The Data Your Business Creates Everyday Is Smarter Than You Think.
You can learn a lot about people by analysing, for instance, how they eat. Not just what they choose, but when they order, how much they spend, and how often they return. Recently, during my postgraduate studies in Machine Learning and Artificial Intelligence at the McCombs School of Business at the University of Texas in Austin, I had the chance to study a real-world dataset from a food delivery platform in New York. The numbers didn’t shout. But they told a quiet, powerful story. It was an assignment designed to build foundational data analysis skills. However, as I worked through it, I kept thinking about ‘Order and Eat’, my own mobile food business here in Ghana. Although it is FDA-registered, I shut it down temporarily, until I could either run it better or find someone who could manage it more effectively. In this assignment, I observed behaviours that felt familiar. I noticed customers who weren’t so different from ours. And I realised that what began as a case study was in fact a window into how we might operate better food businesses at home in Ghana. The Weekend Effect One pattern stood out immediately. Nearly three-quarters of all orders were placed on weekends. People were opting for American, Japanese, and Italian dishes. These weren’t quick weekday bites; they were meals that people sat down to enjoy, often shared and indulgent. This suggests that food delivery thrives most when people feel unhurried. They aren’t chasing deadlines or battling traffic. They’re unwinding. This has shaped how I think about scheduling for Order and Eat. Fridays through Sundays are no longer just part of the week. They are a rhythm. We have to plan our menu around them, preparing higher-demand ingredients in advance and using WhatsApp to send out quiet nudges to regulars. The numbers made it clear. Your best sales often happen when your customers are at their most relaxed. Who’s Spending More Than You Think Another detail stood out. Almost a third of the orders within the dataset in this assignment crossed the twenty-dollar mark. These weren’t one-off splurges. They were part of a steady flow of higher-value transactions. That matters. Too often we assume that most customers are watching every cedi. And while many are, some are not. Some are prioritising convenience, quality, or the joy of a particular meal. In Ghana, this segment exists, even if it doesn’t always advertise itself. I’ve seen it with Order and Eat. There are customers who consistently choose the more premium options, who never ask about delivery charges, and who quietly place repeat orders. When this happens, we have to start testing small add-ons just for them. A surprise dessert. A reusable food pack. These are simple things, but they build loyalty in quiet ways. If you run a small business, it is worth asking yourself who your highest-spending customers are and what kind of experience they are having. Are you treating them like everyone else, or are you giving them a reason to keep coming back? Speed Isn’t Everything There’s a widespread belief that faster delivery always means happier customers. But the data I worked on challenged that. Orders that were delivered quickly didn’t always earn better ratings. And the difference between five-star and three-star reviews had little to do with how fast the food showed up. That was a reminder. Satisfaction is about more than speed. It is about whether the food arrived in good condition, whether it tasted fresh, and whether the packaging felt intentional. For many of us in Ghana, where delivery times can be affected by factors outside our control, this is actually good news. We can’t always deliver faster. But we can deliver better. So, we have to start focusing more on consistency. Double-check packaging before dispatch, make follow-up calls on first-time orders, and even include small notes when we have the time. The goal is to shape the full experience, not just beat the clock. Loyalty is Quiet Until You Listen According to the data, one customer placed thirteen orders, while four others placed between eight and ten. These were not corporate bulk orders, but rather individual customers using the platform repeatedly. Yet, there was no mention of loyalty rewards or special offers for them in this assignment. This is a lesson many of us overlook. Not all loyal customers are loud. Some don’t leave reviews. Some don’t tag you on social media. But they come back. Again and again. If you’re not tracking their behaviour, you may never realise just how important they are. So start tracking basic customer behaviours. Nothing fancy. Just a spreadsheet with names, order frequency, and preferences. It will help us know who to thank, who to offer a new dish to first, and who to message when something goes wrong. Loyalty is built in moments, not marketing campaigns. Only a Few Vendors Stand Out Out of nearly 180 restaurants in the dataset, only four met the platform’s criteria for promotion: at least fifty ratings and an average score above four. That was startling. It showed how rare consistent quality and customer engagement really are. In Ghana, where digital food marketplaces are growing, the same problem exists. A lot of vendors get listed, but few are truly dependable. Many shine for a week and fade the next. The food platforms promote whoever brings in volume for them, not necessarily whoever delivers consistent quality to the customers. There’s an opportunity here. If you’re a vendor and you can maintain high standards, gather regular feedback, and respond quickly to issues, you can rise above the noise. And if you’re running a platform, it’s worth thinking about how your algorithm promotes vendors. Are you rewarding consistency and satisfaction, or just sales spikes? The Small Percentage That Can Ruin Your Brand Just over 10% of orders in the case study took more than an hour to deliver. This is not a majority. Not even close! But they still matter. These are the orders that frustrate people.
The Data Your Business Creates Everyday Is Smarter Than You Think. Read More »