Back in 1989, I led what was then called an MPR II implementation project for Thunderbird Formula boats (http://www.formulaboats.com/). We had many objectives that we wanted to reach and one of them was to be able to look at the quantity listed as “On Hand” and trust the number was correct. For too long, our purchasing personnel simply didn’t trust the numbers. Instead, to be certain that they knew the right quantity on hand, they would go find the inventory and make sure the quantities were right.
While this was understandable, it was also exceptionally inefficient. Why not get correct balances and keep them correct instead? Now, you may be thinking, okay, they’ll take a full physical inventory and everything will be correct. Unfortunately, it isn’t that easy. There are a number of problems associated with a physical inventory:
- Often firms use whatever personnel might be available so the person counting may not be familiar with the parts they are counting. This means there can be errors in quantities or part numbers. For example, if the standard unit of measure is pairs and the item is counted as “each” the total quantity will be overstated by a factor of two!
- Often there is too little time or lack of available resources to investigate discrepancies. As a result, wrong quantities are posted to inventory. Wrong quantities lead to lack of faith in the numbers and the cycle of distrust starts all over again.
- Confusion over open order and allocations can cause personnel not to count items that should be counted or to conclude that the parts are included in WIP when in fact the parts are still located in Raw Materials.
- Often companies call items “inventory” but they don’t have a good method to identify these parts so they become part of an accumulation of unidentified inventory.
- Entry errors can be another source of problems causing incorrect entries to be posted to inventory and the general ledger. These can be errors such as typing in a wrong quantity or typing in the wrong part number.
- When you can’t trust the inventory numbers, morale often lags because you question the ability of the company to perform even the most simple of tasks.
- Research indicated that in general, a physical inventory created more problems than it solved!
So, how does a company reach the point of trusting their numbers? You start by taking one final inventory. About 1989, Thunderbird did just that and there hasn’t been another physical inventory at Thunderbird since!
We started out with one primary goal: we had to maintain a minimum of 95% inventory accuracy.
We probably should define inventory accuracy. In general, if, when you count a part, the quantity you count is plus or minus 5% from the actual counted quantity for the part. As long as the quantity counted is close enough to be within the acceptable range that count would be considered a hit. If the count falls outside that range, it would be a miss. If you count 100 items and 95 or more fall within the acceptable range, then your inventory is above 95% accurate. There may also be times that you can’t afford any variation. For Thunderbird, those parts were boat engines. We felt that boat engines should always be 100% accurate without any variance range allowed. Typically, these parts that fall into this category are very valuable and have a ready market to buy the item.
We kept track of our inventory using a class system. For example, at Thunderbird, we established four classes:
- Control group – This is a group of parts that are representative of the inventory in general. In other words, these parts helped us identify parts that were not being relieved properly from inventory or if we were failing to record activities properly. These items were counted at least weekly and in some cases daily. If these parts were off, it typically meant that we had failed to record some activity properly such as relieving a manufacturing kit from raw inventory. A manufacturing kit is a collection of parts that were needed to manufacture an assembly or subassembly.
- “A” items were high value items that were counted once per week.
- “B” items were generally more readily available and not as expensive. We would count these parts about monthly.
- “C” items were usually inexpensive and readily available like nuts, bolts and screws.
Using this process, we would count every item at least 4 times each year. We assigned an individual as our cycle counter. This person needs to have certain traits. They must be very detail oriented, enjoy the challenge of digging into to a problem and figuring out what happened and then capable of helping to develop procedures to help avoid having the problem continue to occur. Our first cycle counter was a gentleman named Richard Black.
Richard was excellent at what he did. Each day, before production began, he would go out and count the items. Enter the data into our software system and then review the results. When there were problems he would do research and feed solutions back to the proper individuals. If the problem was in our bills of materials he would make sure the bills were updated. If we had failed to record transfer of inventory he’d get that problem corrected or if we had not recorded the proper number of items received from a vendor, he knew how to make those adjustments as well.
The key here was to find a person with the motivation and drive to get to the root cause of the mistakes and fix them. Also, Richard knew all our parts, knew the unit of measure for each part and then graphed and published the results so we could visually see our cycle counting progress. We didn’t face the errors generated by a physical inventory and that created trust in the numbers. When purchasing trusts the numbers, purchasing becomes far more efficient.