Latest / The Operator's Manual with Fexingo: Daily Decisions, Tactical Wins, and Business Execution / How a Single Kanban Card Cut Inventory 40 Percent
Transcript
- Lucas: There's a scene in the documentary 'The Corporation' where a Toyota plant manager holds up a single metal card — about the size of a playing card — and says this piece of paper is worth more than most companies' entire ERP system. Luna: I love that scene because you can see the journalist doesn't believe him at first. Lucas: Right. He looks at this little card and thinks, 'You're telling me this replaced an entire inventory management software suite?' But the answer is yes — when it's used right. And today I want to talk about one company that did use it right: a mid-sized aerospace parts manufacturer in Ohio, call them AeroPrecision Systems. They had a problem with a single high-volume fastener — a titanium bolt used in landing gear assemblies. They were carrying 12 weeks of inventory on it, and still stocking out about once a month. Luna: Twelve weeks of inventory for a bolt, and still running out. That's the classic inventory paradox — the more you hold, the more you need. Lucas: Exactly. Their purchasing manager, a guy named Dave, would order in batches of 10,000 because the unit price was lower. But the bolt came in lots from the supplier, and the production line only used about 200 per day. So they'd get a pallet of boxes, stack them in the cage, and by week eight they'd realize they were down to two weeks of stock, panic-order another 10,000, and the cycle repeated. Luna: And the carrying cost on 12 weeks of titanium bolts — titanium's not cheap. At what, maybe $4 a bolt? That's $40,000 sitting on the shelf. Lucas: Higher. Aerospace-grade Ti-6Al-4V, about $8.50 per bolt at the time. So more like $85,000 in that one SKU alone. But here's the thing — the real cost wasn't the inventory value. It was the space, the handling, the risk of obsolescence when engineering changed the spec. And the stockouts were driving emergency expedite fees — $1,500 per overnight shipment from the supplier. Dave calculated his team was spending 15 hours a week just managing that one fastener. Luna: So they went to a kanban system. But not the whole plant — just this one bolt. Lucas: One card. That's the story. They implemented a single two-bin kanban loop. The supplier — a specialty fastener company in Cleveland — agreed to hold a buffer of 3,000 bolts at their own warehouse. On the factory floor, AeroPrecision kept two bins: each held 500 bolts. When the first bin emptied, the production operator would take a laminated card from the back of the bin and scan it — which sent an automatic signal to the supplier to replenish that bin. The second bin had 500 bolts to cover the lead time. Luna: So the total on-site inventory dropped from 12,000 bolts to 1,000 — a 92 percent reduction just in physical stock. But you said 40 percent earlier. Lucas: That's because they also had to account for the supplier's buffer. The 3,000 bolts at the supplier were still in the system — just not on AeroPrecision's balance sheet. So total system inventory fell from 12,000 to 4,000 — 8,000 bolts eliminated, or about 67 percent. But their on-site reduction was indeed 92 percent. And the working capital freed up was $68,000 from that one bolt — plus the elimination of expedite fees. Over a year, about $240,000 in savings and freed cash. Luna: And they scaled it. I read the case study — they eventually put 47 SKUs on kanban, freed $2.3 million in working capital, and cut stockouts by 60 percent across the board. Lucas: Which is remarkable, but here's the part that I think is more interesting than the numbers: how they got there. Because Dave told me the hardest part wasn't the supplier negotiation or the bin setup. It was convincing the production manager that having only 500 bolts in the second bin was safe. The manager had been burned by stockouts for years, and his instinct was to hold more. So Dave had to show him the math: if the supplier's lead time is five days, and daily usage is 200, the second bin of 500 gives you two and a half days of buffer. The supplier actually delivered in three days on average. So the risk of a stockout was essentially zero — unless the supplier failed. Luna: But the supplier did fail one time — there was a quality hold on a heat-treat batch, and they couldn't ship for a week. Lucas: Right. And AeroPrecision had to air-freight from an alternative source. But here's the kicker: under the old system, they would have air-freighted anyway, because their buffer was 12 weeks but their ordering pattern was chaotic. With the kanban system, the supplier's failure was a one-off event, and they had a contingency plan for exactly that scenario — a pre-qualified secondary source with a standing agreement. The cost of that one air-freight was $8,000. Under the old system, they were spending nearly $20,000 a year on expedites for that bolt alone. So even with the hiccup, they came out ahead. Luna: It sounds simple, but I think that's the deceptive thing about kanban. People hear the story and think, 'We just need cards.' But the discipline is in the sizing — the bin size, the lead time, the reorder point. And in the trust between the supplier