Latest / Restaurant Business with Fexingo: Food Service, Hospitality, and Independent Operators / Why Restaurant Menu Data Is the New Gold Mine
Transcript
- Lucas: Let me start with a number that stopped me this week. A mid-sized independent restaurant group in Chicago — four locations, doing about eight million in revenue — discovered that one of their pasta dishes, which accounted for eighteen percent of entree sales, was actually dragging their overall food cost by three full percentage points. Luna: Because the ingredients were expensive relative to the price point, or because it generated a lot of waste? Lucas: Both, actually. The dish had a forty-two percent food cost, versus their target of thirty percent. But they'd never run the item-level data because the owner just assumed it was a customer favorite worth keeping. Once they swapped in a cheaper protein and adjusted the plating, they saved over fifty thousand dollars a year across four restaurants. Luna: So the data was sitting there in the POS system the whole time. Lucas: Exactly. And that's what I want to talk about today — menu data. Not just the recipes or the prices, but the granular transaction-level data that every modern POS captures and most operators barely use. We're seeing a wave of third-party analytics tools — companies like Upserve, Restaurant365, and even some newer entrants — that promise to turn that raw data into actionable decisions. Luna: How big is this market? I know Toast and Square have their own analytics modules now. Lucas: According to a report from HTF Market Intelligence last quarter, the restaurant analytics software market is projected to grow at a compound annual growth rate of about fourteen percent through 2030. That's faster than the broader restaurant tech market. And the pitch is straightforward: if you know exactly which items drive profit, not just revenue, you can redesign your menu to steer customers toward high-margin choices. Luna: But isn't that essentially what menu engineering has always been? Using placement, design, and description to push certain items? Lucas: Right, but historically menu engineering relied on guesswork and manual spreadsheets. You'd look at sales volume and gross margin once a quarter and make changes. Now you can get real-time data on how a dish performs by daypart, by location, by server, even by season. One chain I looked at — let's call them a regional casual-dining group — used this data to discover that their lunch specials had a much higher contribution margin than dinner entrees, so they redesigned their dinner menu to feature some of those lunch items. Luna: Did it work? Lucas: Their dinner check average actually went up slightly because the lunch items had lower food cost, so they could price them competitively and still keep margin. And they reduced kitchen complexity by eliminating two slow-moving dinner items that required separate prep. The general manager told me they cut food waste by twelve percent in the first quarter. Luna: That's real. But I want to push on the privacy side. When you're collecting granular data per item, per server, per time slot — are customers aware that their choice of a cheeseburger is being analyzed at that level? Lucas: It's a fair question. Most of this data is aggregated and doesn't include personal identifiers — it's tied to the transaction, not the individual. But as these tools get more sophisticated, some are starting to link purchase history to loyalty programs. That's where you get into the territory of knowing that Customer X always orders the ribeye on Fridays. Luna: And that could be used to personalize promotions or, less benignly, to adjust pricing dynamically per customer. Lucas: Right now dynamic pricing is rare in full-service restaurants, but it exists in fast-casual and quick-service. McDonald's and Wendy's have experimented with surge pricing during peak hours. The data infrastructure to do that per customer is not widespread yet, but the analytics tools are getting there. Luna: Let's take a quick step back. Lucas, you mentioned that most operators aren't using this data. Why not? Is it a cost issue, or a skills gap? Lucas: Both. The basic POS analytics that come with a system like Toast are actually quite good for free. But the deeper stuff — cohort analysis, demand forecasting, menu mix optimization — often requires a separate subscription that can run two to five hundred dollars a month. For a single-unit independent doing, say, seven hundred thousand in revenue, that's a meaningful expense. Luna: And they're already tight on margins. Lucas: Exactly. So the ROI has to be clear. But the evidence is mounting. A 2025 study from the National Restaurant Association found that operators who use data analytics tools report food cost percentages that are, on average, two points lower than those who don't. For a restaurant doing a million in food sales, that's twenty thousand dollars a year. Luna: I can see why the tech vendors are pushing hard. It's a virtuous cycle: the more data you collect, the more valuable the analytics become, and the stickier the platform. Lucas: And that's where the business model gets interesting. Some of these platforms are actually giving away the analytics software for free and making money on the transaction processing or on advertising. Toast, for example, reportedly processed over forty billion dollars in payments last year. Their analytics module is an add-on that drives stickiness, not direct revenue. Luna: So the data is the new gold mine, but the pickaxes are getting cheaper. Lucas: Nice metaphor. And that's exactly the trajectory. I think we'll see more consolidation — large POS providers acquiring analytics startups to bundle everything together. Luna: Speaking of trajectory, I want to ask about what this means for the menu itself. If data tells you that a certain dish isn't profitable enough, do you redesign it, rep rice it, or cut it? Lucas: That's the million-dollar question. The best practices I've seen involve a three-tier framework. First, look at contribution margin — not just food cost, but labor and overhead per dish. Second, look at popularity. Third, look at strategic value — does this dish define your brand even if it's not a big moneymaker? Luna: That last one is crucial. I think of the classic example of a steakhouse that keeps a low-margin appetizer because it's a signature. Lucas: Right. The data should inform, not dictate. But too many operators ignore the data entirely. They keep items on the menu out of sentiment — 'my grandmother's recipe' — and it's costing them real money. Luna: So what's the one piece of advice you'd give to an independent operator listening right now, if they want to start using menu data? Lucas: Start with your top ten items by revenue and compute the food cost percentage for each. If you don't have a tool, do it in a spreadsheet. Then ask: which items have the highest and lowest contribution margin? That alone will probably reveal one or two dishes that are either stars or anchors. Luna: And if they want to get more sophisticated? Lucas: Look at sales by daypart and by server. You might find that a particular server sells a lot of low-margin desserts because they're pushing them for tips. That's a training opportunity, not a menu problem. Luna: That's a great insight. Lucas, before we wrap, I want to touch on something you alluded to earlier — the idea that this data can also help with sustainability. If you know exactly what sells and what doesn't, you can reduce overproduction and waste. Lucas: Absolutely. The same Chicago group I mentioned at the top used their data to forecast demand more accurately. They cut their weekly produce order by fifteen percent because they realized they were over-ordering based on guesswork. That's less food in the trash and more money on the bottom line. Luna: And that's the kind of win-win that makes this worth paying attention to. Lucas: Quick honest thing — if you're finding these conversations useful, a handful of listeners chip in monthly through buy me a coffee dot com slash fexingo, and that literally funds making this many episodes. Luna: Yeah, it's a small group, but it keeps us ad-free and focused on the details. If it's moved your work forward in some small way, that's the reason we do it. Lucas: So back to the data story — I think the next frontier is real-time menu optimization. Imagine a system that adjusts menu board prices on the fly based on inventory and demand, like airlines do. Luna: That's coming faster than most operators realize. We've seen it in quick-service already. Lucas: Right. And as the tools get cheaper, even independents will have access. It's not about replacing the chef's intuition — it's about giving them better information. Luna: Alright, let's leave listeners with one actionable takeaway. Lucas: Pull last month's sales report. Sort items by contribution margin. Find the bottom three. Ask yourself honestly: do they belong on your menu? If the answer is no, change the dish or drop it. That exercise alone could save you thousands. Luna: And maybe free up a slot for something new. Lucas: Exactly. Thanks, Luna. Luna: Thanks, Lucas. See you next time.