Maturing supply chain analytics for optimal inventory management

19 July 2021 3 min. read

A growing number of supply chain organisations are embracing prescriptive analytics – the most mature form of analytics – as a means to optimise their inventory management. Luc Baetens, a partner at business transformation consultancy Möbius, outlines the path supply chain can take towards prescriptive inventory analytics.

Adopting data-driven insights can be a gamechanger for inventory management. Using data, companies can come to the optimal trade-off between the often different interests of key stakeholders, such as sales, finance, manufacturing and supply chain. This is where data analytics come in, providing both the oversight as well as the details to take the best decisions, with speed.

To get the most out of inventory analytics, it is recommended to move one step at a time. In every step, people learn and get used to new techniques. And every new technique brings further gains that pay for the additional investment.

Supply chain inventory analyticsThe descriptive analytics step creates visibility on inventory levels across the supply chain. End-to-end reports show the position of the inventory in the supply chain. They can reveal how raw material, intermediate product, and finished goods stocks cover sales. People learn to think in value and not only in units. And they start to see how inventories on different levels add up to serve the same demand.

The utility of descriptive analytics is that it combines information to make it quick and easy to consult. But it does not help the planners to find the items that require attention.

Diagnostic analytics helps to pinpoint shortages or overstocks and understand their causes. It assists planners to go straight to the stock-keeping units that need action. Is there more inventory than required? What can we do about it? How can we avoid it in the future? Are we short? Will that create a problem?

Diagnostic inventory analysis will also encourage the management to look past total inventory value. By showing the quality of the inventory, it shows the difference between healthy and unhealthy stocks. To make this possible, the company needs detailed inventory targets, but those targets do not need to be precise. Simple techniques are usually sufficient to determine the proper minimum and maximum stocks.

Predictive analytics will warn about possible shortages and risks of slow-moving and obsolete stocks. By coupling data on inventories and demand, algorithms can alert planners if stocks risk passing their shelf life. They can create warnings if inventories reach critically low levels and lead to lost sales very soon.

For these applications, analytics needs detailed inventory information (on a batch level, for example), reliable master data, and demand information. The logic becomes more complex, demanding more knowledge of the supply chain people. That is why it is better to start easy.

Finally, prescriptive analytics can suggest actions to improve inventory efficiency. These actions should not simply be “launch a replenishment,” of course. That kind of proposal should come from the planning system. Inventory analytics should suggest solutions that a basic planning system cannot, such as “move excess stock from location A to location B to reduce the risk of obsolescence.” 

Prescriptive analytics can also look at the inventory parameters themselves. By analysing inventory, supply, and demand data, it can propose changes to safety stocks, lot sizes, lead times, or other planning parameters.