Counting the Vault: A Look at Variances

Post by: Jason Bader - Principal, The Distribution Team

Almost every distributor has gone through the painful task of taking a physical inventory. For those who have adopted cycle counting as a way of life in their organizations, annual physicals are becoming a nostalgic memory. I have a good friend who hasn’t taken an annual count in over 10 years. It’s not that he is careless in his accounting practices; on the contrary, his team has achieved a consistent inventory accuracy of 97%. It took him a few years to get there; but for most of us, the high 90 percent arena is a very respectable goal to achieve. Additionally, this company measured itself the hard way. Their inventory accuracy percentage represented the number of times an item count in their system matched the shelf count in the vault.

I recently worked with a distributor who wanted to begin a cycle counting program. After working out some vault logistics, such as bin locations and daily numbers to count, we came to the largest hurdle in the process. Certain branches felt that they should be exempt from the program because their inventories were never off by more than a few hundred dollars. This company made an effort to do bi-annual physical inventories in each of their branch locations. Every time they plowed through the process, the overall dollar variance was never more than a few percentage points. When you are holding a million dollars in inventory, who could argue with a variance of less than a thousand dollars? It could be argued that this location boasts a 99% inventory accuracy.

This view of inventory accuracy may give the owner an idea of what they have in the vault; but from a customer service perspective, it will cost us a great deal of net profit. In order to really judge how well our operation is handling the inventory asset, we need to look at the variance in pieces. Over the years, I have been involved in numerous annual physical inventory counts. I can recall be elated when a particular vendor line would have an overall dollar variance of less than a couple percent. This meant that I would not have to recount the line. It had met the acceptable tolerance criteria. Unfortunately, when I would examine the actual item count variances, there would be huge swings from the positive to the negative. But hey, what did I care? The dollars averaged out.

When creating the argument for switching from dollar variance examination to item variance, we need to ask our self one simple question: When our customers buy from us, do they order in dollars or in pieces? When was the last time a customer called up and ordered $400 in riser clamps? I haven’t run into it yet. The standard by which our customers ask us to fill their needs is pieces. This is how we are judged. In order to maintain a high level of customer service, we must know how many pieces are on the shelf. In most operations, there is that magic number where the inside sales person looks at the screen and begins to question if they can fill the order. A customer may want four packs of recip saw blades, and the system says we have four packs of recip saw blades; but our inside sales person is hesitant to commit. They have been burned before. What typically happens is that our sales person puts the customer on hold, runs out to physically check stock, than comes back in to complete the transaction. Does this happen in your place? We refer to it as “cycle counting by order”.

Wouldn’t it be great if our inside sales people could trust the counts in the system? I have often wondered how many additional inside sales people we employ because we refuse to make inventory accuracy a priority in our companies. I am not suggesting that we should focus on reducing headcount. By giving our inside sales representatives the tools they need, we are freeing them up to do a little add on selling. Would you like fries with your order?

Another argument for establishing a “pieces versus dollars” mentality comes from the buy side. How do we buy products from our suppliers? Obviously we order in pieces or packs. Without accurate shelf counts in our system, how do our purchasing agents make buying decisions? In many multi-branch operations, purchasing decisions are made from a central location. They do not have the luxury of wandering out to the vault to verify counts. They must rely strictly on the data in the system. Purchasing takes the occasional beating from branch managers for not ordering enough product to satisfy the needs of their customers. Yet the branch manager is too busy to cycle count. Besides, during the last physical count, their dollar variance was only off by one percent. Inventory accuracy by pieces is critical when purchasing decisions are made from a remote location. The most common fix to inventory accuracy problems in a branch is to keep a little cushion. Just so we are all on the same page; cushion refers to a bunch of really unproductive inventory costing you money.

The switch in focus needs to come from the top. If the officers of the company are still talking in terms of dollar variance, data accuracy will never become a reality in the organization. Take a look at how your system handles a physical count. Is the variance trigger set in dollars? Switch the system settings to review the item variances. Set an acceptable percentage of variance. Work to lower your pain tolerance. If the goal of this year is to work on a 10% variance, move to 8% next year. As your inventory in pieces becomes more accurate, the accuracy in dollars will follow suit.


Portland, Oregon • Charlotte, North Carolina