Recently while working with a DemandCaster client, we were evaluating the safety stock levels of numerous items that appeared to be excessive relative to their historical demand. Upon running through each SKU, we identified that each of these items had significant intermittent spikes in their demand history that in turn mathematically require higher levels of safety stock to "assure" they always have enough on hand.  Now we all know that safety stock is rarely the answer to improving service levels, so we dug in to see if a pattern would appear.

DemandCaster aggregates demand and also has a feature that allows the user to drill down into the data to see the source.  In doing so we learned that these intermittent spikes were consistently being generated by a small yet consistent group of customers.

To say I was excited would be an understatement. What was once an unmanagable garble of data to the client suddenly fell into order and became very manageable resulting in the proverbial A-HA!

The action was set: If we can work with these small group of customer to change their purchasing patterns, we can in turn eliminate the volatility, which reduces our inventory investment and our overall exposure! The question moved from why do we have so much inventory to why do these three customers purchase in such large and spotty quantities. We moved from reacting to demand to proactively working with these customer to improve not only our inventory cost but theirs as well.   

Moving forward, the key to managing not only the three customers idenitified, but others is to manage the atypical demand right at the front end in customer service.  This, alas, is a topic for another posting.