If a business isn’t currently cycle counting their inventory, then they probably have a goal of switching from a single, arduous annual Physical Inventory to the simpler Cycle Count process. Where almost all businesses fail is by focusing far more time and energy on reporting the values, and not enough on improving the processes that deliver the results. Worse, when facing an ERP transformation project, both client and implementer may make the mistake of lifting the current process and configuration options and shifting them, rather than taking advantage of the moment to revisit the goal and process from the ground up.
If you are on a boat and you notice it is taking on water, you need to simultaneously use a bucket or pump to remove enough water to keep the boat from sinking, while also finding and repairing the leak. To many, the act of bailing water out seems to take precedence; but there is a real risk that you may decide that you could simply always run a small pump and ignore the leak. I propose that finding and repairing the leak is inherently the most important task, and therefore the first goal of cycle counting is not to update inventory counts, but to help an organization identify the root cause of inaccurate inventory, and fix it.
What is accurate inventory? While numbers like 95%, 98%, even 99% sound great, think about the terrifying world of 1% failure rates. Assuming you drive to work, 5 days a week 50 weeks a year, a 1% failure rate means 2-3 times a year, you are suffering a break down or a car accident. Your insurance and repair bills would be sky high!
Assuming your inventory is less accurate, how can we fix it? Every inaccuracy is an opportunity to find a gap in a process and fix it, so lets approach them with as positive a mindset as we can when we know that a missing piece may mean lost revenue, delayed production, or worse.
There are many continuous improvement methodologies, and advanced training in those is never a bad thing, but you can start today with simple, common sense approaches.
First, make sure you know your current processes. As they are on the shop floor, not those crafted in an ivory tower and scoffed at by the real warehouse employees. Also consider the tools, reports, environment and documentation that keeps things humming along.
Next, devise a method of counting, and go measure your inventory accuracy! Think practically, do not treat a 1-piece-of-10-million discrepancy with the same root cause finding urgency as a larger one that actually impacted production. Once you have some inaccuracies, it’s time to dig deep and find out why! The word that comes to mind is “triage!”
You may pick the largest and most material discrepancies to start. For example, one client after a cycle count, was missing 4 entire containers worth of the main component necessary to hold multiple millions of dollars worth of inventory that was in process and ready for final assembly. This was their most material discrepancy and it required immediate diagnosis to avoid a financial disaster.
In other situations, you may have a group of discrepancies that are not material individually, but in whole they represent a large drain on the organizations. One client managed a catalog of small labels. If they got mixed up, identifying the correct part could be difficult as the differences between each item were very minor, and tiny. In this situation, rather than analyzing each individual incident or inaccuracy, they grouped them and picked one to focus on, knowing that it would solve multiple others at the same time.
Once you have something to investigate, the fun starts! Put on your Sherlock Holmes hat and go find out what caused the issue! Here are some examples:
- Full and partial boxes stored on the same pallet, and the warehouse issued a partial box of items to the production floor because they didn’t realize it had been opened and partially used. The previous users had not marked the labels to indicate, and they had opened and re-closed the box carefully; the new pickers did not notice the weight difference because they were moving quickly and not using any technology to weigh the pieces.
- A pick list called for Eaches, but the product was wrapped in pairs or sets, and the picking team misunderstood that one wrapped item was 2 Eaches.
- The partial box stated it had a certain quantity, but it did not, and no counting scale was used to verify.
- Two suppliers use similar labels, but with item number and quantity fields (or some others) in different locations; by coincidence the part number was a small number and the quantity was a similar number, for one box, and the pickers mistakenly issued thinking the part number was the quantity.
- Paperwork indicating work completed was rubber stamped, and not confirmed, so pieces were marked as delivered when they were still left in the warehouse.
Those may seem like cut and dried examples, but you should always try to dig as deep and as far out as possible to make sure you have the real root cause. When you find a leaking seam on a boat, you need to know if you hit something in the water, if the metal was to specification, how the rivets were installed, and more, in addition to whether this is the only leaking spot, or even the worse leak, before you start to weld.
A popular concept is the “5 Why’s,” but the key isn’t really to always ask ‘Why?’ exactly 5 times, but to ensure you are really probing to get to the real root cause; this is because there is a strong tendency to simply blame human error. This makes workers feel insulted, less valuable, and doesn’t do anything to resolve the issue. Humans are inherently fallable, and machines aren’t perfect either.
It also helps you to categorize issues and appreciate that there are factors such as training, processes, technology, and machines, to review. Many times, a comprehensive fix requires changes in multiple areas.
For example, to resolve issues where transactions were done in the system but not physically, you may consider
- Training the team on proper steps and that they are empowered to not process a transaction if they are concerned about it being a mistake, without risk of punishment
- Improving the lighting, temperature, and other physical aspects of the workspace, and giving more breaks or alternate tasks to avoid workers being overtasked
- Supply or update the technology used, such as RF devices and weighing devices
- Reviewing the pick lists and changing the pick structure to find a more convenient way to pick items
- Include a review step where an alternate employee verifies details
After implementing changes, it’s time to count again and verify that your changes had the expected impact. But, recall that finding old errors from pre-change era, should not mislead you into thinking your changes were off mark. That is to say, after you fix the leak, you will still have water in the boat from that leak, and it may take time to dry out.
To that end, the secondary goal of cycle counting is to correct and update inventory counts, akin to removing the water with a bucket or pump. And for this purpose, you may consider using two lesser utilized ABC classification settings, or manual settings. Most organizations want to classify ABC based on valuation, or weighted valuation. Auditors may love this, but inventory managers will recognize that, assuming you have no theft and people are not touching inventory that is not required for work orders–who among us has time to move things needlessly?–then the only way inventory gets inaccurate is because it was used. Indeed, even the act of counting introduces a risk of injecting inaccuracies, further highlighting that we must count to correct processes, not in a misguided attempt to simply correct the inventory.
Therefore, you may consider using ABC classifications based on quantity usage instead of value. Alternatively, if you are counting items frequently, then you might also consider classifying based on prior inaccuracies. Finally, in a real emergency situation, you may consider using a custom query to identify inventory that meets certain criteria and generate manual count schedules.
For example, if you realize a major root cause of inventory inaccuracy, such as a defect in a label counting machine, and you suspect that all inventory used in a specific section of the label storage racks is at risk, then you could query for all inventory that has been used in a work order, regardless of count history, in that section, and generate manual counts. While ABC classification by use could help in this situation it may well be that in the overall scheme of items in your warehouse, those labels were a C class item by usage, but you need to count them all immediately without waiting.
A fundamental theme here in inventory management is identifying a valid KPI that will allow you the freedom and flexibility you need for proper root cause analysis, while still demonstrating your progress in a timely manner. Far too often, managers focus just on the numbers, like count of locations inspected, and fail to even consider how the impacts of a process change may not be immediately reflected in the count accuracy values.
Worse, when you are aware of a problem, the secondary type of count to get the water out, only results in a harm to your KPI and in some organizations, results in shadow counts being organized in Excel worksheets to avoid making the team look bad.
Trust and empowerment build a stronger team and that strong team running a Cycle Count program will build an organization that ships items on time without losing sleep at night. The best moment to cast a new vision for inventory management and cycle counting, is during the implementation of an advanced ERP like Oracle Cloud, which brings strong tools and reporting to help achieve the goal, and the team is receptive to change.