APICS is the premier professional association for supply chain management.

One-of-a-Kind Inventory Management

By Mark E. Neuman | July/August 2012 | 22 | 4

One-of-a-Kind  Inventory ManagementDesigning the right cycle counting system for your business

As supply chain and operations management professionals, we all are familiar with ABC analysis, Pareto’s law, and the 80-20 rule. When it comes to cycle counting, these basic principles have their place—but a generic, cookie-cutter approach won’t work every time. Take a step back from the most basic guidelines, and consider the following questions:
  • Does Pareto’s law really fit your business? Are 80 percent of your inventory dollars wrapped up in 20 percent of your items (and vice versa)?
  • Does ABC analysis hold up at your company? How many low-dollar, high-volume items such as screws, nuts, bolts, and washers do you have? 
  • What type of industry do you represent? Is your company a job shop, a custom builder, a process or discrete manufacturer, or something else? At what pace are you producing? Do you deal with high or low volumes? Push systems or pull systems?

Many leaders don’t have a good sense of their organizations’ identities or have lost track of what they have become. Before you can fully understand your business, you may have to do some digging. Only then can you implement the right cycle counting program to support your company. 

Evaluate the current state

Regardless of the nature of your industry or your experience level, a successful cycle counting program begins with the following steps. 

Determine inventory mix. In other words, figure out what percentage of inventory dollars represents what proportion of items, and vice versa. As this figure approaches 50-50, it makes more sense to treat each item as having the same cycle time. But when inventory begins to more closely resemble the 80-20 rule, classic cycle times are more likely the best fit: 4 weeks for A items, 8 weeks for B items, and 26 weeks for C items. You may wish to tweak these numbers further. As most enterprise resources planning (ERP) packages include only a standard analysis, it may be necessary to work closely with your information technology department to modify the software program’s input and reporting structures. 

Count the number of locations and parts. This will aid in determining the number of resources necessary to support whatever cycle is necessary.

Examine material flow through the warehouse. Answer questions such as: Do your items have dedicated storage bins? Are the bins placed at random? Do some items stay in inventory perpetually? Are there items with short lives due to only custom demand? This will aid in deciding whether to focus on locations, materials, or even materials within locations.

Set accuracy goals. Try not to get hung up on this step—just pick a number. Once the program begins and you start seeing accuracy data, you will know where you are and can update the goal accordingly. All cycle counting systems should have longer- as well as shorter-term goals, usually annual with monthly increments. 

Define the program’s purpose. This step actually is quite easy. There is only one purpose to cycle counting—to find out how errors occur and eliminate the practices that created them. As you eliminate the root cause of the errors, inventory accuracy rises. Those who say they cycle count to maintain accuracy do not have the right focus. 

My company, TSE Industries, is a Clearwater, Florida-based manufacturer of custom-machined plastic parts, custom rubber extrusions, compression-molded rubber, and cast urethane. We maintain a job shop environment, and most manufacturing is performed make-to-order. Additionally, there are a number of specialty chemicals produced in a process environment, sometimes make-to-stock and sometimes make-to-order.

As far as inventory mix goes, 60 percent of our inventory dollars are wrapped up in 40 percent of goods, and vice versa. There are about 1,800 locations and 5,000 individual parts to deal with. 

We use a random location system, as there’s not enough room for dedicated bins. The number of parts per bin varies. Bins are mixed and matched to ensure an even amount of work each day. This enables counting to begin and end and research into errors to be completed the same day. We examined how items move through the warehouse to and from work in process, subcontractors, and customers.

Eventually, after redefining the organization, we identified a cycle counting program that worked best for us.

One company’s system
Our lead times are relatively short, and we carry an average of three months of inventory for the custom orders that represent most of our volume. The majority of our materials have a similar ability to shut down production, and heavy investment in inventory as with classic C items is undesirable. The 60/40 split of our inventory further supports the notion that each item deserves the same amount of attention. With these factors in mind, it did not take much effort to figure out that a quarterly cycle for the entire inventory was appropriate.

