To analyze the stock model, select the Inventory Analysis button. When the calculation is complete, a new window will appear with the results page of the analysis.
Any warnings with data entry will be listed in the Notifications pane that will appear at the bottom of the UI. Warnings will also appear at the top of the results page. Another indication of errors in data entry is a results page that has charts missing and/or skewed to the right. If this happens, review and correct the list of notifications. Note that some notifications are to inform or warn the modeler and are not a sign of an error in data entry that needs to be corrected.
If all data was entered correctly, the results page should resemble the following image.
Inventory Results Page Layout
Like the Production System Analysis, the Inventory Analysis groups analyses by a specific setting. In the Production System Analysis, analyses are grouped by Product Flows. In the Inventory Analysis, analyses are grouped by Stock Points.
Select the appropriate stock point from the drop-down list at the top center of the page. In this model, there is only one ending stock point, “Component Inventory”.
Under the stock point selection, the current inventory policy for the stock point is shown. The data displayed is the entered information about the selected stock point. However, the last three data, Minimum Possible Fill Rate, Maximum Possible Fill Rate, and Minimum Possible Investment are computed values based on the information provided for the items associated with the stock point.
The Minimum Possible Fill Rate is a demand-weighted fill rate computed by looking at the minimum fill rates of all the items associated with the stock point. In each item menu, there is an input field for the user to set a minimum fill rate. Occasionally, the user will want to specify a fixed reorder quantity for an item due to logistical considerations, e.g. a supplier always ships in quantities of 300. However, if an item has a fixed reorder quantity associated with it, this is used to compute the fill rate instead of using the optimal calculated minimum fill rate of that item. If none of the items are assigned minimum fill rates and none have fixed reorder quantities, then the Minimum Possible Fill Rate will be zero.
The Maximum Possible Fill Rate is a demand-weighted fill rate computed by using the defined fixed reorder quantity values of the associated items of the stock point. If an item does not have a fixed reorder quantity, then its maximum fill rate will be 100%. If none of the items have fixed reorder quantities, then this value will be 100%. If all of the items associated with the stock point have fixed reorder quantities, then the Minimum Possible Fill Rate and Maximum Possible Fill Rate will be equal.
The Minimum Possible Investment is the dollar value of the inventory needed to support the Minimum Possible Fill Rate using the predicted average number of re-orders per period and the total unit cost of the items.
Inventory Tradeoff Plot
The Inventory Optimizer uses the input data like unit cost, mean and variance of demand, and mean and standard deviation of replenishment time to create efficient frontier curves that show the tradeoff between fill rates and inventory investment. Any point along these curves is an Optimal Policy for a chosen, fill rate, inventory investment and order frequency. Each line in the tradeoff plot represent a different order frequency specified in the stock point table. The tradeoff plot shows improvement potential from the current situation and the additional reduction potential associated with increasing order frequency.
The Actual (yellow dot) and Predicted (grey dot) policies are also shown in the Tradeoff Plot to contrast optimal policy options with current performance. Predicted represents the planned inventory control policy. The “prediction” comes from using current replenishment policies input by the user in the item tables. Actual values come from a sum product of all on-hand quantities and total unit cost of each item (also in the item tables) and the Current Fill Rate (%) set in the stock point table. Current Fill Rate can be identified by pulling a report from their ERP system. Many ERP systems have built-in reports or dashboards that display fill rate information. Look for terms like “Order Fulfillment,” “Fill Rate,” or “Order Accuracy” in the reporting section. If the ERP system does not directly calculate the fill rate, you can export the necessary data to perform the calculation manually. The basic formula for fill rate is:
By comparing Actual and Predicted performance, one can conclude if a deviation from the current policy has happened. For the XYZ Assembly Company, the predicted values are quite low both in fill rate and inventory investment. This explains why the company was experiencing frequent stock outs. A Fill rate of 52.22% means that almost half of the raw materials needed for internal production orders could not be fulfilled. What is interesting, is the deviation between predicted and actual values. Clearly, someone in the purchasing department made a decision to forego the current ordering policies and order much more raw materials to combat the frequent stock outs. And clearly, the decision is working. The company experienced a 26% increase in fill rate.
