Latest / The Operator's Manual with Fexingo: Daily Decisions, Tactical Wins, and Business Execution / How a Single Dashboard Saved One Thousand Restaurants
Transcript
- Lucas: So I was reading through Sweetgreen's latest impact report — the one they published back in March — and one number jumped out at me. They claim their internal operations dashboard reduced food waste by twenty-three percent across their entire fleet in the first twelve months. Luna: Twenty-three percent across roughly a thousand locations? That's not just a nice to have. That's real money. Lucas: Exactly. And what's interesting is that this wasn't some OpenAI partnership or a robotic arm that folds bowls. It was a dashboard. A collection of maybe six or seven core metrics that every single store manager sees on a screen in the back of house. The story is really about what they chose to measure and — more importantly — what they chose to ignore. Luna: Because in a thousand-store operation, if you try to track everything, you track nothing. Lucas: Right. Sweetgreen's operations team told a conference last year that they initially had a list of a hundred and forty possible KPIs. They cut it to twelve. Then to seven. And the ones that survived aren't what you'd expect. They don't track 'smiles per customer' or 'Instagram mentions per bowl'. They track things like 'seconds from ticket print to bowl in hand' and 'ounces of prepped ingredient wasted per shift'. Luna: So they're measuring the stuff that actually hits the P&L. Labor, waste, throughput. Not vanity metrics. Lucas: Precisely. And the way they rolled it out is worth studying. They didn't just install a screen and walk away. They paired every metric with a specific action threshold. If 'prep waste per shift' exceeds a certain number, the system triggers a coaching prompt for the shift lead. Not an email to the regional VP — a prompt right there, on the screen, saying 'Hey, your avocado prep yield was six percent below target. Here's a two-minute video on how to portion correctly.' Luna: So the dashboard becomes a teaching tool, not just a report card. That's smart. Because a number without context is just noise. Lucas: Exactly. And the context is delivered at the moment of the problem. They essentially embedded their best-practice training into the operational feedback loop. I think that's why the waste reduction stuck — it's not a quarterly initiative. It's a shift-by-shift habit. Luna: What about the line speed metric? Eighteen percent improvement — that's huge for a lunch rush. How did they avoid just pressuring crews to go faster and burning them out? Lucas: Great question. They told the same conference that they deliberately separated speed from labor cost in the dashboard. So a store manager sees line speed alongside 'labor percent of sales'. The goal isn't just to be fast — it's to be fast within a target labor efficiency. That stops managers from simply throwing more bodies at the line. They have to find process improvements. Luna: So you can't cheat the system by overscheduling. You have to actually redesign the station layout or change the flow of ingredients. Lucas: Right. And that's where the real operational gains come from. One store in Boston realized their line was bottlenecked because the dressing bottles were stored on the opposite side of the station from where the crew member actually dressed the bowl. Moving those bottles saved three seconds per bowl. Across a thousand bowls a day, that's nearly an hour of cumulative labor. Luna: That's the kind of thing that would never show up in a high-level corporate report. You have to be inside the store, looking at the process, with the right data. Lucas: Exactly. And the dashboard is designed to surface those micro-bottlenecks. It tracks 'time at each station' for every crew member — not to monitor individuals, but to spot which station is consistently the slowest. Then the system suggests a specific improvement: 'Your dressing station is running ten seconds slower than the assembly station. Consider moving ingredients or pre-portioning.' Luna: It's almost like a continuous improvement engine. Kaizen, but with a screen. Lucas: Yeah, it's digital kaizen. And here's the part I find most impressive: Sweetgreen built this entirely in-house. They have a small internal team — maybe eight or nine people — who are part operations experts, part software engineers. They spent eighteen months iterating on the dashboard before they rolled it out to all stores. Luna: So they didn't buy an off-the-shelf solution. That's a bet. Most companies would've gone with a vendor and customized it. Why build? Lucas: The operations VP said in an interview that the off-the-shelf tools were either too generic — think Tableau dashboards that just visualize data — or too rigid, locked into a specific industry framework. Sweetgreen wanted a tool that would change as their operations evolved. They also wanted full control over the user experience. They wanted the dashboard to feel like an extension of the store, not a corporate reporting tool. Luna: And that matters. If the dashboard feels like it's for the regional manager, the store team won't engage with it. You want them to own it. Lucas: Exactly. So they designed the interface to be used on a tablet mounted right next to the expo station. It refreshes every sixty seconds. It uses big, color-coded tiles — green is good, yellow is warning, red is action needed. No spreadsheets. No drill-downs that require a finance degree. Luna: What about the human side? Change management is usually the hardest part of any ops tool rollout. Did they face resistance from store managers? Lucas: They did. Especially from veteran managers who'd been running stores for years without a dashboard. Their attitude was, 'I know my store. I don't need a screen to tell me what's wrong.' Sweetgreen's response was interesting: they didn't force it. They piloted the dashboard in ten stores, then let the results speak. After three months, those pilot stores had significantly lower waste and higher customer satisfaction scores. Eventually, the holdout managers came around. Luna: So they used proof, not policy. That's a good lesson for any operational change. Lucas: Totally. And now, two years in, the dashboard has become part of the culture. New managers are trained on it from day one. They even have a 'dashboard champion' in each region — a store manager who's particularly good at using the data and who mentors others. Luna: That's a nice organic scaling mechanism. peer to peer, not top-down. Lucas: Right. And it's worth noting that Sweetgreen isn't a tech company. They're a salad chain. But they invested in building internal operations software because they saw the ROI clearly. Twenty-three percent waste reduction across a thousand stores — at an average ingredient cost per bowl of about three dollars — that's millions saved annually. Not to mention the sustainability angle: less food in the landfill. Luna: And customers notice when bowls are consistent and lines move faster. That drives repeat business. Lucas: Exactly. So the takeaway here isn't 'build your own dashboard.' It's: think deeply about what you measure, pair metrics with action, and design for the person who's actually doing the work. If you can do that, you don't need a hundred KPIs. You need the right seven. Luna: That's a powerful way to think about operations. And if this episode gave you a new angle on your own business, you know how we keep these conversations ad-free and independent. Lucas: Yeah, it's listener support that makes this possible. If today was useful to you, you can throw a few bucks toward the show at buy me a coffee dot com slash fexingo. No pressure, just if you feel like it. Luna: Every bit helps us keep digging into stories like this one. Now back to the dashboard: one metric I didn't mention that they track is 'crew engagement score' — a simple daily pulse check where each team member rates their shift on a scale of one to ten. Lucas: Oh interesting. And they found a correlation between engagement scores and waste — stores with higher engagement had lower waste. So they don't just measure the process; they measure the people. That's the kind of holistic operational thinking that makes a dashboard more than a toy. Luna: Exactly. Because ultimately, operations is about humans making decisions in real time. The best tools amplify their judgment, not replace it. Lucas: Well said. And with that, episode twenty-five in the books. Next time, we'll look at how one logistics company used the same principle — simple metrics, local ownership — to cut delivery delays by thirty percent. Luna: Looking forward to it.