Inventory Optimisation

Inventory Optimisation is the process of maintaining inventory levels in a way that provides high levels of customer service, while also minimising supply chain costs such as costs of inventory and warehousing.

Many Inventory Optimisation projects focus on demand forecasting as a way to reduce inventory levels and improve availability. It is true that a more accurate demand forecast enables companies to hold less inventory while maintaining high levels of customer service. However, it is equally true that the vast majority of medium-large companies already have machine learning-driven demand forecasts. While there is usually room to further improve machine learning demand forecasting (see our article on Demand Forecasting), what else can be done to optimise raw material and finished product inventory? 

As much time and energy as you put into your demand forecasting, it will never be 100% accurate. Hopefully that is not bursting anyone’s bubble! But how do we use an understanding of how accurate our demand forecast is to better optimise inventory levels?

Imagine a scenario where your demand forecast accuracy for products in Category A is good, while accuracy for products in Category B has a much wider variation in accuracy. You should look to hold more inventory for Category B products, to help meet orders even when demand is well above the forecast. 

In a similar vein to forecast accuracy, if you had high supply reliability for one category, but a number of supply disruptions or otherwise being let down by a supplier for another category, it would make sense to hold more inventory in the later category to help meet demand even when there are supply disruptions.

By using statistical analysis or machine learning of forecast accuracy and supply reliability, it is possible to predict the likelihood and impact of future forecast errors and supply disruptions. This can be translated into the probability of different scenarios occurring in the future to help you strategically account for those in your plans.

By combining different factors together, including the costs of holding inventory, the risks of forecast errors and supply disruptions, it is possible to get to an overall view of the total cost of inventory based on holding different amounts of safety stock. This in turn can be used to identify an optimum safety stock target.

  Similar principles can be applied to calculate optimum safety stock targets, optimum production quantities and optimum raw material sourcing.

  Better still, in our recent project with a leading CPG food firm, we used the latest data on forecast accuracy, material and product pricing, demand run-rate and a barometer metric of overall supply reliability to create dynamic inventory targets each week. This was transformational in increasing responsiveness to changes in internal and external factors, making their planning much more agile and increasing security of supply.

Thinking back to the forecast accuracy discussion from earlier, when there is a system shock such as a natural disaster, war or pandemic, demand forecasts can take time to respond and adjust. However, while that happens, dynamic safety stock targets will enable the material planners to quickly re-optimise inventory levels based on the current reality. This includes responding to lower forecast accuracy and increased supply disruption risks to hold more inventory, all driven by data. And as things return to normal, your safety stock targets automatically adjust to reduce your inventory levels again.

With inventory planning normally leveraging tools such as Kinaxis RapidResponse or BlueYonder Luminate, how can you leverage these safety stock targets to improve your inventory planning process?

To get the most out of inventory planning tools, many leading CPGs are augmenting them with custom intelligence as a way to get a competitive advantage. Kinaxis call this a Best of Breed (BoB) approach. For the leading CPG food firm, we integrated their weekly dynamic inventory targets into their Kinaxis RapidReponse cloud instance, which highlighted to the material planners when their planned inventory dropped below the safety stock targets. This enabled them to strategically respond to changes in demand, prices, forecast accuracy and demand security. The next time a natural disaster, pandemic or other market shock happens, the supply chain plans will rapidly respond to maintain supply and reduce costs. This is what Inventory Optimisation should look like.

  Purpose AI is an AI and Analytics Consultancy delivering data-driven solutions for complex Supply Chain and Commercial decision making. We work across the Consumer Products and Retail Industries and have worked with many Top 20 CPGs and Retailers in Europe and North America. Our USP is that we have developed industrialised solutions to common industry problems, reducing the time and costs to deliver value. But unlike SaaS providers, there is no licensing costs, we customise our solutions to your requirements and embed a strong business process to ensure they deliver value.