If the past few years have taught distributors anything, it’s this: Uncertainty isn’t a temporary condition, it’s a constant.
In a recent Distribution Strategy Group technology leader panel, Epicor’s Greg Hartunian and ORS Group’s Chandra Subramanian tackled the hard truth about inventory planning today.
It’s about building systems that thrive on volatility.
Here’s what we learned from the conversation.
Missed the webinar? Watch it on-demand.
Accept That Uncertainty Is the Norm
Economic uncertainty has always been a part of business. What’s changed is the frequency and severity of disruptions. As Subramanian said: “Volatility and uncertainty have become the norm now. It’s been the norm since 2008.” Black swan events are not as random or rare, and distributors must evolve their approaches to accommodate higher risk.
Distributors who continue to build inventory models based on stability will be outpaced by competitors who evolve with chaos. The goal now is to develop plans that flex with turbulence instead of breaking under it.
Start Forecasting for Risk
The panelists said that forecasting isn’t about predicting the exact number of anything anymore. Distributors need to understand the range of possibilities and how much risk you’re willing to take.
Many distributors fixate on forecast accuracy, but as Subramanian said: “Accuracy only matters if you’re allocating tomorrow.” For mid- to long-term planning, focusing only on accuracy can be misleading. What you need to understand is the volatility in the forecast – how much it might swing – and then plan accordingly, said Subramanian.
Ditch the Average: Plan by Percentile
Inventory planning based on averages leads to a predictable problem: stockouts. Planning to percentiles means asking questions like: “What inventory level covers 95% of potential demand scenarios?”
That’s a shift in thinking that requires modern AI simulation tools to get right.
Hartunian said: “What we want to focus on at that level is how accurately you can predict the upper percentile. That’s critical for a distributor.”
Use Scenario Planning to Manage Uncertainty
What could happen, and how should we respond? Modern inventory models can simulate thousands or even hundreds of thousands of possible demand scenarios. These simulations generate a range of possible outcomes and help distributors model everything from demand volatility to supplier variability to customer buying behavior, Subramanian said. They can also account for factors that can’t be precisely predicted, like seasonality, economic shifts or weather events but that can still be modeled probabilistically.
The result is a more adaptive, dynamic approach to inventory planning. You’re no longer just forecasting demand, you’re simulating risk. That means asking tough but necessary questions: What if demand exceeds our forecast by 20%? What if lead times suddenly double? What’s the cost of being wrong? What if a real crisis hits?
Simulations help you build a resilient response. By modeling the ripple effects of unexpected scenarios, distributors can understand where their inventory strategies are most vulnerable and adjust before it’s too late.
Then you must evaluate the tradeoffs, Subramanian said. For example, weigh the cost of holding extra inventory against the cost of a missed sale or a lost customer.
Hartunian explained the thought process: “Do I want to cover 99% of the possible demands that might happen? If I do, I might never stock out. But now I’m going to have excess inventory and obsolescence risk.”
Much of this hinges on AI-powered simulation tools that use ERP data such as sales, shipments and lead times, as well as external market indicators to generate predictive ranges.
“It’s hard to tie macroeconomic indicators to specific SKU-level reorder points,” Hartunian said. “But if we simulate thousands of scenarios, we can optimize based on historical patterns and present realities.”
Bring Human Insight Into the Loop
No model, no matter how sophisticated, knows your business better than you do. A model can’t see your competitor’s new warehouse opening across town or predict that a local event will spike demand next month. These kinds of insights only come from the frontlines of the business: the real-world, subjective information often critical to making the right decisions.
Hartunian and Subramanian emphasized the importance of blending machine learning with human intuition. As Subramanian summed it up: “The model spits out indicators. You make the decision.”
This blend of machine precision and human expertise creates a feedback loop that strengthens the inventory planning process, ensuring that data-driven insights are grounded in the practical realities of the market.
Just in Time vs. Just in Case
While just-in-time models were once the gold standard, disruptions of recent years have exposed their vulnerabilities. More leaders are now leaning into a just-in-case mindset to prioritize flexibility and service continuity.
“In reality, a just-in-case model ensures that you’re in control of your own destiny. It may mean that you absorb more inventory costs, but the more you hold, the more you can ship to your customers, and the more buffer you have when the supply chain disruptions come,” said Hartunian.
That doesn’t mean stockpiling excess, he said, but planning more intentionally. How often do you need to fulfill orders from existing stock? In other words, Hartunian said, “how just in case do you want to be, and how just in time do you want to be?” The answer may be different on different items.
AI-based tools can help distributors predict performance and identify where they may be over- or under-stocked, and adjust accordingly.
Subramanian said many distributors still rely on static reorder points and minimum order quantities that don’t reflect real-world dynamics. Smarter AI-based tools can model variables like seasonality, lead time variability and borrowing costs to turn those gut-based decisions into data-backed strategies. They also allow companies to factor in profitability, cash flow and capital costs to strike the right balance between service levels and financial health.
Build Toward Resilience
The panel ended on a forward-looking note: Create systems that flex and adapt.
Distributors can no longer rely on fixed safety stock rules, monthly planning cadences or rearview-looking models. The most successful companies will be most prepared for what they didn’t see coming.
“You cannot predict the black swan, but you can plan for how you’ll respond when it hits,” said Subramanian.
Want more? Watch the webinar on-demand.
