A customer requests a product and it’s not in stock. They can’t wait for you to obtain the item so they purchase it elsewhere. You’ve lost a sale. That means you have lost the opportunity to earn a profit, disappointed the customer, and probably put some doubt in his or her mind about your reliability as a supplier. A lost sale is not a good thing.

But a lost sale is made even worse if you don’t use it as an opportunity to evaluate and possibly revise the replenishment parameters for a product – tThat is, to try to prevent future lost sales. Indeed, many companies try to capture all unfulfilled customer requests and add them to the actual usage recorded for a specific inventory period. They believe that by including these lost sales in usage history, future demand forecasts will be adequate to cover the unfulfilled sales or usage experienced during the current inventory period.

But there are some inherent flaws in this practice:

Your Salespeople May Not Consistently and Accurately Record Lost Sales: Most companies do not experience a consistent level of sales activity throughout the day. Often, customer orders (or customers themselves) arrive in spurts. During these hectic periods, salespeople may not have the time to accurately record what customers request. They are too busy processing actual orders.

A Single Request May Be Repeated: A customer calls today and asks for a product that is not in stock. A lost sale is recorded. They call again tomorrow to see if the item has been received, but they talk to a different salesperson. The product is still out of stock and another lost sale is recorded. You have recorded two lost sales for the same request. If lost sales are automatically added to usage, you have captured false or “phantom” demand for the product.

An Area-Wide Shortage May Cause a Greater Than Normal Number of Requests: If your competitors are also out of a product, their customers may call your salespeople looking for the item. If all of these requests are treated as lost sales and added to usage, you may again capture phantom demand for the product.

As part of a comprehensive customer relationship management (CRM) program, it is important to capture lost sales in order to see what customers are not being adequately served. Indeed, your sales manager probably wants to know if your most important customer requested an out-of-stock product five days in a row. But as we’ve seen, adding lost sale quantities to actual usage may not truly reflect the quantities of a product actually needed. The adjusted figures may distort future demand forecasts and result in either additional lost sales or excess inventory.

We’ve found that there is a better way to adjust actual usage for lost sales. This method does not require sales people to accurately record customer requests and protects you from capturing phantom demand. As with other aspects of our inventory management philosophy, we have different recommendations for items with recurring activity (i.e. those that are sold or used on a regular basis) and those with sporadic usage.


Recurring Items

These products are sold on a regular basis. As these are the products customers request most often, it is important to correct for out-of-stock situations in order to ensure a high level of customer service.

  1. Specify whether each of these inventory items will or will not accumulate backorders. If an item will accumulate backorders, customers will wait for you to receive the product. As a result, these items tend not to experience lost sales. If the product will not accumulate backorders, customers will go elsewhere to obtain the item. These are the items that will probably experience lost sales.
  2. At the end of each inventory period (i.e. week, month, four-week period, etc.), record the number of days each product was out of stock.
  3. For items that will not accumulate backorders, multiply the days out-of-stock by the forecast demand/day and adjust monthly usage by this quantity.

For example, say a very competitive popular product was out of stock for six days during the month that just ended and the forecast for the item was seven pieces/day. Forty-two pieces (six days x seven pieces/day) will be added to the month’s actual usage to compensate for possible lost sales.


Sporadic Items

These are items that are not sold on a regular basis and whose replenishment parameters are based on the normal quantity sold or used in one transaction as opposed to the forecast demand for an upcoming inventory period.

  1. Record the number of times a product is out of stock (or its available quantity drops below the normal or average sales quantity).
  2. If the product is out of stock more than one or two times in a six month period, automatically increase the minimum quantity for the product by the normal quantity sold or used in one transaction.

For example, a specific type of seal is used when a certain engine is rebuilt. As two seals are required in the process, a distributor maintains the stock of the item with these replenishment parameters:

Minimum Quantity = 3
Maximum Quantity = 4

When the available quantity falls below the minimum of three pieces (i.e. when two seals are left on the shelf), another two pieces will be ordered to bring the maximum stock quantity to four pieces. But because the item was out of stock two times in a six month period, the system will increase the minimum to five pieces and the maximum to six pieces. Now the item will be reordered when the available quantity falls to four pieces – that is, when there is still enough stock to rebuild two engines.

Note that to avoid an unrealistically large inventory investment, most organizations will tolerate a certain number of stock-outs of non-critical sporadic-usage items. That is why we normally will wait for several stock-outs to occur over a certain period before making the adjustment. However, this rule is not “cast in stone,” and should be adapted to each company’s specific situation and needs.

As I’ve said many times, a good forecast is the foundation of an effective inventory management program. The better the prediction of future demand of a product, the easier it will be to provide a superior level of customer service while minimizing your overall inventory investment. Correcting actual usage for lost sales opportunities is an essential part of the forecasting process. But to be effective, these corrections must predict, as accurately as possible, what would have been sold or used had the item continuously been in stock.