A-02 Janet Steiner Soukup, CPIM How Small Businesses Can Use a Cycle-Count Program to Control Inventory INTRODUCTION In today s economic environment, it is more critical than ever that small businesses manage their inventories effectively. Inventory is a reflection of management s inability to control itself. In a perfect world, all manufacturing companies would only have inventory that it receives, converts, and ships in the same day. Reality does not usually allow for companies to operate perfectly, but there are many things that can be done to minimize the amount of inventory needed to effectively support operations. There are many ways that a company can minimize the amount of inventory that it has to carry. Some of these steps would include accurate forecasting, reliable suppliers, a well-organized warehouse, and an efficient means of supplying the production floor with raw materials. One of the first things that companies must do is to make sure that inventory is organized so that it can be located quickly and easily. The best method to ensure accurate inventories is to establish an effective cycle count program. The goal of a good cycle counting program should be to find and fix processes that cause inventory to be inaccurate. Many companies implement cycle counting and achieve a high level of inventory accuracy. However, the mistake is often made to view the cycle count as a report card rather than as a tool to help the improvement process. INVENTORY PROCESS CONTROL Eighteen months ago, Westell Technologies, Inc., created the inventory process control department. The objective of this group is to identify methods to monitor and measure key processes within operations that would ensure maximum inventory accuracy. To do this, process control had to be involved whenever new processes were implemented to ensure that the resulting change improved current processes. Additionally, process control worked with manufacturing to review current processes and identify potential process improvements. Most importantly, process control was charged with developing an effective cycle count program. The goal was to achieve exemption from the annual physical inventory, and we had exactly six months to achieve the goal. CYCLE COUNTING GETTING STARTED The first 90 days were spent getting the newly created group organized and trained on their new job responsibilities. One of the first tasks was to revitalize the cycle count program that had been suspended the prior year. Accounting and process control agreed on certain criteria for the program: An ABC analysis is done quarterly by accounting. A predetermined number of As, Bs, and Cs are counted on a daily basis. Results are published weekly, monthly, quarterly, and annually. An activity-based costing (ABC) analysis is a means of segmenting inventory into groups based on velocity of the parts. Whether it is based on dollars, units, or a combination, it should be determined by what makes the most sense for each individual corporation. The goal of doing this analysis is that maximum focus should be on the most critical (high velocity) parts in inventory. Typically parts might be counted as follows: As are high-velocity/high-dollar inventory where maximum focus is required 12 times per year. Bs are moderately-used parts four times per year. Cs are low-unit/low-dollar parts one time per year. Results need to be published regularly in order to have an impact and should be summarized in order to view trends. REPORT PREPARATION Cycle count reports are prepared in Microsoft Excel. While the raw data is gathered from the enterprise resources planning (ERP) system, Westell has found that Excel provides the opportunity to report the data in a fashion that is most suitable to its needs. Key data that is collected would include: part number, supply area (departments or sub-departments), original system count, and adjusted system count. Information is summarized by supply area (such as stock, manufacturing, shipping, etc.) so that we can better understand where the inventory 2003 International Conference Proceedings, 2003 APICS The Educational Society for Resource Management 1
variances are located. This information is then published weekly, monthly, quarterly, and annually. Table 1 is a sample of the weekly results. Looking at the same results by department helps to focus in on where the inventory issues are the most significant. (See Table 2.) ANALYSIS OF THE CYCLE COUNT REPORT By late December 2001, we had the cycle count program up, and initial results were good. We then began weekly and monthly reports that helped us to identify some trends. The basic cycle count reports are published weekly. (See Table 3.) Table 1. Table 2. 2003 International Conference Proceedings, 2003 APICS The Educational Society for Resource Management 2
Table 3. Table 4. Total locations counted means each individual location a pat might be in not the number of parts. Hits means that 70.95 percent of all locations had exactly what the system indicated. Hits indicate the accuracy of each location. Hits were absolute, except on our production floor. We had a few tolerances because significant portions of our raw materials come on reels, and many of those parts are extremely small. As a result, small amounts of the parts can be lost loading and unloading reels on our machinery. (See Table 4.) Absolute dollars indicate how many dollars were lost or gained overall. Again, net dollars can be deceiving and hide certain process flaws. Net units indicate the actual number of units to be written off. Net dollars indicate the actual dollars to be written off. Second pass results are quite different: Hits improved slightly. Absolute units improved by 13.5 percent. Absolute dollars improved by 19.4 percent. Net units improved by 48.3 percent. Net dollars improved 140.3 percent. Notice on the chart above, these results are first pass. This is one of the first places good inventory analysts can step in and make a huge difference. At Westell, any inventory variance that is considered significant in either dollars or units is investigated by one of the process control analysts. Their job is to recount the current locations, review transactions, look at prior inventory locations, and attempt to either physically find the part or determine why the inventory is inaccurate. An example of second pass results can be seen in Table 5. What changed? The analysts found two of the locations to be correct when they recounted the parts. Additionally, they managed to find 5,000 pieces of a very expensive component, and that resulted in the net inventory loss changing to a net inventory gain. By just having someone else check the counts, we managed to save the company $14,000. Is this enough? In the long run, the cycle count data does not give us enough information, because we do not root cause every variance rather, only the significant ones. Therefore, we do not always have enough data. Process control looked for ways to gather more information, and the result was the inventory adjustment database. BEYOND THE OFFICIAL CYCLE COUNT THE INVENTORY ADJUSTMENT DATABASE While the cycle count program was certainly helping to determine some of the more obvious issues, too few parts were being counted to identify smaller issues. However, like most companies, there was a constant stream of unofficial requests to look for missing parts and/or to cycle in parts that had been found. In late 2001, process control started collecting the data from these unofficial counts on a Lotus Notes database in order to determine the root causes for these counts. Figure 1 is a sample database form. 2003 International Conference Proceedings, 2003 APICS The Educational Society for Resource Management 3
