If you're in the consumer packaged goods (CPG) industry, you know how hard it is to find your optimal safety stock. But do you know how to make it easier?
What is safety stock?
Safety stock is the extra inventory companies hold to mitigate the risk of stockouts. It's often calculated by looking at the average daily demand, delivery lead times, and reorder point.
However, supply chains are seldom simple. When daily demand varies, lead times are irregular, and reorder points are uncertain, it becomes difficult for inventory planners to find the perfect amount of safety stock. And if they get it wrong, the impact is huge.
Too much safety stock leads to increased costs and waste. Too little, and you risk losing sales, stockouts, and unhappy customers.
Using the Monte Carlo model
But there is another way of working.
Imagine a day in the life of an inventory planner managing thousands of products across the globe. A big challenge they face is ordering inventory wisely when every product has different lead times, which vary even more at the SKU, plant, and vendor levels. And what if one of these products unexpectedly runs out? How quickly can they respond?
The Monte Carlo model – a computer simulation technique that models the probability of different outcomes in situations that are difficult to predict – turns this challenging task into something more manageable. The model can produce a simulated environment, backed by company data and variables, that helps inventory planners find the optimal safety stock.
And it's not just a pipedream; it's a reality that some companies are already living.
An example in the enterprise
With a diversified product portfolio across multiple plants and distribution centers, one of our CPG clients faced major problems in its supply chain, including:
Volatile lead times from suppliers
High demand variability across SKUs and regions
Inaccurate static safety stock formulas
Identifying plant and SKU combinations tying up working capital
Too many decisions made on assumptions instead of data
We brought in our safety stock optimization solution, built on the Monte Carlo model, to help. The model uses historical and forecasted data to simulate thousands of inventory scenarios across every SKU, plant, and vendor combination to find the optimal safety stock. And this wasn't just used for finished goods – the model could also be applied to raw and packing materials.
After some tailored training sessions to support adoption, the company saved $4.3 million in working capital costs and $320,000 in warehouse costs in just six months.
Lessons from the leaders
If you're ready to improve the way you manage safety stock, here are five pieces of advice from our firsthand experience:
Shift from static rules to data-driven decisions: The static rules of traditional safety stock formulas struggle to support today's volatile and complex supply chains. With the data behind the Monte Carlo model, you can account for real-world uncertainty
Turn inventory into a strategic capability: Safety stock shouldn't be viewed as merely a buffer. When it's managed correctly and backed by data, it becomes a strategic way to optimize working capital and service levels
Drive results at a granular level: A high-level, low-touch approach to safety stock simply does not reflect the complexities of many modern businesses. The Monte Carlo model can handle nuance at the SKU, plant, and vendor levels
Build an analytical culture: Working with clients, we always talk about the need for a company culture that supports the adoption of advanced technology. If you invest time in empowering planners with visual tools and scenario testing, this reduces the reliance on tribal knowledge and embeds predictive analytics into everyday planning
Use explainable AI for transparency: Many businesses and employees fear machine learning and AI, mainly because they feel it's shrouded in mystery. Consider using a SHapley Additive exPlanations (SHAP) model for greater explainability to highlight which variables matter most in safety stock decisions
Remember, safety stock is a balancing act. Unfortunately, it's become a task too big for inventory planners to tackle alone. Now is the time to invest in solutions that empower inventory planners to consider every data point to make the right decision at the right time.
