Note to my readers: This is part 2 of a series.


In my nearly 30 years of helping make supply chains more profitable, one of the toughest challenges has always been how to set stock levels for items with sporadic demand. And, of course, we must remember that the most sporadic of sporadic demand occurs at the retail site level.


Sporadic demand at the retail outlet has additional challenges accompanying it:

  • Usually a relatively small stock on the shop floor for such items
  • Frequently no back-stock on site
  • No future demand visibility (sales are off-the-shelf, not ordered ahead for pick-up days later)


What is sporadic demand?

Sporadic demand goes by other names, as well. Sometimes known as intermittent demand, sporadic demand might be exemplified by an item that fits one of the following scenarios:

  • Average Daily Utilization (ADU) is about 2 units; however, it sold only 41 times over the last  year
  • ADU is about 1.15 units, but variability is very high (perhaps selling large quantities one day, then not selling any for several days in succession); with a very short lead-time


In such cases, traditional methods for calculating stock levels and safety stock have never yielded good results in terms of ROI (return on investment) for the supply chains involved. Either customer service levels were high, but relative stock levels were so high that profits were consumed by the costs of carrying the required inventories; or customer service levels were quite low, and profits were low due to lost sales, lost customers, or expediting and excess shipping costs trying to satisfy the sporadic demand.


In “Solving the Sporadic Demand Problem – Part 1,” we covered the first method that can contribute appropriate buffer (stock) sizing for items with sporadic demand—especially in a POS (point of sale) situation at a retail location. This article will cover an alternate method.


Demand driven supply chain thinkers to the rescue

This second method was developed by Patrick Rigoni of SmartChain, LLC and introduces a new factor in DDMRP buffer calculations. We (that is to say, I) have given this new factor the name: Sporadic Demand Factor (SDF).


The SDF is calculated as square-root of average number of days between demand for an item with sporadic demand. In the example shown in the accompanying figure, there were 41 days with demand for this SKU-location in the last year. When we divide 365 days (the lookback period) by 41 days with demand, we get 8.9 average days between demand (DBD). The square-root of 8.9 is 2.983, or rounded to 3.0.

Figures showing calculations


As shown in the accompanying figure, the parameters surrounding the SKU (stock-keeping unit) are as follows:

  • Lead-time: 7 days
  • DDMRP Lead-time factor applied: 50%
  • DDMRP Variability factor applied: 33% (Note: When using SDF calculations, very high demand variability factors are not required.)
  • Minimum Order Quantity (MoQ): None
  • Visibility: None (no spike level or spike horizon available)
  • Average Daily Usage (ADU): 2


Using these factors and the DDMRP standard of creating three standard zones (RED, YELLOW and GREEN), the factors are applied in the following manner:

  • RED ZONE: ADU times Lead-time times Lead-time Factor times (1 plus Variability Factor) times SDF (Sporadic Demand Factor)
  • YELLOW ZONE: ADU times Lead-time
  • GREEN ZONE: the maximum of
    • MoQ
    • Lead-time times ADU times Lead-time factor times SDF


The results of these calculations (as shown in the figure above) are:

  • RED ZONE = 28 units
  • YELLOW ZONE = 14 units
  • GREEN ZONE = 21 units


This results in an average on-hand quantity, calculated as RED ZONE plus (GREEN ZONE divided by 2) of about 38.5 units (or about 19.3 days’ on-hand). The 19.3 days is just over two demand cycles, or 2 times the average number of days between demand (8.9), in this case.


Amazing results

This method, too, proved to be extremely effective in protecting flow when run in simulations against historic activity for such SKU-locations. It certainly demonstrates the value of such an innovative approach in calculating stocking levels on SKUs with sporadic demand.


An additional benefit

Since the square-root of any number tends to 1 rapidly as the number itself tends to 1, the SDF may be used in the calculation of buffer sizes whether or not the SKU-location is subject to sporadic demand. For example:


Average Days Between Demand

Sporadic Demand Factor (SDF)











1 (daily demand)



Now it’s your turn

As is said at the outset, SKU-locations with sporadic demand have always been a challenge for setting stock (buffer) levels that produce high ROI and adequately protect FLOW. These two new options show a lot of promise, I believe.

How are you handling stock buffer calculations for items where you experience intermittent or sporadic demand? How are your methods working?


Let us know by leaving your comments below, or by contacting us directly, if you prefer. Thank you.



Follow us on Twitter: @RKLeSolutions and @RDCushing
LIKE us on Facebook: RKL eSolutions and GeeWhiz2ROI