This article is Part-1 of a 2-part series on planning kanban inventory reductions.
Part-1 covers estimating inventory reductions.
Most manufacturers set goals for areas such as revenue, inventory performance and on-time delivery. Holding and managing inventory can cost tons of time, money and other resources so a common goal many manufacturers have for their kanban initiatives is reducing on-hand inventory.
Unfortunately, most manufacturers have a poor goal setting process, consisting of guesstimation akin to throwing darts in the dark. They may have a rough idea about where they’d like performance to be, but they don’t use a formal mathematical model or scientific framework to inform their goal setting process.
One of the most common mistakes manufacturers make is setting site or organizational goals first, and then working down form there. Unfortunately, there’s often no evidence to suggest whether the organizational goal is even achievable until further analysis of the various areas has been conducted. Often, the reputations and egos of bosses are already at stake by that point, sabotaging the effective resolution of the initial misguided goal-setting process.
It’s far better to first set goals by supplier, value stream, or work-cell and then aggregate those combined goals to arrive at an overall goal that is much more realistic. Accordingly, within the target areas (value-stream, work-cell, etc.), start by setting goals by part-number, then by ABC class, product family, etc. By starting at the bottom and working up, you achieve a much more accurate and achievable goal. Fortunately, kanban data makes it easy to isolate information at various levels based on category and class.
Let’s get started.
Phase 1: Estimate Inventory Reduction
Inventory Reduction Benchmark
Every inventory plan needs a benchmark to evaluate inventory reduction efforts. While, many inventory managers assume that inventory reduction should be measured against maximum estimated inventory, that is wrong. The proper benchmark is the sum of all items’ estimated average inventory. This is because total on-hand inventory dollars should be in the ballpark of the sum of each item’s expected average inventory.
The Central Limit Theorem tells us why this is true. It states that the average of a large population of variables can be estimated by summing the average of the individual variables. Inventory levels grow and shrink all the time and at different rates for each part. On any given day, you can have a handful of parts that are at their max and others that are near stock out. An individual part’s inventory levels can fluctuate drastically over a period of time, but across a large population of parts (i.e. 1,000+), the total on-hand inventory dollars will hover around the same level, and that level is the sum of each part’s estimated average on-hand dollars.
In fact, we can use deviation to drive improvement inquiries. For example, anytime the two values differ by more than 5 to 10 percent it’s often a good idea to find the root cause of the difference (e.g. supplier compliance, demand shifts, broken Kanban processes, etc.).
Estimate Inventory Reduction per Item
For each inventory item, calculate predicted inventory change by finding the difference between estimated average on-hand dollars and current on-hand dollars. By subtracting estimated on-hand dollars from current on-hand dollars you will be compliant with conventional operational and finance reporting where positive numbers always represent a directionally desired outcome. In this case, a positive number will present the amount of the reduction. You don’t necessarily need to follow this convention, but you should be aware of it. Use whatever works best for your organization.
Estimated Inventory Reduction = Current On-hand Dollars – Estimated Average On-hand Dollars
When analyzing any individual part’s current level at a given moment, you are likely to see that the part should either increase or decrease by a certain amount to get closer to the expected outcome. Remember that fluctuations are normal, but there tends to be balance across all inventory items. That said, if an item is suffering from stock-outs, its levels are too low. Similarly, if an item’s levels are above the estimated maximum on-hand balance, then levels are too high.
Estimate Inventory Reduction Total
The population (or total inventory) reduction potential, assuming all kanban solutions are immediately put in place, should also be calculated:
Total Potential Inventory Reduction = SUM (Estimated On-Hand Dollars – Actual On-Hand Dollars) for all items.
This formula accounts for each individual item at whatever level of inventory it is currently at. When kanban is deployed, it will begin improving those levels, starting with those items that are at or near stockout, requiring immediate replenishment. The result is that items that had low levels will now have higher levels, while items with high levels will be ignored.
This is vital to understand, because it means that properly deployed kanban naturally increases inventory levels before it reduces them.
Not only that, but the calculated potential inventory reduction value is not necessarily what your goal for reduction should be. While that is the number that is ideal, it serves more as a signal of the proper direction. It’s more of a target rather than a destination.
It’s good to know there’s a few reasons that explain why ideal perfection isn’t achievable:
- Perfect and immediate kanban deployment isn’t a realistic expectation. That would be akin to flipping a switch and achieving instant results. There are simply too many uncontrollable factors that can and will delay rollout of kanban.
- Most sites choose a soft start. Instead of hitting the ground full speed with complete kanban deployment, a site might choose to begin the process with a small selection of those part numbers likely to have the largest impact on reducing inventory. If this is the track your facility plans to follow, then, for planning purposes, it’s good to assume while the initial kanban items will achieve nice reduction results, delayed kanban items won’t achieve any reduction at all.
- Items that are not on kanban are not likely to achieve an inventory reduction, if their replenishment method is not changed.
- Suppliers may let you down. This can be understandable. After all, there’s always a learning curve to any new system. Internal or external suppliers might act as barriers to success rather than as facilitators. During Kanban deployment, keep a close eye on your suppliers. If supplier compliance is holding you back from improved performance, you will need to address that situation.
That does it for Phase 1 planning kanban inventory reduction. In part-2 of this series we dig into Phase-2 of the Kanban Inventory Reduction planning process; planning supplier negotiations.