and the customer. Lucas: That's the part that usually breaks. I've seen companies try kanban and fail because they skip the lead time calculation. They guess the bin size. Or they don't really commit to the pull system — someone on the floor sees an empty bin and says, 'I'll just order 2,000 to be safe,' and suddenly you have five bins and the system falls apart. The whole point is that you cannot over-order. The card is the authority. If you don't respect the card, you don't have a kanban system — you have a suggestion box. Luna: And that's where Toyota's culture is so different. In a Toyota plant, if you see a bin with two cards, you stop the line. Because that means someone broke the rules. Lucas: Exactly. The kanban system is a self-regulating mechanism, but only if everyone treats it as a rule, not a guideline. AeroPrecision's Dave told me the first month was a constant battle — operators would see the bin getting low and start hoarding bolts in their toolboxes. They had to physically remove all other storage locations and make the two bins the only source. That forced compliance. And within three months, no one questioned it, because the stockouts had stopped and the work of managing inventory had essentially vanished. Luna: I think there's a broader lesson here about simplicity. We've covered a lot of episodes — checklists, rounding, one question — and the pattern is always the same: the most elegant solution is often the one that removes the most complexity, not adds to it. Lucas: Right. And kanban is really just a visual signal that says 'make more' or 'send more.' It's not a software algorithm. It's a physical card. But it forces you to understand your actual demand and your actual lead time, and it creates a rhythm that is predictable. And if these conversations have moved your work forward in some small way — maybe you've tried a daily check-in or a post-it note system because of something we talked about — it's worth mentioning that listener support is what keeps this show ad-free and independent. Luna: Yeah, we don't run ads, and we don't want to. But a couple of dollars a month from people who find value in the episodes genuinely makes a difference. It covers hosting, transcription, the research time Lucas puts in. Lucas: So if you've gotten something out of these episodes, you can find us at buy me a coffee dot com slash fexingo. That's it — no pressure, no special offers. Just a way to keep the show going. And now back to kanban: the other thing AeroPrecision did was they used the cards as a training tool. New operators learned the flow of the line by following the cards — where they came from, where they went. The card became a map of the production process. Luna: That's smart. Instead of a binder you never read, the card is a living document you touch every day. Lucas: And when they scaled to 47 SKUs, they color-coded the cards by supplier. Blue for Cleveland Fastener, green for Precision Metals, red for a local heat-treat shop. That way, the forklift driver could see at a glance which supplier's bin was empty and consolidate the pickup. They cut internal logistics time by 30 percent just from that color coding. Luna: So the card is not just a signal to replenish — it's a visual management tool, a training tool, and a communication tool. All from a piece of laminated paper. Lucas: And the cost to produce each card? About 15 cents. The whole system for 47 SKUs cost less than $10 to set up — plus the bins, which were just repurposed plastic totes they already had. That's the part that still blows my mind: a $10 investment that freed $2.3 million in working capital. You don't need a consultant or an ERP module. You need a clear understanding of your demand, a reliable supplier, and the discipline to trust the card. Luna: But I imagine the resistance to simplicity is strong. People want the big software solution because it feels like you're doing something. Lucas: It's a status thing. 'We're implementing a kanban system' sounds like process improvement. 'We're using index cards' sounds like a grade-school project. But the results don't care about the optics. And I think that's the final lesson from AeroPrecision: start with one thing. One fastener, one card, one supplier. Prove it works, then scale. Don't try to kanban the whole plant in a weekend. Luna: Because if you do, you'll end up with 200 cards you don't understand and a warehouse full of bins that aren't really pulling anything. Lucas: Exactly. Dave told me they almost made that mistake. Their CEO wanted to roll out kanban across all 2,000 SKUs after the first month. Dave said no — he insisted on running the one sku pilot for 90 days, collecting data, and only then expanding. That discipline probably saved the whole initiative. Luna: So the takeaway: pick one high-volume, high-pain item. Calculate your actual lead time and actual daily usage. Size your bins to cover lead time plus a small buffer. Then commit to the card as the only authority. And don't scale until the pilot is boring. Lucas: Boring is the goal. A kanban system that works is one you don't think about. The bins cycle, the cards flow, the line never stops. That's the quiet magic of a well-designed pull system. And it starts with a single card.