Latest / Franchise Conversations with Fexingo: Buying, Running, and Scaling Franchise Businesses / How Franchisees Use Predictive Analytics to Forecast Labor Needs
Transcript
- Lucas: You know how a lot of franchisees still build their weekly schedule by hand, or with a spreadsheet that just uses last year's sales as a rough guide? Luna: Yeah, I see it all the time. Maybe they look at last week's numbers and add a body if they're expecting a busy Friday. Lucas: Exactly. And that's fine for a baseline. But there's a growing group of franchisees who are taking it a step further — using predictive analytics to forecast exactly how many staff they need, hour by hour, days or even weeks in advance. Luna: And the payoff is real. We're talking 10 to 15 percent lower labor costs, and often better customer service because you're never understaffed. Lucas: So let's get specific. I was looking at a case study from a McDonald's franchisee in central Florida — operates about a dozen locations. They started using a tool called CrunchTime, which pulls in point of sale data, historical transaction patterns, and even external data like local school calendars and weather forecasts. Luna: Weather forecasts — that's a smart one. A rainy afternoon can tank drive-thru traffic or send it indoors, and if you've scheduled based on a sunny day, you're overstaffed. Lucas: Right. This franchisee told me that before the predictive system, they'd typically schedule about 28 labor hours per thousand dollars in sales. After six months with the analytics, they got it down to 24.5 hours per thousand. That's a 12.5 percent improvement. Luna: And that's pure margin. In a QSR where food costs eat up about a third of revenue and rent another chunk, labor is one of the few levers you can really pull. Lucas: Exactly. So how does it actually work? Most of these systems use a machine learning model trained on at least two years of historical data. They look at transaction counts by half-hour increments, average check size, and then layer in variables like day of week, holidays, local events, and yeah, weather. Luna: And the model learns that, say, every home football game at the local high school boosts sales by 22 percent between 6 and 8 PM. Lucas: Exactly. So now the franchisee can see, two weeks out, that next Friday the system predicts 475 transactions between 5 and 9 PM, and it recommends scheduling three cashiers, two grill cooks, and one manager for that window. Not just a raw headcount — it tells you the mix. Luna: And that's key because a lot of stores run into trouble when they have enough bodies but the wrong skill sets. You might have five people on shift but nobody on the register. Lucas: Right. The better systems also track individual employee performance — who's fastest on the grill, who's best at upselling — and they optimize the schedule for who should be working when. Luna: That sounds great in theory, but I imagine a lot of franchisees are wary of handing scheduling over to an algorithm. There's a human element. Employees have preferences, limited availability. Lucas: That's the thing — these tools don't replace the manager's judgment, they inform it. The system generates a recommendation, but the manager can adjust based on known constraints. Most platforms have a drag and drop interface where you can swap people in and out. Luna: So what's the barrier to entry? Is this only for big multi-unit operators with IT budgets? Lucas: It's getting more accessible. A basic labor forecasting module from a provider like 7shifts or ScheduleFly starts around 50 to 100 dollars per location per month. For a single-unit franchisee, that's probably less than one shift of overstaffing per month. Luna: And the ROI is pretty clear. If you save even 5 percent on labor, on a store doing a million a year in sales, that's maybe five to seven thousand dollars annually. The software pays for itself in a month. Lucas: Exactly. But here's the catch — the quality of the prediction depends heavily on clean data. If your point of sale system has missing or messy transaction records, the model will spit out garbage. Luna: So you need to audit your data first. Make sure every transaction is tagged with the correct time, and that you're capturing all sales channels — online orders, third-party delivery, in-store. Lucas: Right. One franchisee I spoke with spent two months just cleaning up their POS data before they even turned on the predictive feature. They found that about 8 percent of their transactions were missing timestamps. Luna: Once they fixed that, the model accuracy jumped significantly. But that's a step a lot of people skip. Lucas: So if you're a franchisee listening and you want to try this, I'd say start small. Pick your busiest day of the week — maybe Saturday — and run a parallel forecast for that day only. Compare the prediction to what actually happens for a month. Luna: That's a low-risk way to validate the model. And if it's within 5 percent of actual sales, you can expand to the full week. Lucas: And don't forget to account for seasonality. A beach-town franchise might have a completely different pattern in July versus December. The model needs at least one full year of data to capture that. Luna: You know, I've been thinking about this in the context of the broader labor market. With unemployment still low in most metro areas — around 3.5 percent nationally — it's harder to find good staff. So using analytics to maximize the productivity of the people you have is becoming a competitive advantage. Lucas: Absolutely. It's not just about cutting hours — it's about having the right people in the right place at the right time. That improves employee morale too, because they're not bored during slow periods or overwhelmed during rushes. Luna: And that reduces turnover, which is a huge cost for franchisees. Replacing a crew member can cost as much as 1,500 dollars in training and lost productivity. Lucas: Right. So the analytics aren't just saving money on labor — they're saving on recruitment and training costs too. Luna: If today's conversation gave you a new idea or helped you see a way to improve your own scheduling, honestly, that's worth a coffee. You can show it at buy me a coffee dot com slash fexingo. It's a small way to keep these episodes ad-free and focused on what works. Lucas: Yeah, we appreciate that. And we're always looking for more real-world examples — so if you're a franchisee using predictive scheduling, drop us a note. Luna: Alright, let's circle back to implementation. What's the one piece of advice you'd give a franchisee who's on the fence? Lucas: I'd say: start with the data cleanup, get a free trial of a tool like 7shifts, and run it in parallel with your current process for four weeks. The numbers will speak for themselves. Luna: And if you see a 10 percent labor cost drop, you'll wonder why you didn't do it sooner. Lucas: Exactly. That's the franchise edge — small improvements that compound across every unit.