Current Inventory Policy
Underneath the Inventory Tradeoff Plot is the Current Inventory Policy. This table displays the details about the current policy for each of the items associated with the stock point. The “Environment” and “Current Policy” headings represent input data and the “Predicted Results Under Current Conditions” heading represents predicted results computed by the system.
The information in the “Current” row uses data entered by the user in the item table and the stock point table. Current refers to the Actual point in the Tradeoff Plot. The information in the “Predicted” row is what the model predicts would be achieved based on the current inventory policies entered into the item tables: ROQ and ROP or Planned Lead Time, Safety Stock, and Days of Supply. This is the same plotting data for the Predicted dot in the Tradeoff Plot.
A frequent source of confusion within the Current Policy table is which values to take and input into a company’s ERP system. The Inventory Optimizer purposely forces the user to use either ROP and ROQ or Safety Stock and Days of Supply depending on if the user is operating with a fixed order policy or a time-phased policy. Regardless, each pair of data can be calculated using the other pair as follows:
ROP = Replenishment Time Demand + Safety Stock = (Lead Time)*(Daily Demand) + Safety Stock
ROQ = (Days of Supply)(Daily Demand)
Therefore, if a user’s ERP system uses one of the parameters from each pair, they can be calculated using the formulas above.
Calculating an Optimal Inventory Policy
Recall, that the objective of the company was to maintain a 95% fill rate with less than $3 million of cash tied up in stock inventory. Additional data indicated that the purchasing team would be overburdened if more than 25 orders were placed per week.
With this information, any inventory policy with less than 25 orders per week, falling under $3 million investment and having a fill rate of ≥ 95% is a win. So, which should be picked?
By adjusting the sliders below the Tradeoff Plot, a user can zoom into smaller section of the chart. It is clear that all three lines run on top of each other. This means that all three reorder frequencies chosen in this example will produce the same Fill Rate for any inventory investment value.
Despite its common usage, fill rate is not always the best measure of customer service. The fill rate is a “zero-one” measure. If it is not available when it is needed the “grade” is a zero, otherwise, a one. Indeed, the result is a zero if it is a month late or even an hour late. A better measure is that of backorder days—the average number of days that a part is back ordered. The greater the backorder days the longer it will take to fill the order when there is a stock out. Unfortunately, most people do not have a good intuition for backorder days. That is why it is important to use both fill rate and backorder days. The fill rate predicts how often there will be a problem while the backorder days reveal just how bad the problem will be. For instance, it might be better to have a 80% fill rate with a 0.5 average backorders days than a 95% fill rate with 20 average backorder days.
The inventory analysis must consider Fill Rate, Backorder Days, Inventory Investment, the order frequency and the space and infrastructure required to deal with smaller, more frequent or larger, less frequent replenishments.
In this example, the same fill rate and inventory investment can be achieved for a replenishment frequency of 10 orders per week and 5 orders per week. 5 orders per week initially seems like a better choice because it creates less burden on purchasing and receiving teams but 10 orders per week gives an average 40% decrease in backorder days for all items in the stock point.
What if there is not a clickable point on the curves for the policy you want?
If this occurs, like what can be seen from the above image, the user can use the input fields just above the Tradeoff Plot. Simply provide the reorder frequency in the first box and the ideal fill rate in the second box and click Calculate. The results of this optimal policy will appear below.
Optimal Results
Once an optimal policy is chosen, whether it is from a point on a curve or from the calculation in the fields above the Tradeoff Plot, 3 things will happen.
- The Predicted Inventory Investment Profile simulation can be ran
- The Comparison of Fill Rates Chart will become visible
- The details of the Optimal Inventory Policy become visible
The Predicted Inventory Investment Profile simulation can be ran using the input fields directly below the Fill Rate slider bar. Start by selecting the time horizon for which the simulation should be run. The user can pick any number of Days or Periods (weeks). A good starting point is 365 Days. The input fields on the right enables the user to specify a factor to increase or decrease average demand and visualize what happens to the inventory investment profile over time given the fluctuations. Selecting calculate, runs the simulation and plots the profile(s). If desired, this data can be exported to a Microsoft Excel file by selecting the Excel icon.
