Inventory Optimization for High-Mix Manufacturers & Distributors
Whitepaper
Inventory Optimization for High-Mix Manufacturers & Distributors
Today’s supply chain landscape is being transformed from high-volume to high-mix. This means a broader product mix with more demand variability, more complex global supply networks, and more decisions to be made when it comes to balancing inventory holding costs against customer service.
Every company is dealing with the same problem: for some products, they have too much inventory, but for other products they don’t have enough inventory, which leads to stock outs, expedited shipments, and potentially even the loss of customers.
There is often an internal debate inside companies when it comes to inventory; on one side, senior leadership and sales often prefer to store large amounts of inventory in order to guarantee on-time shipment and provide their customers with high levels of service. On the other side of the debate, finance, supply chain and logistics managers struggle because of the many costs associated with high levels of inventory: storage & handling, holding costs, and obsolescence. It helps when both sides can understand best practices for inventory optimization.
Written for manufacturing & distribution executives and supply chain managers, this whitepaper describes strategies companies can deploy to “right-size” their inventory in high-mix industries where SKU counts are in the 1000s.
In this whitepaper we will summarize the purpose of inventory and look at the four maturity levels of inventory management, with the pros and cons of each method.
Understanding The Need For Inventory
Inventory is a means to an end, not the end product itself. In a perfect world where demand is constant, a manufacturer or distributor would only stock what was required for sales each day; inventory would be nearly zero and customer service would always be a 100%. For example, if you knew your business was going to sell 100 pallets of a widget every month, you would only produce 100 pallets to sell, and then you’d take the rest of the day off to golf and celebrate your good fortune in working in such a stress-free industry!
Unfortunately, the real world isn’t so simple. The problem arises when you introduce the complexity of modern high-mix supply chains into your business models. What happens when you’re making 1000 products instead of one? What happens when some products began to have variation in demand patterns, where you might need 100 pallets one week, 10 the next, and 300 the following? What happens when you have seasonality, long and unreliable supplier lead times, and other sources of variability in your Sales and Operational Planning process?
The answer is you’re going to need inventory, and lots of it.
The image below illustrates an analogy using inventory as “water” and the “rocks” in the water as sources of variability that need to buffered against in order to ensure the boat (the product) reaches the shore (the customer). Notice all the different forms of variability in supply and demand that might require a high water” level to prevent a collision (stock out or late shipment).
Figure 1: Sources of Variability Requiring Inventory to Buffer
In short, Inventory is a buffer against variability and uncertainty in your supply chain. To understand how much inventory is required in your company, you need to understand this variability, and then set your inventory “water” levels high enough to cover up your high-mix variability “rocks.
The Trade Off Curve Between Inventory and Customer Service
It’s same problem everywhere, over and over again: too much inventory for some SKUs, and not enough for others. If you’re a planner making inventory stocking decisions in your warehouse, you probably tend to err on the side of caution and overstock your finished products rather than risk hurting customer service. Having high inventory handling costs might earn you a slap on the wrist, but losing a customer due to late shipments could cost you a job. As a result, most high-mix companies nowadays are carrying more inventory than they need, and it’s not unusual to see some SKUs that have over a years worth of inventory.
The following trade-off curve is typical for most high-mix manufacturers and distributors. The X-axis shows what percentage of the time you are able to satisfy your customers orders from existing inventory. The Y-axis shows how much average inventory is required if it’s correctly sized.
The green curve is the “optimal” allocation of inventory for each service level:
Figure 2: Trade-off between Inventory and Customer Service in High-Mix
Using the water and rocks analogy: the green curve shows how high the inventory waters needs to be to cover up the rocks a certain percentage of the time to guarantee “% demand satisfied” i.e. customer orders shipped on time, in full.
The trade-off curve is consistent for most high-mix manufactures. As your targeted customer service level approaches 100%, the inventory investment required to achieve that service approaches “infinity”. In plain terms, it means that as you invest more inventory to buffer against wider sources of variability in supply & demand, the cost of inventory will rise non-linearly, i.e. buying a 1% change in service will be cheaper at lower service levels than at higher service levels.
Once you’re granted visibility into the trade-off between your inventory costs and customer service, you are now equipped with the information necessary to make business decisions. Do we need a 99% service or is a 95% service enough? Should we aim for the same target service with all of our customers, or do certain customers warrant a higher investment in their service? Do we want to invest $5 million for 90% service or $8 million for 95% service? These are all business decisions you can make after you understand your trade-off curve.
Notice the blue dot in Figure 2 of the Trade-off Curve. That tends to be the area that most companies operate prior to optimizing inventory. The blue dot is above the optimal green trade-off curve because currently there is more inventory than required for some products, and it’s also to the left of the curve because, for other products, inventory is too low, which lowers customer service.
The histogram further illustrates how companies can have both “too much” and “not enough” inventory simultaneously. The charts illustrates a before/after picture of inventory volume before and after an inventory optimization analysis. You can see at the start of the analysis, the company had ~120 SKUs with 0-2 days of on-hand inventory (risking stock out), while simultaneously there were 180 SKUs with over a years’ worth of on-hand inventory. Using some operational research math, we can calculate the recommended case to compare it to the current case. Notice the chart shifts inwards. Inventory is added where it was needed, and removed where it was in excess. The end result is a more right-sized mix of inventory.
