I come from Ireland, a country better known for its golf, Gaelic, and Guinness than for its distributors. 

So when Matt Rojas, my dormmate at Stanford, started telling me about his father’s work in distribution, I was fascinated. It was an industry I had never thought about, yet the more I learned, the more it became clear: the U.S. economy quietly depends on distributors every single day.

Matt and I decided to dig deeper. What started as curiosity turned into our MBA research focus. Over the course of a year, we interviewed more than 200 distribution executives. 

We ended every conversation the same way: What is the biggest challenge you’re facing right now?

The answer was remarkably consistent. Forecasting

Distribution has a forecasting problem.

Industrial distributors have a forecasting problem

For industrial distributors, inventory isn’t just part of the business, it is the business.

It’s the largest asset on the balance sheet. It’s what customers rely on. And it’s often the clearest signal of whether a company is running well or falling behind. But forecasting how much inventory to hold is incredibly difficult.

Buy too little, and you face stockouts, missed sales, and frustrated customers who may not come back. Buy too much, and you tie up cash in products that don’t move, slowly filling your warehouse with dead stock. 

And this isn’t a one-time decision. It’s a constant balancing act across thousands of SKUs, every week. How much to buy. When to buy it. Where to place it. All while demand shifts and information remains imperfect.

It’s no surprise that many industrial distributors end up with both stockouts and excess inventory at the same time. The worst of both worlds. 

Why this problem exists:

This isn’t happening because buyers aren’t doing their jobs. If anything, they’re working harder than ever.

The real issue is that the environment around them is changing faster than ever before. Demand is more volatile. Lead times shift unpredictably. 

As businesses grow, this complexity doesn’t just increase, it compounds. What once could be managed through experience and intuition becomes increasingly difficult to control, even for the best operators.

In the industrial vertical specifically, the problem is made worse by having to manage a very large catalogue of SKUs, a meaningful portion of demand being project-driven, and the fact customer ordering behavior doesn’t always mirror actual consumption.

Why existing solutions fail

Despite all this, most inventory decisions are still guided by rules, averages, and gut feel. These approaches were built for a more stable world. They assume the future will look like the past. But that assumption no longer holds.

When demand shifts quickly, lead times move unpredictably, and assortments expand constantly, static systems break down. They react too slowly, and they miss what’s actually changing.

Inventory is too important to be managed this way.

Dinner with our first customer

The Lantern Approach

During our research, Matt and I met our third co-founder, Jay Shitole. Jay had spent five years leading Walmart’s Applied AI Lab for demand forecasting. He showed us a fundamentally different way to approach the problem.

As demand, lead times, and assortments change, inventory decisions need to adapt in real time. That requires moving from reactive decision-making to something proactive, to something that continuously learns and improves.

That something is called machine learning. 

By leveraging all available signals, this technology can uncover true demand patterns and align purchasing to actual demand, not outdated assumptions. 

It can use information across the entire catalog to make better predictions for long-tail items with limited history. 

And it can separate one-off anomalies, like project-driven spikes, from the early signs of sustained demand shifts.

In a world that doesn’t stand still, inventory decisions shouldn’t either. This insight became the foundation of Lantern and how we can help industrial distributors forecast better.

We felt we had a once-in-a-lifetime opportunity to bring a new approach to an industry that needed better tools.

So we built Lantern to help distributors:

• Forecast demand
• Forecast lead times
• Optimize order frequency
• Rank products
• Assign safety stock to hit service level targets

Lantern guides purchasing decisions so distributors have the right goods, in the right place, at the right time.

The results are tangible:

• Fewer stockouts and happier customers
• Less dead inventory and more available cash
• Hours of manual work saved

On average, our industrial customers reduce inventory by 30–40%, improve turns by 2, and increase fill rates by 10%.

Thank you, distributors.

The impact of our product matters, but what’s meant the most to us is the community we’ve found in the businesses we work with.

Our team comes from very different communities across Ireland, America, and India, each with its own culture, pace, and way of doing things. But underneath those differences, we’ve been shaped by the same values: hard work, integrity, and pride in doing things well.

What’s been most striking is seeing those same values show up in every distributor we meet. No matter the geography or the business, there’s a shared mindset, a commitment to doing the job right and taking responsibility for the outcome.

Building alongside industrial distributors has been incredibly rewarding. And we’re just getting started. 

To any industrial distributor reading this, it’s an honor to serve you.

My email is conor@lanternhq.com. Please reach out, I would love to meet you. If you are attending ISA ’26, let’s connect.