The Comparison of Fill Rates Histogram is depicted next. This chart compares the fill rates between the current and optimal policies. Each column counts the number of items from the current and optimal policies that fall within a Fill Rate percentage range. Ideally, a user would want most of their items clustered around their target Fill Rate. In this example, the spread of fill rates among items using the current policy is wide (30% – 100%), revealing that many of the items have very poor fill rates. Alternatively, the optimal policy clusters the fill rates of each item between 75% and 100%
Beneath this chart in the results page are two tables contrasting the details of the current policy and the optimal policy. For ease of comparison, both of these tables can be combined into one table by selecting the toggle switch at the top-right corner of the Current Inventory Policy table. In the combined table, the bold font represents the optimal policy and the normal weighted font represents the current policy.
If the system uses a (ROQ, ROP) policy for managing inventory, the Reorder Point and Reorder Quantity columns will be used. If the system uses time-phased reorder points, the Planned Lead Time, Safety Stock, and Days of Supply columns will be used. All current and optimal policy parameters can be exported to a Microsoft Excel file using the buttons on the top-left hand side of the combined table. These values can be uploaded into the current materials planning system for immediate adherence to the optimal inventory policy.
Inventory Policy Data Element Descriptions
| Data Element | Source | Description |
|---|---|---|
| ID | Item Table | Unique identifier given to each item upon creating |
| Description | Item Table | Description of the item |
| Unit Cost | Item Table | Value of the item, usually valued at Cost of Goods Sold |
| Avg Dmd in Period (Units) | Item Table | The total average demand of the item |
| CV LT Dmd | Item Table | Standard deviation of lead time demand divided by mean lead time demand |
| % LT Dmd Variability (Demand) | Computed | Indicates level of inventory requirements driven by a combination of replenishment time and demand variability |
| % LT Dmd Variability (Supply) | Computed | Indicates level of inventory requirements driven by a combination of demand and replenishment time variability |
| Planned Lead Time (Days) | Item Table | Used for Time-Phased policy. Planned Lead Time (PLT) is the length of the planning horizon used to determine when to place a replenishment order. For example, if the PLT is 10 days, when projected on-hand balance goes negative 10 days from now, a replenishment order should be placed today. |
| Reorder Point (Units) | Item Table / Computed | Inventory Position at which next order for product is triggered. Inventory Position = On Hand + On Order – Back orders |
| Safety Stock (Units) | Item Table / Computed | Used for Time-Phased policy. Amount of stock to keep on-hand to buffer demand and replenishment time variability. |
| Reorder Quantity (Units) | Item Table / Computed | Current standard order quantity of part |
| Days of Supply | Item Table / Computed | Used for Time-Phased Reorder policy. Current standard order quantity of part. |
| Avg No. Reorders (Units) | Computed | Average total demand per period for an item / order quantity |
| Fill Rate (%) | Computed or entered by user | Probability of having no backorders. Red line (Actual) fill rate is entered by user, not computed. Yellow line (Predicted) fill rate is demand weighted average fill rate for all items in the stock point. For each item, fill rate is calculated from current inventory parameters specified in the Item definitions. |
| Avg Backorder Days (when short) | Computed | Average number of days that it will take to fill a backorder when there is no inventory on hand the day the order appears. |
| Value of Inventory | Computed | Average number of units on hand * unit cost. Actual inventory is sum of dollar value of all items in stock point. Predicted inventory is amount required for predicted fill rate resulting from reorder points and reorder quantities entered by the user. |
A Note on Order Frequencies
The best point to select is the one that provides a policy with the highest fill rate, the least possible investment, and the lowest frequency of replenishments. When picking an order frequency, it is important to consider the infrastructure required to handle each replenishment and the space required to store the replenished inventory. Less frequent replenishments typically come with larger-size replenishments and require more storage space. The opposite goes for more frequent replenishments.