Latest / The Operator's Manual with Fexingo: Daily Decisions, Tactical Wins, and Business Execution / How One Coffee Chain Cut Waste 40 Percent With a Single Metric
Transcript
- Lucas: There's a regional coffee chain in the Pacific Northwest — about forty locations — and last year they figured out they were throwing away around three hundred thousand dollars worth of food annually. Pastries, breakfast sandwiches, yogurt parfaits. Stuff that looked fine but had just passed the freshness window. Luna: And that's not unusual, right? Most food service places accept a certain amount of waste as the cost of having enough variety. Lucas: Exactly. But what this chain did was really specific. Instead of tackling everything at once, they picked one metric to focus on: production waste per customer. That's the dollar value of items they made but didn't sell, divided by the number of customers that day. Luna: So not total waste, but waste per person. That's interesting because it normalizes for traffic. Lucas: Right. A busy store might throw away more in absolute terms, but if waste per customer is low, that's actually efficient. The chain looked at their data and found that on a typical weekday morning, they were producing about twelve percent more than they sold. That buffer accounted for almost all the waste. Luna: Twelve percent overproduction sounds small, but on three hundred thousand dollars of waste, that's the whole thing. Lucas: Exactly. So they asked: what if we trimmed the buffer to five percent? And then they had to get really tactical about how to do it without running out of stock during the breakfast rush. Luna: How do you even cut overproduction without annoying customers who show up at 7:30 and want a bacon croissant? Lucas: That's the core of it. Their solution was to shift from a fixed production schedule — make a batch every thirty minutes — to a dynamic one based on real-time traffic data from the point of sale. They installed a simple dashboard in each store that showed the last hour's customer count and the next hour's forecast based on the same day last week. Luna: So they weren't guessing anymore. Lucas: Right. The morning prep team started using a rule: only make what's needed for the next hour, plus a five percent cushion. For example, if the forecast said forty customers in the next hour, and historically forty percent of customers order a breakfast sandwich, they'd prepare sixteen sandwiches, not twenty-two. Luna: Sixteen versus twenty-two — that's a twenty-seven percent cut in production right there. Lucas: And that was just the start. The chain also introduced a 'make to order' window for certain items after 10 a.m., when the rush died down. So the hot breakfast sandwiches — those were still pre-made during peak, but after ten, they'd only cook them when someone ordered. Luna: That shifts the labor cost, though. You need a cook on standby. Lucas: True, but they found that the labor cost was actually lower because they weren't wasting time making food that got thrown away. And the cook could help with cleaning or stocking during slow periods. So the net effect was positive. Luna: What kind of savings did they end up seeing? Lucas: Over six months, they cut total food waste by forty percent. That's about one hundred twenty thousand dollars annually across the chain. And they didn't buy any new software — just a whiteboard and a timer in each store, plus the existing point of sale data they already had. Luna: So the metric itself was the change, not the technology. Lucas: Exactly. And that's what I love about this case. It's not about AI or fancy forecasting. It's about picking the right number to look at and then aligning your process around it. Most operations track waste as a percentage of total production, which tells you how much you're throwing away relative to what you made. But that number can be misleading because if you make a ton, waste percentage might be low even if you're overproducing wildly. Luna: Yeah, because the denominator is production, not demand. Lucas: Right. So by switching to waste per customer, they tied it directly to actual sales. Suddenly, every store manager could see: 'I made ten bucks of waste per hundred customers, but my sister store across town only makes four.' That creates a clear target. Luna: It's almost like a unit economics lens applied to waste. Lucas: Exactly. And once they had that target, they started sharing best practices across stores. One location discovered that if they pre-wrapped sandwiches in paper rather than plastic, they lasted about forty-five minutes longer under the heat lamp before being written off. That small change alone cut waste by another eight percent. Luna: So it's a continuous improvement loop. The metric gives you visibility, and then the team finds small fixes. Lucas: Right. The key was that they didn't try to fix everything at once. They picked one metric, set a target, and then empowered the store teams to figure out how to hit it. And it worked because the metric was simple enough that everyone — from the barista to the regional manager — could understand it. Luna: I think a lot of businesses overcomplicate this. They try to track twenty KPIs and end up with none. Lucas: That's the lesson here. And there's a parallel to something we talk about on this show all the time — the idea that a single, well-chosen metric can transform an operation. It's not about having the most data; it's about having the right data. Luna: And then actually doing something with it. Which brings me to a quick note — a lot of you have asked how we keep this show ad-free and independent. The answer is pretty straightforward: some listeners support us through buy me a coffee dot com slash fexingo. It's not a big campaign, just a small group whose contributions cover the production costs. Lucas: And honestly, that support is what lets us dive into stories like this one — cases that aren't flashy but are genuinely useful in day-to-day work. So if these conversations have moved your work forward in some small way, that's the only reason we mention it. Luna: Yeah. And back to the chain — they're now expanding the metric to their afternoon and evening shifts, where the waste profile is different. More pastry waste, less breakfast sandwich waste. So they're tailoring the approach. Lucas: Right. And they've started sharing their results with other small chains in their region, basically saying: 'Here's the metric that changed our business. Try it.' No proprietary system, no consulting fee. Just a number and a commitment to watch it. Luna: That's the kind of operational thinking that scales far beyond coffee. Lucas: Absolutely. Whether you're running a bakery, a hospital cafeteria, or a manufacturing line, the question is the same: what's the one number that, if improved, would make everything else easier? For this chain, it was production waste per customer. For your business, it might be something else entirely. Luna: And the only way to find out is to dig into your own data — even if it's just a whiteboard and a timer.