Latest / The Operator's Manual with Fexingo: Daily Decisions, Tactical Wins, and Business Execution / How One Grocery Chain Cut Spoilage 50 Percent With One Ordering Change
Transcript
- Lucas: So there's this mid-sized grocery chain based in Ohio — I'll call them Midwest Grocers, not their real name — and they were losing about four point two percent of their total perishable inventory to spoilage every single month. That's about one point six million dollars a year in produce, dairy, deli meat, just thrown away. Luna: And that's before you factor in the labor cost of pulling expired product off the shelf, marking it down, or hauling it to the compactor. Lucas: Exactly. Their industry benchmark for a well-run grocery is under two percent spoilage. They were more than double that. And here's the interesting part — they'd already tried markdown optimization, better rotation training, even a partnership with a food-rescue app. Nothing moved the needle more than a couple of basis points. Luna: So what finally worked? Lucas: They changed one thing in their ordering system. They switched from a weekly demand forecast — where a buyer in the back office would project what each store needed for the next seven days — to a daily forecast that updated every morning based on what actually left the shelves the day before. Luna: And that alone cut spoilage in half? Lucas: Fifty-three percent, to be precise. And I want to be clear — the forecast change wasn't just about frequency. The key was the data source. The new daily forecast used real-time inventory sensors on the shelf, not the back of house stockroom counts. Luna: So they shifted from what was in the building to what was actually selling in the moment. Lucas: Right. Because the old system was ordering based on what the warehouse shipped to the store, not what customers bought. So if a case of avocados arrived Monday but didn't move until Wednesday, the system thought demand was low and ordered less — even though the shelf was full. Then by Thursday the shelf was empty and you lost sales. Luna: That's a classic bullwhip effect inside a single store. The ordering signal gets distorted by inventory that's sitting in the back room. Lucas: Exactly. The sensor gave them a true demand signal. And the results came fast. Within two weeks, the daily forecast reduced over-ordering on long-shelf-life items like apples and potatoes. But the bigger impact was on high-turn, high-spoilage items like bagged salad, fresh berries, and deli chicken. Luna: How did the store teams react to the change? I imagine the produce manager might be skeptical about giving up ordering control. Lucas: That was the biggest hurdle. The produce managers had been ordering by gut feel for years. They knew their customers, they knew their foot traffic patterns. And honestly, they were pretty good at it — their spoilage was lower than the chain average. The problem was the deli and the dairy departments, where ordering was done by part-timers with no training. Luna: So the sensor-based forecast actually leveled the playing field. It made the worst orderers better without making the best ones worse. Lucas: That's exactly how it played out. The company piloted the change in six stores for three months. The produce managers who initially resisted — after a month, they started using the forecast as a check on their intuition. They'd look at the daily number and say, 'Huh, I would have ordered two cases, but the sensor says one. Let me trust it.' And it was right eighty percent of the time. Luna: That's a nice cultural win. The data didn't replace their judgment — it sharpened it. Lucas: Now, the real kicker — there was a second-order effect nobody predicted. When the stores stopped over-ordering, the shelf sets actually looked fuller because product turned faster. Customers saw fresher stock, and the chain saw a twelve percent increase in full-price sales on high-margin perishables like organic produce and specialty cheese. Luna: Because they weren't constantly marking things down to move them before expiration. Lucas: Right. The markdown rate on deli items dropped from eighteen percent to six percent. That alone covered the cost of the sensors in about four months. And the labor savings — store teams were spending two to three hours per week per store doing manual inventory counts. The daily forecast eliminated that entirely. Luna: If today was actually useful to you — and it sounds like it might be if you're in any kind of perishable inventory business — the way these episodes stay ad-free is listener support. You can buy me a coffee at buy me a coffee dot com slash fexingo. It's what keeps the whole library of three hundred daily shows free. Lucas: Yeah, we don't do sponsorships, so it's listener-funded. That link — buy me a coffee dot com slash fexingo — is the only reason we can keep diving into these one-change case studies without selling your attention. Luna: Okay, back to the grocery chain. Lucas, what about the suppliers? If the stores started ordering less of some items, did the distributors push back? Lucas: They did, initially. The chain's biggest produce supplier — a large distributor out of Cincinnati — was used to shipping a certain volume each week. When the daily forecast caused a fifteen percent drop in weekly orders for certain items, the supplier's sales rep called the VP of procurement and complained that the chain was 'destabilizing the supply chain.' Luna: Classic. The supplier's efficiency depended on predictable order patterns, not actual demand. Lucas: Exactly. But the chain held firm, and after two months, the supplier realized the chain's orders were actually more stable on a day-to-day basis — just lower on average. The volatility came from the old weekly ordering, because the store would order a huge case count Monday, sell through by Wednesday, then panic-order an emergency delivery Friday. The daily forecast smoothed all that out. Luna: So the supplier ended up with a more efficient delivery schedule too. Lucas: Yeah, their route optimization improved, and their own spoilage on the warehouse side dropped, because they weren't holding as much buffer stock to cover the weekly lumpiness. The chain ended up sharing the cost savings — they negotiated a three percent discount on the next year's contract. Luna: That's a great example of a change that ripples beyond your own four walls. One ordering tweak, and it reshapes the supplier relationship. Lucas: And the final piece — the chain also used the daily forecast data to adjust their in-store promotions. They used to run 'two-for-one' sales on strawberries based on a calendar, regardless of supply. Now they run the promotion only when the sensor data shows a surplus building — basically, they mark down before spoilage happens, rather than after. Luna: So they flipped from reactive shrink management to proactive shrink prevention. Lucas: Exactly. And the result: total spoilage across the six pilot stores dropped to one point nine percent — below the industry benchmark. The chain is rolling it out to all forty-two locations this year, and they expect to save about two point eight million dollars annually. Luna: And all because they changed the time horizon and the data source for one forecast. No new software suite, no consulting engagement. Lucas: No. The sensors cost about eight hundred dollars per store. The forecast algorithm was written by an internal data analyst in about three weeks. The whole thing paid for itself in under one quarter. Luna: I think the lesson here is that a lot of operations problems are actually information problems. The product was fine, the staff was fine — the signal was just lagging and noisy. Lucas: Right. And fixing the signal doesn't have to be expensive or complicated. It just has to be the right signal. For this chain, weekly warehouse shipments were the wrong signal. Daily shelf-level sales were the right one. Luna: So next time you're staring at a spoilage report or a shrink number that feels baked into the business, ask: is my ordering system looking at the shelf or the stockroom? Lucas: That's a good question to leave on. For The Operator's Manual, I'm Lucas. Luna: And I'm Luna. See you tomorrow.