To ensure the entire inventory is cycled through, we assigned each of our 1,800 bin locations a count date. After a day’s locations are counted, they are tracked in a computer database. It’s easy to refer back to the last date a bin was cycle counted when an error occurs. Seven workers perform the daily counts along with one data administrator. The counts are performed blind—that is, the material handlers do not know what is in the database when counting. If parts are found in a location the system has not accounted for, workers record this information on a form, which later is used to aid in reconciliation. 

About 27 locations are counted per day, with an average of 75 parts. The best element of this system—assuming the work is divided correctly—is that it requires only about 30 minutes of work for each material handler and about three hours for the administrator. No extra people and no overtime are necessary.

Once material handlers turn in their count sheets, the administrative work can begin. Counts are checked against the system. If a discrepancy is found, the administrator first checks for any pending updates that would bring the quantities into sync. If that does not provide a match, the area is recounted. If the recount doesn’t match, we perform a root cause analysis. Most importantly, this evaluation begins and ends on the same day—an absolute must for an effective program.

Initially, the program was 93 to 95 percent accurate. We attributed the high level of accuracy to existing discipline and reconciliation programs. In subsequent years, we have raised the bar, as our annual goals inch toward 100 percent. In the last three years, we have maintained a consistent 99.5 to 100 percent accuracy.

When we discover an error, it becomes coded and tracked among a range of issues. Our list of causes includes 
  • material transferred incorrectly
  • material received incorrectly
  • information recorded incorrectly
  • material issued incorrectly
  • material shipped incorrectly
  • policy violation
  • unknown.
Of course, we consider the elusive “unknown” category to be our nemesis. But its existence helps us focus on eliminating it altogether. We maintain a laundry list of possible reasons for unknowns, and we continue to chip away at them.

One way we’ve found to speed the process is to recognize sealed containers from previous cycle counts. As long as the container is still properly sealed and unopened, the count is taken directly from the container. Units that continue to be unused are moved to remote locations where the counting is done even more quickly. These materials transfer to a slow-move candidate list, and hopefully eventually out of inventory altogether.

Despite the opinions of many, cycle counting is not a process to maintain accuracy; it is designed to discover how errors occurred and to eliminate the possibility of them recurring. We are tremendously satisfied with our program in keeping track of the causes of errors and the overall health of our inventory. Parts are available nearly 100 percent of the time, and there are very few occasions when production is stopped or a customer is shorted due to an inventory error. Finally, one of the greatest benefits is that our people now believe in the inventory figures because they know how the data were generated. This has instilled our workforce with confidence, pride, and boosted morale. 

Mark E. Neuman is vice president of materials at TSE Industries, a global plastics, rubber, and specialty chemicals manufacturer. He has nearly 40 years of experience in materials management in the manufacturing arena, and he is a former member of the APICS board of directors and past president of the Florida West Coast chapter. He may be contacted at mark.neuman@tse-industries.com

To comment on this article, send a message to feedback@apics.org.

All comments will be published pending approval. Read the APICS Comment Policy.


  1. RadEditor - HTML WYSIWYG Editor. MS Word-like content editing experience thanks to a rich set of formatting tools, dropdowns, dialogs, system modules and built-in spell-check.
    RadEditor's components - toolbar, content area, modes and modules
    Toolbar's wrapper 
    Content area wrapper
    RadEditor's bottom area: Design, Html and Preview modes, Statistics module and resize handle.
    It contains RadEditor's Modes/views (HTML, Design and Preview), Statistics and Resizer
    Editor Mode buttonsStatistics moduleEditor resizer
    RadEditor's Modules - special tools used to provide extra information such as Tag Inspector, Real Time HTML Viewer, Tag Properties and other.

Please log in to see content on this page

This page article is available to APICS members and APICS magazine subscribers only. APICS members and APICS magazine subscribers: Please log in at the top right of this page to view this content.

Not an APICS member? Join today to receive instant access to valuable APICS member benefits like this.