Summary
This article walks through how lost sales is calculated and what it means.
Applies To
For Lost Sales to be calculated, a few things are necessary:
- The item must have a "Recurring" Usage Pattern - When trying to determine if an item has had lost sales you must have some confidence that there was a repeating pattern of sales before the stockout, otherwise it is impossible to differentiate a period of 0 sales due to a stockout from a period with 0 sales due to...simply no sales. So, items with "Sporadic", "Slow", and "Dead" usage patterns will not have lost sales.
- The Item Must Have Forecast History - In order to figure out what your sales should have been, we must have some reference point for what the expected amount was. StockIQ first looks for what your forecast sales were in this period, and uses this to come up with what your daily run-rate of sales would have been if this forecast were accurate (and you had inventory to support that forecast).
- ...Or Demand History - If this forecast history is not present, StockIQ will look for the average rate of sales in periods leading up to the stockout by looking for a complete period (e.g. a previous month) with no stockouts, and use that to set "here's what your average usage or run rate was before you started running out of this", as a fallback to a forecast number. This is less preferred than a forecast history number in case there is some seasonality to your forecast, but for relatively stable items, it can provide a good replacement for a run rate from which to calculate estimated lost sales.
Process
The idea of Lost Sales is trying to estimate or capture sales that you lost due to insufficient inventory levels. StockIQ can estimate lost sales for you based on on hand, forecast history, and if no forecast history is present, demand history.
Since we know the periods of time that you’re out of stock, we can measure if that correlates with a reduction in demand, and if so, we can “fill in” the gaps to help inform the forecast algorithm of what might have been, had you not run out of that product.
This helps improve future forecast accuracy, and keeps it from happening again in the future.
Detection of lost sales is turned on by default, and appears as a purple area in the Forecast Manager Summary chart.
You can optionally have the lost sales also applied to your forecasts as well, using the Lost Sales Settings Tab in Forecast Manager, or by setting the default in the Default Lost Sales settings under the System Configuration.
Calculation Steps:
- Convert the forecast for the period of the stockout into a daily demand amount.
- a. If there is no forecast present - calculate an average historical, non-stockout daily usage by referring up to 6 months in the past for other periods without outages and events.
- Examine number of workdays stocked out.
- Multiply the number of days out by the average daily demand which then equals Lost Sales for that stockout and date periods it touches.