Figure 3: Histogram of Current vs. Recommended
Four Maturity Levels for Optimizing Inventory Levels
With over 15 years of experience working with a large number of high-mix manufacturers and distributors, Invistics has developed the following maturity model to help companies assess their current inventory management practices, and to help them improve their inventory management success:
1
Maturity Level #1: “Acoustical” Inventory Levels
The first maturity level for setting inventory is jokingly called “acoustical” because whoever yells the loudest gets their inventory stocked. In effect, inventory rises until the yelling stops. A surprisingly large percentage of
companies use this immature method for setting inventory levels; they raise inventory if customers or sales complains about stockouts, or they lower inventory when finance (or managers with limited space) complain that inventory is too high.
Acoustical inventory management is stressful for all parties involved, on either side of the megaphone. It’s usage doesn’t allow for root cause analysis and it enables endless loops of ‘fire-fighting’ and fluctuating inventory levels.
2
Maturity Level #2: Heuristics, aka. “Rules of Thumb”
Companies at the next maturity level use heuristics or “Rules of Thumb”. For example a company might hold 4 weeks of supply for all materials. Typically, a company will have an experienced planner or a group of planners who use “intuition” to make adjustments to the Rules-of-Thumb weighting in factors such as seasonality and historical demand averages. It’s a messy, unscientific approach that tends to err on the side of caution forcing a heavy investment into inventory in order to maintain customer service.
While a Rules-of-Thumb approach is easy to implement, it can lead to terrible results, particularly in the high-mix world of today, where variability is prevalent in both supply and demand. Due to variability, i.e. rocks in the water, Rules-of-Thumb approaches tend to fall short. Some products will be stocked with too much inventory, and others will have not enough.
3
Maturity Level #3: Excel using Textbook Analytical Approaches
The third maturity level goes beyond Rules-of-Thumb approaches by incorporating variability into the math. There are many textbook formulas to use, but the standard one most companies adopt is provided by the manufacturing organization, APICS. The following are their two formulas for calculating safety stock and reorder points:
Notice the equation not only looks at averages, but considers the variability of demand and leadtime, reflecting real-world variability in supply & demand. This is how maturity level 3 improves on previous Rules-of-Thumb approaches.
Companies that reach this level of maturity typically implement these formulas in spreadsheets such as Microsoft Excel. The inputs required for such a task are minimal and can be readily extracted from most ERP systems. The APICS formula in home grown spreadsheets, will work fine if certain criteria are met:
- No minimum order quantities
- Products with normal distribution of demand and leadtime, with relatively low variability
- Planners are available to manually update the Excel spreadsheet inputs on a regular basis
- Planners can use spreadsheet to convince management where to operate on trade-off curve
Unfortunately, most high-mix manufacturers and distributors have both minimum order quantities and products with high, non-normally distributed variability. Even if this is not the case, most planners are already swamped with their planning responsibilities, and lack the time required to build, update, and maintain these home-grown spreadsheets.
4
Maturity Level #4: Software Analytics to Right-Size Inventory Levels
The highest level of maturity for right-sizing inventory is to employ advanced analytics to optimize the entire product mix, understanding the impact of multiple forms of supply and demand variability. Invistics offers the Inventory Advisor software tool to assist planners in performing inventory optimization. The Inventory Advisor uses advanced analytics to calculate optimal inventory levels, such as Reorder Points or Safety Stock. These optimized numbers can then be fed back into your MRP/ERP systems to replenish your stock based on smarter numbers. The module extends current safety stock formulas (such as the APICS formula above), and provides many features that are unavailable for an Excel-based tool.
A software tool such as Inventory Advisor can drastically reduce the effort it takes to manage your inventory levels, particularly in an environment with hundreds to thousands of products. Additionally, a software tool has the advantage of built-in security and data integrity features that are unavailable to an Excel-based method, and can facilitate planning across multiple planners, divisions, and sites.
How does it work? Inventory Advisor automatically retrieves input data from ERP and other source systems, optimizes the inventory numbers, and then sends the updated targets back to the ERP systems for execution. The software expands upon the APICS math to have real world constraints and planning challenges that are often ignored by Excel-based tools. This is especially valuable in high-mix environments where demand variability is extremely volatile.
The APICS formula’s “z-score” assumes a Normal distribution for supply and demand, which becomes increasingly inaccurate as variability increases. The Inventory Advisor has several other distribution models that are more accurate than Normal when it comes to modeling high variability. For example, Inventory Advisor has the ability to take sporadic demand into consideration, in environments where a finished product may have many weeks of zero demand due to seasonality or promotional periods. Additionally, Excel based approaches often ignore real-world constraints such as minimum lot sizes, periodic review, and customer allowed lead time which can have a massive impact on stocking levels.
Download the full whitepaper by clicking the button below. We are passionate about helping companies improve their performance. Call us at 1-800-601-3456 for a free demo or consultation or to see if we can help reduce inventory costs and compliance risks, within a single facility or across the extended enterprise.