Table 5. Figure 1. ROOT CAUSE ANALYSIS With the creation of the inventory adjustment database, enough data was now being collected to understand what generated the inventory variances. Root causes should reflect the issues that most commonly impact the company s inventory accuracy. To utilize root cause analysis to improve processes, the following steps should be followed: Step #1 Identify the most common root causes (Below is a sample of the root causes Westell utilizes.) Analyst error: Means that a prior analyst correction was not accurate. Count error: First pass count was incorrect. Floor error: Either someone from the production floor failed to complete a process or completed the process incorrectly. 2003 International Conference Proceedings, 2003 APICS The Educational Society for Resource Management 4
Stockroom error: Someone from the stockroom failed to complete a process or completed the process incorrectly. Transaction error: An incomplete or incorrect transaction was done or an transaction was omitted where one was required. Vendor error: The vendor created the problem. Within tolerance: The count is wrong, but the number of units and associated dollars are so small that analysis is not warranted. Year-end physical error: The error was created in prior year-end physicals. Other: No specific error can be identified. This is often the case when the error is not identified immediately after it occurs. Picking error: When a replenishment order was being picked to be sent to the floor, an error occurred. Missed scrap transaction: Scrap transactions that should have occurred were missed. Autoscrap reversal: On some of our production lines, we have an automated report that collects and reports scrap. Sometimes, the machine skips a placement but does not actually lose the part. When this happens, a manual reversal might be in order. Step #2 Pareto the root causes to determine where the greatest numbers of errors have occurred A Pareto analysis of these counts turned up very compelling results. Each analyst is responsible for assessing what went wrong and correcting the system if transactions were missed, duplicated, etc. Again, this data was collected by supply area, so once the information is downloaded to Excel, we could collect the same data as the official cycle count. Table 6 shows the results from a nine-month period. We found that our most significant number of errors came from count errors. Step #3 Create teams to address process issues and implement appropriate solutions COUNT ERROR ANALYSIS Analysis determined that the large number of count errors was due to multiple causes. The most significant ones included disorganization in the stockroom, large amounts of unneeded work in process (WIP) on the production floor (thus making it very difficult for the cycle counters to find all the parts), and lack of training of employees. From October to May, count errors declined sharply due to reducing the amount of WIP on the production floor and reorganizing the stockroom. (See Table 7.) Table 6. 2003 International Conference Proceedings, 2003 APICS The Educational Society for Resource Management 5
FLOOR ERROR ANALYSIS When we investigated the floor errors, we determined that they were related to the count errors. This was obvious to the analysts, but the team validated that we had way too much inventory on the production floor. When production could not locate a part, it simply scrapped it in the system. When it was eventually found, a database was submitted to correct the inventory. By significantly reducing the volume of parts on the floor, floor errors dropped dramatically from October to March. (See Table 8.) TRANSACTION ERROR ANALYSIS Transaction errors were not really improving for most of the prior year due to a lack of adequate, and more importantly consistent, training to new employees. As a result, transaction errors continued to plague us. A team was put together to discuss options for correcting this problem. We are now implementing a much more robust training program that will include both classroom and hands-on training. Every employee who utilizes our ERP system will be certified as they complete each section of the training program. (See Table 9.) YEAR-END PHYSICAL ERROR ANALYSIS These errors, not unexpectedly, were huge. This was one of the most urgent reasons for gaining exemption from the physical inventory. Because so many of the counters were not regular stockroom Table 7. Table 8. Table 9. 2003 International Conference Proceedings, 2003 APICS The Educational Society for Resource Management 6
Table 10. personnel, many errors were created as a result of the year-end physical inventory. There was no expectation of improving in this category without getting the exemption from an annual physical. Even with physical exemption, it took through March to truly clear the system of all prior errors. (See Table 10.) PHYSICAL INVENTORY EXEMPTION In January 2002, Westell met with the external auditors and requested the exemption from physical. Over the next 90 days, multiple test counts were done with the auditors and proved that the cycle count program results were consistently coming in with overall net dollar accuracy of 99+ percent. As a result, in March 2002, Westell gained exemption from a year-end physical inventory. CONCLUSIONS In the 18 months since the inventory process control group was established, Westell has utilized results of the cycle count program along with other initiatives to improve the flow of inventory within manufacturing. The warehouse is now organized in an effective and efficient manner to maximize space utilization and maintain control of inventories to help drive the overall inventory levels down while on-time delivery of product improved. The creation of the inventory process control department and the impacts of both the cycle count program and inventory adjustment databasehave been profound. It is estimated that the annual physical inventory cost Westell about $250,000 per year. This includes the costs of downtime, overtime to make up for the downtime, costs of having auditors and counters here for the count, etc. Additionally, the analysts found more than $900,000 of lost inventory on just the official cycle counts, while the costs of running the department is about $400,000 per year. Clearly, this department is self-funding. While companies cannot always operate in a perfect world, it is possible with good tools and techniques to minimize the amount of inventory needed to effectively support operations and continuously improve processes. Cycle counting is a tool that every company can afford because it will always pay back more than it costs. ABOUT THE AUTHOR With more than 27 years of experience in operations, information technology, and marketing, Dr. Juan Montermoso, CPIM, CIRM, focuses on content management in his current role as manager, Business Planning, for Worldwide e-business at the Hewlett-Packard Company. 2003 International Conference Proceedings, 2003 APICS The Educational Society for Resource Management 7