Organizations that correctly and thoroughly apply inventory optimization techniques can typically reduce inventories by up to 50% while improving or maintaining service levels. 

And yet the sad truth is that even the organizations who think they are already optimizing their inventories are almost certainly not. “Inventory optimization” is too often used to describe any effort to improve inventory management, while true optimization is barely even understood, let alone applied. 

Sometimes the assumption is that improved inventory visibility alone will deliver optimization. Or sometimes the feeling is that because a “model” is being used to define safety stock levels, that inventories are being optimized, regardless of whether or not the model is even appropriate for the type of planning being used. 

This is not to say that enhanced inventory visibility or safety stock models are not good things. They are. It is just to say that there is still a huge gap between current practice and optimal inventory levels at the vast majority of organizations. 

Why is Inventory Optimization Such a Neglected Opportunity?  

We find there are three main root causes. 

The first is simply unawareness of the potential. For many organizations, inventory management is simply about considering what is to hand, estimating what is needed, and planning to have enough, plus a bit more to be safe. With this type of approach, real-time inventory visibility is extremely valuable. But this way of thinking leads organizations to over-estimate the importance of forecasting, regardless of their forecast quality, and to ignore the other key levers in inventory optimization. 

The second is complexity. Supply chains are complex networks, with many variables and almost constant change. Accounting for this complexity, understanding the sensitivities of optimization models and knowing how best to apply them is no small task. 

The third is supply chain technology. Many tools promise inventory optimization, but don’t really apply optimization techniques and are excessively reliant on heuristics. Others do follow optimization methods, but with so many assumptions and simplifications that their usefulness is very limited. While the best tools on the market do offer a range of optimization functionality, they are usually poorly documented, very difficult to understand or use and, like all these tools, are severely limited by underlying data quality.  

The net result is that a lot of money is spent on “inventory optimization” technology without much optimization being achieved. Planners often switch off or overwrite optimization engines, preferring their own Excel models, which at least they understand and can rely on to deliver predictable results. Or even worse, perhaps, tools are manipulated to deliver the results that were expected in their absence. 

So What is to be Done?  

While the barriers to total inventory optimization are formidable, the good news is that you can achieve most of the benefits through a number of pragmatic steps that will lay the foundation for continuous improvement over many years. 

The first step is to focus on utilizing inventory optimization techniques in a decision-support system. You can avoid building optimization into the heart of the planning process itself, which leaves you in the situation of needing to accept or reject your analytical outputs wholesale. Instead, running advanced inventory analytics alongside your live planning system allows you to identify and address the biggest opportunities first. That also allows you to model different scenarios, and mitigate data quality problems.  

Advanced inventory analytics can consider a range of statistical considerations (such as demand and lead-time variability) that aren’t obvious to humans. But it needs to do so in a way that complements planning processes without interfering with them — it must provide planners with easy-to-follow values, a clearly documented basis for them, and the ability to segment, prioritize and complement them with information known by the planners but not by the system. 

The second step is to put equal focus on the people and processes involved. While technology is essential for the heavy statistical work involved in optimization calculations, it is useless if it cannot successfully be adopted by planning teams in their daily work. You need a number of optimization specialists  internally, and a plan for training and developing planning teams. Organizations need to combine the flexible knowledge and experience of human planners with the best analytical insights. By utilizing advanced inventory analytics in decision support, businesses gain valuable insights into the different variables that impact inventory management. By improving master data quality, automation and the use of existing inventory planning tools can be increased and improved over time. 

Thirdly, do not underestimate the wider organizational barriers to inventory optimization. If incentives are mis-aligned (think of sales or production teams who have no incentive to minimize inventory excesses, or of procurement teams who are disproportionately measured on acquisition cost), then you will struggle to make progress. The whole organization needs to pull in the same direction to optimize inventory.  

But overall, do not lose sight of the fact that inventory optimization is almost certainly a major untapped opportunity for your organization. The utilization of advanced analytical techniques allows businesses to better understand how much inventory they really need and what key levers to influence in order to achieve it.  

Ask yourself how far from optimal your current inventory levels are. If you don’t know the answer in percentage terms, or if you resort to top-down measures like DIO benchmarks or arbitrary cover targets, then you will have significant improvement potential. 

Ultimately, a strategic business focus on inventory optimization allows companies to align their overall supply chain set up with customer needs and other market dynamics, enhancing overall operational efficiency and customer satisfaction.  

Matthew Bardell is managing director at nVentic.