Latest / The Operations Podcast with Fexingo: Process, People, and Profit in Modern Business / How a German Grocer Cut Food Waste by 65 Percent
Transcript
- Lucas: So earlier this week I was looking at a disposal log from a grocery store — not a small store, a proper supermarket — and every night it was throwing away what amounted to a full shopping cart of food. That was one store. Multiply by three hundred. Luna: That image is honestly kind of sickening. But also a massive operations problem. Lucas: Massive. And it's not just the waste itself — it's the cost of disposal, the lost margin, the reputational risk. But there's a German chain called Tegut that decided to do something about it. They set a target: cut food waste by sixty-five percent across their fleet. Luna: Sixty-five percent. That is not a pilot. That's a company-wide overhaul. Lucas: Exactly. And they got there — in about three years. I want to walk through how they did it, because the playbook is really replicable. So Tegut is a regional grocer in Germany, about three hundred stores, mostly in the southwest. They operate in a fairly premium space — think Whole Foods but German. Luna: And culturally, Germany is pretty waste-conscious already. So I imagine they started with a baseline? What were they wasting before? Lucas: They tracked it carefully. Pre-program, Tegut was wasting about two point three percent of total inventory by weight. That's roughly eighteen million euros annually in lost product, not including disposal fees. The breakdown: forty percent was produce, twenty-two percent was dairy, fifteen percent was baked goods, and the rest was a mix of packaged goods, meat, and prepared foods. Luna: So the big wins would have to come in fresh — things with the shortest shelf life. Lucas: Exactly. And the first lever they pulled was dynamic pricing. You've probably seen this in some grocery apps: as a product approaches its sell-by date, the price drops automatically. Tegut implemented that store-wide, but with a twist. Instead of a simple percentage off, they used a machine learning model that factored in the remaining shelf life, the product's historical sales velocity, and even the weather forecast. Luna: Weather — because people buy different things when it's hot or rainy? Lucas: Right. If it's thirty degrees and sunny, you might still buy a salad with two days left because you're eating it tonight. But if it's cloudy and you're shopping for the week, you want a salad that'll last four days. So the model learned to adjust not just the discount, but the timing of the discount. Some items got marked down twenty-four hours before sell-by, others only six hours. Luna: That is really operational. I love this level of granularity. How much did dynamic pricing alone save? Lucas: They estimate it cut waste by about twenty-eight percent. But here's the thing — dynamic pricing only works if customers actually see the discounts. So Tegut made sure the reduced items were placed on a dedicated 'rescue shelf' near the checkout, clearly labeled. They also added a digital shelf tag system that updated the price in real time, so no staff had to physically re-label items. Luna: Okay, that handles the items that still have a few days left. But what about the stuff that's literally expiring that day? You can't sell that. Lucas: That's where the second lever comes in: demand forecasting on the ordering side. Tegut had been ordering based on historical averages plus a manual override by the store manager. The problem was that managers over-ordered to avoid stockouts, especially on weekends. The new system used an AI that predicted demand per store per SKU, with a confidence interval. The order quantity was set to the lower end of that interval — essentially accepting a small risk of stockout to reduce waste. Luna: That must have been a hard sell to store managers. 'Hey, you might run out of milk on Saturday morning.' Lucas: It was. And they didn't mandate it. They ran a pilot in ten stores for six months, comparing waste and stockout rates. The pilot stores saw waste drop by forty percent, while stockouts only increased by half a percent. That was enough to convince the rest. Luna: Half a percent is basically nothing. Customers probably didn't even notice. Lucas: Right. And Tegut also used the forecasting system to adjust orders for holidays and local events — things a human might not remember. For example, during a local festival, one store's bread sales spiked forty percent. The AI caught that pattern from the previous year and ordered accordingly. Luna: So dynamic pricing handles the stuff already in the store, and better ordering prevents stuff from entering the waste stream in the first place. What about the stuff that still goes unsold? The day-old bread, the bruised