will happen in the
future, and that you cannot change these outcomes or the
behaviours influencing them. This approach forces you to
work within constraints that may or may not be real.
And this is why optimization doesn’t truly optimize, it
only recalculates within a set of assumed constraints,
it doesn’t challenge those constraints.
About The Author
Phillip Slater is an
Engineering Materials and Spares Management Specialist
and the developer of Inventory
Optimization™. Phillip is the author of several
Inventory Solutions, A
New Strategy for
Continuous Improvement, and The Optimization
Trap. For more information visit his website at