The phrase "customer data" makes most restaurant owners think of something they do not have and cannot afford — a system to buy, a consultant to hire, a dashboard full of charts nobody reads. That is backwards. You are already collecting customer data, every single service, and have been for years. It is sitting in your POS, your booking system and your loyalty program, unread. This course is not about buying anything. It is about opening what you already own and learning the small amount of arithmetic that turns it into decisions.
🗄️ You are already collecting it
Take a normal Friday. Every check that closes records what was ordered, when, at what table, for how much. Every booking records a name, a phone number, a party size and whether they turned up. Every loyalty member who scans records that this specific person was here again. Every online order records an address and an entire history attached to one account. None of that required a decision from you — it is a by-product of running the restaurant.
The gap is not collection, it is that nothing joins it up and nobody looks. The POS knows the check was $84 but not that it was Ana's fourth visit. The booking system knows Ana books every couple of weeks but not what she spends. So the restaurant ends up with a great deal of information about transactions and almost none about people — which is why most owners can tell you last month's sales to the peso and cannot tell you how many different human beings produced them.
🧩 Three kinds, and only one of them is hard
It helps to sort what you have into three buckets, because they differ enormously in how easy they are to get and how useful they are.
| Kind | Examples | How you get it |
| What they do | Visits, dates, spend, dishes, channel | Free — your systems already log it |
| Who they are | Name, email, phone, birthday, area | They have to give it to you, so ask for little |
| What they say | Reviews, survey answers, complaints | The previous course; words, not numbers |
The first bucket is the one that matters most and the one owners neglect, because it feels less like data than a list of email addresses does. It is the opposite: behaviour is honest. What people actually did last quarter predicts what they will do next quarter far better than anything they told you about themselves. A customer who has come three times in six weeks is telling you something a demographic profile never will.
The second bucket is worth having but is where restaurants overreach, asking for birthdays, anniversaries, home addresses and preferences that then sit unused forever. Ask for the minimum that lets you recognise someone and reach them. The third bucket was the whole of the last course — this one deliberately stays on the numbers, and the two get reunited at the end, in topic 6.
🚧 What the data cannot tell you
Being honest about the limits early will save you from the two ways this goes wrong. The first is that your data only sees a slice of your customers. Walk-ins who pay cash and never join anything are invisible; you will know a lot about your loyalty members and online orderers and almost nothing about the couple who came last Tuesday. Any conclusion you draw describes the customers you can see, not all of them, and forgetting that turns a partial picture into a confident wrong answer.
The second is that numbers tell you what happened and never why. You will see that Thursday visits fell by a fifth since February. The data cannot tell you whether that is the new menu, the roadworks, the manager who left, or a competitor opening. That is what the previous course was for: the numbers find the question, the words answer it. Anyone who tells you the data speaks for itself is about to sell you something.
🧭 What this course covers
Bringing what you already have into one place, with one record per person instead of five contradicting lists, and what to ask for and what to leave alone (topic 2). Then the core of it: three numbers per customer — how often, how much, how long ago — and the handful of groups those numbers create, which is where this stops being theory (topic 3). Using what you know without being creepy, and what personalisation is genuinely worth the effort (topic 4). Spotting the regular who has quietly stopped coming, while you can still do something about it (topic 5). And finally putting numbers and words together into two decisions a month, and noticing when data is being used as decoration for a decision already made (topic 6).
Two things this course will not do. It will not recommend software: the arithmetic here works on a spreadsheet, and if you outgrow that you will know exactly what to look for. And it will not turn you into an analyst — a restaurant needs about five numbers, understood properly, not fifty on a screen.