The chart above, shows a hypothetical cost optimization example. In particular, it shows how stockout cost risk is reduced as inventory is added, but inventory costs increase. There are no inventory costs at a zero-inventory level, and stockout costs are reduced to near zero at higher inventory levels.
In the above example, stocking a level of 1 is projected to reduce the number of stockouts by more than half, so the expected stockout costs decrease accordingly, but adding that 1 until of on-the-shelf inventory results in inventory costs – but the reductions in stockout costs is significantly more than the added inventory cost, so an inventory level of 1 is a lower total cost than 0.
Adding another part, to bring the stocking to a level of 2, causes another large decrease in stockout costs, and a roughly linear increase in inventory costs. At this point, the stockout risk becomes very small – so small that any additional inventory added can’t reduce enough additional stockout cost to counter-balance the added inventory expense.
As the lowest total cost option, a stock level of 2 is the optimized stocking position for this part at this location.
What about service level?
Most inventory planning systems target a specified service level goal. In cost optimization, service level is an output of the process at the lowest cost point, rather than the target of the optimization.
The problem with service level goals is that they are often arbitrarily defined. If your company is targeting 95% service level, you need to ask, “Why is 95% the target?” The likely answer is that it was found to be a convenient point where it felt like inventory costs and stockout costs were balanced. It was an arbitrarily set goal that felt right, which is not a true optimization.
In fact, some organizations create complex, overlapping service level goals for categories of parts using ABC classifications and other groupings, which can be very hard to manage effectively and can cause inventory churn. The downsides of using ABC stratification for optimization will be a future article.
Sometimes, due to contractual commitments, a minimum service level goal is necessary, even if it moves the inventory target to a non-cost-optimized state. Targeting a specific service level goal should be viewed as a constraint on an optimized state, rather than the core optimization.
A key benefit of cost optimization is that it allows for easy identification of situations where changing a stocking target from one level to another doesn’t provide a cost justification. For example, if your optimized target moves from 1 to 2, but the optimization result is close enough that the cost benefit of changing the stocking level does not exceed the operational costs of changing the stock level, then the planning system can be intelligent about leaving the stocking target at 1. This biases the system to the current state, reducing churn and further reducing operational costs.
A planning system that only allows its users to identify service level goals is not truly optimized. Optimization is only achieved when the minimum of total operational costs of inventory investment plus stockout costs are at their lowest possible level.
Only a planning system that provides cost optimization can provide this lowest total cost approach to stock levels. This, combined with understanding the operational costs of changing a target stock level, means Prophet by Baxter Planning is your best solution for service parts planning. Ready to learn more about our service parts planning software? Contact us today!