DemandCaster analytics are designed to help users focus on the most important items so that appropriate forecasting, planning, and supply policies may be applied. This article describes the organization of the DemandCaster analytic processes.
Analytics are designed to help set an items inventory basis at each location. The critical options for each item/location are:
- Stock versus not stock
- Replenish by forecast or min/max
- What level of safety stock to apply – the more volatile the demand the higher the inventory
- Higher service levels for A, lower for B and C
- What should the lead time be – the longer the lead times the higher the inventory
- What is the distribution network
Set supplier for a given item to be the most optimal preferred supplier
Set the distribution ship from location and its source to the most optimal path for fulfillment
The Importance of Classification
Classification is a key principle behind inventory management. It is based on the premise that not all items are alike and not having a 100% fill rate on every item carried is a perfectly reasonable business outcome.
Set demand and supply policy based on:
- Business Importance
- Order Frequency
Recognize that the goal is not 100% in stock. Increase in inventory for higher fill rate is significant:
- 90% to 95% requires a 65% increase in stock
- 95% to 99% requires a 133% increase in stock
The analytics are built with a logical sequential work flow as described below and organized under the Analytics menu group. Click on each of the hyperlinks below for a deep dive on each analytic.
- The Classification analysis sets the Business Importance (BI), Order Frequency (OF), and Accuracy (A) by item and location (if applicable). The BI ranking is based on revenue and margin contribution, cost, and units of an item over the last 52 weeks. The basis is set in system settings. The Classification analytic also calculates the Accuracy (A), Forecastability (F) and Order Frequency (OF) of an item and location (if applicable). Order Frequency calculates the ranking of an item based on how many customer order line items have been created for the item over the last 52 weeks. These metrics are used to apply appropriate demand and supply planning approaches for ordering and inventory management.
- The Forecast analysis provides the detail of the forecast source including statistical measures. It also includes a recommendation on the forecasting/demand planning approach to employ for a given item.
- The Lead Time analysis calculates the actual versus stated (planned) lead time by item and location (if applicable). Lead time is the typical number of calendar days to expect an item to arrive at a location from order to receipt. Correct lead times are necessary to ensure proper order point calculations. The analytic also calculates the on-time performance of supply from purchase order and production order data.
- The Stocking analysis helps to determine what should be stocked and maintained in inventory versus not stocked and made or purchase to ordered.
- The Stocking analytic and Lead Time analytic drives the Order Point Calc which calculates the safety stocks and order points for those items that are stocked. This analytic also provides guidance regarding the validity of the order point being recommended based on historical performance.
- On a periodic basis (monthly is recommended) the Review analytic is reviewed to assess actual to plan performance among other metrics to assess if any of the set parameters need to modified. This could mean changes to lead times, forecasts, stock versus not stock among other variables.
All these analytics may be automated to run on a periodic basis using DemandCaster's Automation process.