Email
sinha.rk.dr@gmail.com
Sizing of inventory buffers is a key parameter to improve both availability as well as freshness.
Most consumer businesses operate in a scenario where consumer tolerance time is much lower than the supply lead time. Hence, we need to keep inventory buffers at various strategic decoupling points to ensure uninterrupted product availability during demand fluctuations. If the buffer size is low, it may lead to stockouts and if the buffer size is high, it would affect product freshness on retail shelves. How do we set the right inventory buffers in a dynamic demand environment?
Theoretically, inventory buffers should be set at the maximum demand during replenishment lead time (RLT). While RLT is easier to calculate, companies typically look at the past data to estimate maximum demand during RLT. Is it the right way?
Let’s take the case of a retailer with RLT of 3 days. If we look at the past data, maximum demand is likely to be during the weekends, say Friday-Saturday-Sunday. Setting inventory buffer at this level would mean that we carry much higher stock during the other weekdays when demand is lower. Shouldn’t we set buffer for Monday-Tuesday-Wednesday separately considering the weekday demand?
That’s where continuous demand sensing at the most granular level gains importance. Using AI/ML models, we can predict daily demand of various items much better and see the patterns changing for each day of the week. Buffers can then be set at the expected forward RLT demand and top it up to account for the prediction errors at that granularity.
Implementing this methodology across various implementations has resulted in higher availability with lower inventory!