THE PROFIT LAB // 10 Steps to Finding Profit in Localized Replenishment
9. YOU CAN’T ALWAYS GET WHAT YOU WANT //
Step #9: Scarcity and abundance
Retail is a volatile environment. Change is the norm. One area where this is often apparent is in the receipt of orders. Whether due to demand changes, supplier error, poor initial allocation decisions, shipping damage or other constantly evolving activity, it is generally more common to have too much or too little product than it is to have precisely what you need. Since this will have an impact on how you are fulfilling demand, it is critical to have a strategy for how to deal with it.
Simple methods will just scale by spreading the overage or underage across the contribution base of the target inventory. This has proved to be unreliable because it frequently fulfills high volume locations and completely eliminates low volume locations. As this became known as an undependable practice, a number of more sophisticated strategies evolved such as using need percent (which evaluates the importance of need relative to total need). The results are often complicated by packs or prepacks which traditionally rely on the use of somewhat arbitrary rounding rules to get to a result. This again will typically lean toward prioritizing high volume locations, which will come closer to rounding cutoffs.
To make matters more difficult, when we are dealing with 100’s or possibly 1000’s of locations, it is common that many of them will have exactly the same targets or remaining need. How to break the tie? It is eye opening to see how many traditional environments end up giving the next unit to the location with the lowest store number in these situations. It would be tough to get less optimal than that!
What can you do now?
First off, if you have a situation where a tie goes to the lowest numbered location, find a way to make a change. Any change, even random selection, would be an improvement.
If you are using a simple spread in areas of shortage, try to get a more representative base in place. This can be done a number of ways depending on how your targets are being derived. For example, if you are using model stocks, consider reworking models to reflect values that are more aligned with your constrained inventory position.
If you have control over how the shortage/overage itself is being divvied out then consider the use of percent needed rather than straight unit needed. For example:
Location A needs 4 units to fulfill its target of 16
Location B needs 3 units to fulfill its target of 8
Simple logic will give the next available unit to Location A. Location B, however, has a higher percent needed since the three units it needs represent 38% of its target whereas Location A’s need is only 25%. Therefore, the importance of the next unit is greater for Location B than it is for Location A.
While not a perfect solution, it can be a simple way to move in the right direction.
What should you consider in the future?
As we look at solutions that produce a more ideal result, we need to carry forward some of the concepts discussed earlier in the series to solving this problem. If we know what your objective is (profit, revenue, etc.), we can utilize that to assist in making the shortage or overage determination by location.
First, we can evaluate the importance of each unit of need. Rather than just relying on the target units, or even the slightly better percent need, it becomes important to understand the variables that lead to that target. Two locations with a need of six units may have gotten there for very different reasons. One location may be very volatile selling anywhere from one to 10 units in the period, while another may sell five, six or seven very consistently. These differences represent very different prioritization of each needed unit. Better, modern, functionally integrated replenishment solutions can look beyond the unit need target into components of the forecast including volatility, seasonality, lifecycle, etc. and use those to properly weigh the importance of each unit of need.
Learn more about establishing targets here >>
The best solutions then take that understanding and convert it onto the value of each unit of need as it relates to your defined objective. For example, what is the likelihood that sending an additional unit to one location will result in greater profitability vs. sending it to another location?
Carrying the logic down to individual units can net out an even better answer when dealing with packs. This allows the system to determine not only the value of the needed units, but the cost of unneeded units – resulting in, for instance, sending packs to the location that can sell more of the units contained in the pack profitably than another location.
Images courtesy of Evgeni Dinec and creationzs/FreeDigitalPhotos.net
Stay tuned for the last piece in this series out next week.
Sign up for the Profit Lab Blog here >>