Latest / The Operations Podcast with Fexingo: Process, People, and Profit in Modern Business / How a Machine Tool Builder Cut Lead Times by 80 Percent
Transcript
- Lucas: Luna, I've got a number for you. Sixteen weeks to three weeks. That's the lead-time compression a machine tool builder called Mazak pulled off at its plant in Florence, Kentucky. And I want to talk about how they did it, because it's not a story about buying faster robots. Luna: Sixteen weeks down to three — that's an 80 percent cut. Was this a custom job shop environment where every machine is built to order? Lucas: Exactly right. Mazak makes CNC machine tools — lathes, mills, multi-tasking machines — and every single one is configured to the customer's specs. So it's not like they can build to stock. Before the transformation, an order would sit in queue for weeks, then move through departments in giant batches. Typical lead time was 16 weeks, and on-time delivery was below 60 percent. Not great. Luna: So what was the first thing they changed? Did they go to a cellular layout? Lucas: They did. And this is where it gets interesting. They had a traditional functional layout: all the milling machines in one area, all the lathes in another, assembly in its own building. Parts would travel across the plant, waiting at each step. So they regrouped everything into product-focused cells. Each cell has a mix of machines — milling, turning, drilling — arranged in the sequence the part actually flows through. One cell builds one family of parts from start to finish. Luna: That sounds obvious in retrospect, but I bet the floor space reconfiguration was a nightmare. Did they have to shut down production? Lucas: They phased it in over weekends and planned shutdowns. Took about 18 months to fully roll out 12 cells. But the payoff came fast. Work-in-progress inventory dropped by 60 percent almost immediately, just because parts weren't sitting in queues between departments. But the real bottleneck wasn't the layout — it was the mindset of the machinists. Luna: Ah, the culture piece. I was wondering when we'd get to that. Veteran machinists who've been running the same machine for 20 years probably don't want to hear 'now you'll rotate between three different machines.' Lucas: Exactly. Mazak's workforce averages about 15 years of tenure. These are skilled tradespeople who take pride in their craft. The old system rewarded deep specialization — you owned your machine, you knew its quirks. Cross-training felt like a demotion to some of them. So the operations team had to make the case that flexibility meant job security. If you can run three machines, you're more valuable when order mix shifts. Luna: And that's a hard sell when you're asking someone who's been doing it one way their whole career. How did they actually get buy-in? Lucas: They started with volunteers. A handful of machinists agreed to pilot the cross-training and the new cell layout. Once those cells started showing results — shorter lead times, less rework, less firefighting — the skeptics started paying attention. Within a year, over 90 percent of the shop floor had completed multi-machine certification. And they tied it to a skill-based pay progression: the more machines you're certified on, the higher your base rate. Luna: That's a smart incentive. So after the layout and the cross-training, what was the next piece? Kanban? Lucas: Right. They implemented a pull system for parts. Previously, the production planning department would push work orders to each department based on a forecast. Now, each cell only produces what the downstream cell needs. They use kanban cards — physical cards that signal when to replenish a part. The number of cards in circulation limits the amount of work-in-progress. It's a classic lean technique, but it required a complete rewiring of the IT system and the purchasing department's rhythm. Luna: I imagine the purchasing team was used to ordering in bulk to get volume discounts. Kanban means smaller, more frequent orders — which can look more expensive on a per-unit basis. Lucas: Exactly. They had to shift from thinking about purchase price variance to total cost of ownership. A bulk order of castings might have a lower unit cost, but if those castings sit in inventory for 12 weeks, you're paying carrying costs, obsolescence risk, and floor space. With kanban, they order smaller quantities but much more frequently. Suppliers had to adapt too — Mazak worked with key suppliers to set up milk-run deliveries, sometimes daily. Luna: And the suppliers went along with that? I imagine some of them pushed back. Lucas: Some did. Mazak ended up consolidating its supplier base. They went from about 1,200 suppliers down to 600. The ones that could handle the just-in-time delivery got more volume and longer contracts. The ones that couldn't — or wouldn't — got phased out. It was a tough decision, but it made the supply chain more reliable. And with shorter lead times, they could also give suppliers a more stable forecast, because the order horizon shrunk from 16 weeks to 3 weeks. Luna: Right, because the customer order is firmed up much closer to the build date. So the forecast error drops dramatically. Lucas: Exactly. That's one of the hidden benefits. When your lead time is 16 weeks, you're forecasting demand 4 months out. With 3 weeks, you're forecasting 3 weeks out. Much easier to be accurate. Their on-time delivery went from below 60 percent to over 98 percent. And that's not just a vanity metric — when a customer is waiting for a $500,000 machine tool, every week of delay costs them production revenue. Luna: So what did the financials look like? Was there a measurable ROI on this transformation? Lucas: Mazak doesn't publish a detailed breakdown, but they've said inventory turns went from about 4 times a year to over 12. Work-in-progress dropped by 60 percent. Floor space utilization improved by 30 percent because they didn't need as much space for queues and storage. And they estimate the payback period for the entire reconfiguration — including new machines, cell layout changes, and training — was under 2 years. Not bad for a plant that was already running three shifts. Luna: That's a strong case study. But I wonder — is this replicable for a smaller job shop? Mazak is a big company with deep pockets. What if you're a 20-person shop? Lucas: I think the principles scale down, but the implementation looks different. A small shop doesn't need a multi-year phased rollout. You can pick one product family, set up one cell, and prove the concept in a month. The key is the pull system and cross-training — those are cheap. The expensive part is moving machines. But even there, you can do it gradually. The mindset shift is the same regardless of size: stop thinking in batches, start thinking in flow. Luna: Yeah, and the cultural resistance is probably even more acute in a small shop where the owner has been doing it one way for decades. So how did Mazak sustain the change over time? Did they have any backsliding? Lucas: They did have some. When they hit a demand surge a couple of years after the transformation, some cells started building buffer stock 'just in case.' The kanban system started getting ignored. They had to reinforce the discipline with daily stand-up meetings and visual management boards. The plant manager would walk the floor every morning and look at the kanban status. If a card was missing or a bin was overfilled, they'd stop and ask why. It's that constant vigilance that makes lean stick. Luna: It sounds like the operations team became almost like coaches rather than managers. That's a different skill set. Lucas: Absolutely. And that's something I think a lot of operations transformations underestimate. You can redesign the layout and implement kanban in a few months. But changing how people think about their work — that takes years. Mazak's production manager told me that the hardest part wasn't the machinery; it was convincing a 25-year veteran that letting go of his personal inventory hoard was actually safer for his job. That's a conversation, not a spreadsheet. Luna: So if someone listening is thinking about doing something similar in their own operation, what's the one piece of advice you'd distill from this case? Lucas: Start with one value stream. Don't try to transform the whole plant at once. Pick a product family that represents a significant chunk of revenue, map the current state, identify the biggest source of delay — it's almost always waiting — and then design a future state cell. Run a pilot. Prove the metrics. Then use that success to get buy-in for the next cell. And invest heavily in the cross-training piece early. The machines are the easy part. Luna: That's a good note to end on. I think the Mazak story is a great reminder that operations improvement is as much about people as it is about process. Thanks, Lucas. Lucas: Thanks, Luna. And if you want to read more about Mazak's lean journey, there's a detailed case study from the Lean Enterprise Institute that covers the full timeline and the specific metrics. We'll link to it in the show notes.