apples? Lucas: Third lever: donation logistics. Tegut set up a system where each store has a designated 'food rescue' bin. Items that are still safe but can't be sold — dented cans, day-old bread, produce that's a bit soft — go into that bin. Then a refrigerated truck from a local food bank picks it up every morning. Key detail: they don't use third-party apps like Too Good To Go. They built their own partnership network with local charities. Luna: Why not use the apps? Those are popular. Lucas: Two reasons. One, the apps take a cut, and Tegut wanted the full value of the donation to go to the food bank. Two, the apps are consumer-facing — they're designed for people to buy a surprise bag. But Tegut wanted to donate the food, not sell it at a deep discount. They felt that selling it, even at a discount, would cannibalize their rescue shelf. Luna: That makes sense. So the food bank gets it for free, and Tegut gets a tax benefit. But what about the operational cost? Picking and storing those items, coordinating the pickup... Lucas: They calculated that the total cost of the donation program — including labor, bins, and transportation coordination — is about two percent of the value of the donated food. So for every euro worth of food donated, it costs them two cents. That's cheaper than the disposal cost, which was about five cents per euro. Luna: So it's actually cheaper to donate than to throw away. Lucas: Financially, yes. And you get the goodwill. And on that point — let me talk about the measurement piece, because Tegut didn't just track weight. They tracked waste as a percentage of revenue per store. That way, stores that sold more didn't look like they were wasting more. They set a target of below one percent waste per store. The average before was two point three percent, so that sixty-five percent reduction brings them to about zero point eight percent. Luna: That is a concrete number. And it's also a great incentive — store managers want to be below one percent. Do they get bonuses tied to that? Lucas: Yes. Store managers get a bonus for hitting the waste target, but it's weighted against customer satisfaction scores. So they can't just slash orders and leave shelves empty. The bonus is fifty percent waste reduction, fifty percent customer satisfaction. Luna: That's smart. Aligns incentives with the actual business goal. So the results: sixty-five percent waste reduction. But how much of that is from each lever? Lucas: They break it down as roughly: dynamic pricing contributed about thirty percent of the reduction, demand forecasting contributed about forty percent, and donation logistics contributed the remaining thirty percent. And the total financial benefit? They saved about eleven point seven million euros annually in reduced product loss and disposal costs, against a program cost of about one point two million. Luna: So a roughly ten-to-one return. That's huge. Lucas: Right. And it's not just about the money. The environmental impact: they kept about eight thousand tons of food out of landfills each year. That's the equivalent of taking about three thousand cars off the road in terms of methane emissions. Luna: You know, this kind of episode is exactly why I love working on this show. These are real, practical changes that save money and reduce waste. And if today was actually useful to you — if you're thinking about how to apply any of this in your own work — the way these conversations stay ad-free is through listener support. Lucas: Yeah, it's simple. If you've gotten value from this, you can support us at buy me a coffee dot com slash fexingo. That's it. No tiers, no perks, just keeping the show independent. Luna: And we mean it. Every coffee helps us keep doing deep dives like this one. So, back to Tegut — one last thing I want to touch on. Customer perception. Did customers react negatively to the rescue shelf? Did they see it as 'old food'? Lucas: That was a concern. But Tegut did a couple things. First, they branded it positively — called it 'Die Rettung', which means 'the rescue'. Second, they made sure the prices were compelling enough that customers felt they were getting a deal, not second-rate goods. And they actually saw an increase in overall basket size among customers who bought from the rescue shelf — those customers tended to also buy full-price items. Luna: So the rescue shelf acted as a loss leader in a way. Draws people in. Lucas: Exactly. And that's the kind of win-win that good operations can create. You reduce waste, you save money, you build goodwill, and you even increase sales. It's not a trade-off — it's a redesign. Luna: Alright, I'm going to check if my local grocer has a rescue shelf. If not, I might have to ask. Lucas: That's the spirit. And if you do, let us know how it goes.