I Know How to Use Data. I Just Haven’t Figured Out How to Use It at My Bar

I work in continuous improvement at a food manufacturing company. For 14 years my job has been to take a data set, figure out what’s working, what isn’t, and what needs to change — and then act on it. I build dashboards. I create triggers. I run 12-week trend analyses. I set center lines and control limits so that the moment something goes out of range, an action fires automatically.

I am very experienced and like to think I am good at this. I even coach others on how to use data to make daily decisions. 

And yet, when I look at my own bars, I struggle to apply the same thinking. This article is me being honest about that gap — and asking for help closing it.

WHAT DATA LOOKS LIKE IN MANUFACTURING

In a manufacturing environment, data isn’t just something you review at the end of the month. It’s operational. It tells you what to do right now.

Here’s a simple example. A cooler should run at 38 degrees, plus or minus one degree. The moment it drifts outside that range, it triggers an alert. Someone investigates. Someone acts. The data isn’t just descriptive — it’s prescriptive. It tells you when something is wrong before the wrong thing becomes a bigger problem.

We also run rolling averages, not point-in-time comparisons. If I had one bad week and then a normal week, comparing those two weeks to each other is meaningless. What I want to know is how this week compares to the rolling 12-week average. That’s the real center line. That’s the standard.

And when we change something — a process, a standard, a way of working — we document it. We set a new baseline. The data starts measuring against the new normal.

I want that for my bars. I just haven’t built it yet.

WHAT I ACTUALLY LOOK AT RIGHT NOW

At both bars I look at daily sales. I look at hourly sales to understand when we’re busy and when we’re not. Occasionally I look at refunds and returns at The Beer and Now since we have a retail store.

That’s basically it. And I know it’s not enough.

The POS systems give me so much more — sales by product, sales by server, sales by hour, customer frequency, loyalty program data, coupon redemption rates. The data exists. I’m just not using it in a systematic way.

Here’s the question I keep coming back to: what should trigger an action?

In manufacturing that question has a clear answer. Cooler out of range — action. Efficiency below standard — action. Quality defect rate above threshold — action. The trigger is defined. The action is defined. The system runs.

In a bar I don’t have that clarity yet. If my Wednesday sales are tracking 80% below my rolling average by noon — what do I do? Can I even do anything? Is there a coupon I can fire off that would actually move someone off their couch and into my bar in the next two hours? Would a text blast help or would it just annoy people?

I genuinely don’t know. And I think that’s worth saying out loud.

WHERE I THINK THE OPPORTUNITY IS

A few areas where I believe data could be genuinely useful — if I could figure out how to operationalize it:

Event efficacy. When I run a karaoke night or a bingo event, did it actually increase sales versus a comparable non-event night? Or did I just get the same people who would have come anyway, plus give away some beer? Right now I have a feeling about the answer. I don’t have the data to prove it.

Coupon and promotion ROI. Square tracks coupon redemptions and ties them to transaction totals. If I send a Memorial Day $5 off coupon to 200 people and 20 redeem it, I can see what those 20 people spent. That’s real measurement. I’m not doing this consistently but I should be.

Beer velocity and ordering triggers. Which beers are selling faster than expected? Which ones have been sitting for three weeks? That should trigger an ordering decision or a promotion decision, not my gut feeling on a Monday morning.

Customer frequency trends. My loyalty program collects names and contact info. I should be able to see whether my regular base is coming in more or less frequently over time, and whether new customers are converting into repeat visitors. That’s a health metric for the business.

Rolling sales average vs. daily target. Instead of comparing this Wednesday to last Wednesday, I want to compare it to a 12-week rolling average. That tells me something real. One bad Wednesday followed by a great Wednesday tells me nothing.

THE HONEST ADMISSION

I built Al’s Brain — my own operations dashboard — that pulls live Square data and surfaces it every morning. It tells me yesterday’s sales, where we are versus breakeven, what’s trending. It’s useful.

But it’s still mostly descriptive. It tells me what happened. It doesn’t yet tell me what to do about it.

That’s the gap I’m trying to close. Moving from data that describes the past to data that triggers action in the present.

I don’t have the answer yet. If you’re a bar owner who has figured out how to use operational data to make real-time decisions, I genuinely want to hear from you. Leave a comment or reach out directly. This is one of those problems I’m still working through in public.

That’s kind of the whole point of this site.