Straight answer: your EPOS system is already telling you exactly where your money leaks, which hours are wasting staff wages, and which products you should stop stocking, but only if you read past the total revenue figure on the front screen. Most owners glance at that one number, feel briefly relieved or panicked, and close the tab. That’s the whole problem in one sentence.
What an EPOS report tells you
EPOS stands for electronic point of sale, and if you’re running one (Square, Epos Now, Lightspeed, Vend, Zettle, whatever), it’s quietly logging every transaction with a timestamp, an item, a price, a staff member, and often a payment method. That’s not a sales figure. That’s a diary of your entire trading week, written by a machine that never forgets and never lies.
The mistake I see constantly, from cafes to salons to small retail shops, is treating the report screen like a scoreboard rather than a diagnostic tool. You wouldn’t go to a doctor, get a full blood panel done, and only look at your weight. But that’s exactly what happens with daily takings: “we did £1,400 today” gets written on a whiteboard and the other forty data points in the report get ignored.
The five reports worth opening every single week
Not every report your system spits out matters. These five do:
- Sales by hour – shows you exactly when money comes in, down to the hour, so you can match staffing to demand instead of guessing.
- Sales by product or category – your actual bestsellers and, more usefully, your actual worst sellers.
- Average transaction value (ATV) – how much a typical customer spends per visit, which tells you whether upselling is working at all.
- Void and refund report – every cancelled sale, discount applied, and refund given, which is where theft and human error both hide in plain sight.
- Staff performance report – sales per team member, per shift, which is uncomfortable to look at but tells you who’s driving revenue.
If your system only lets you export a CSV rather than showing pretty dashboards, that’s fine. A spreadsheet with these five views built as pivot tables will do more for your business than a shiny dashboard you never open.
A real example: the bakery that thought Saturdays were the problem
A few years back I worked with a small bakery and coffee shop in the south east, three staff on weekdays, five on weekends, using Epos Now. The owner was convinced Saturdays were underperforming because footfall “felt” quiet compared to how busy it looked. When we pulled the sales by hour report for eight weeks, the picture was completely different from her instinct.
Saturday revenue was up 12% year on year. The real problem was Tuesday and Wednesday between 2pm and 4pm, two hours a day, five days a week, where the till was averaging under £40 an hour against a staffing cost of roughly £24 an hour for two people on the floor. That’s a two-hour window, twice a week minimum, running at a loss once you factored in rent apportioned across trading hours.
The fix wasn’t dramatic. She cut one staff member from that specific window, ran a small “afternoon treat” bundle (a coffee and a slice for £4.50 instead of buying separately at £5.60) to try and lift the average transaction value during the dead patch, and moved the second staff member’s shift to cover the Friday lunch rush the sales by hour report showed was consistently understaffed. Within six weeks, that Tuesday to Wednesday window’s ATV rose from £3.20 to £4.10, and labour cost as a percentage of revenue for the week dropped from 34% to 29%. Nobody did anything clever. She just read the report for the first time.
The report most owners never open, and why it’s the one that matters
Here’s the uncomfortable bit nobody wants to say out loud: the void and refund report is the one small business owners avoid because it can implicate their own staff, and sometimes their own decisions. Nobody enjoys discovering that one team member is applying £3 “manager discretion” discounts to friends every shift, or that a till is showing a pattern of small cash refunds right before closing.
I’ve sat with clients who didn’t want to run this report because they were worried about what they’d find, and that avoidance cost them real money. One retail client discovered, after finally pulling twelve weeks of void data, that a single staff member accounted for 60% of all voided transactions across the whole team, despite working roughly a fifth of total shifts. It wasn’t necessarily theft. It might have been a training gap or a till they found confusing. But she’d never have known without looking, because the top-line revenue number looked perfectly healthy the whole time. Growth isn’t only about adding more sales. Sometimes it’s about stopping the ones you’re quietly losing.
How to turn a report into a decision, step by step
Reading the report is step one. Most people stop there. Here’s the actual process worth running monthly:
- Step 1: Pull sales by hour for the last four full weeks, not just one week, because a single bad Tuesday could be a fluke, not a pattern.
- Step 2: Rank your top ten and bottom ten products by units sold, then check margin, not just revenue, on each one. A product that sells brilliantly at a 15% margin can be worse for you than a slower seller at 45%.
- Step 3: Cross-reference low-performing time slots against your staff rota. If a slot is consistently under target, that’s a staffing decision, not just a “quiet day” you write off.
- Step 4: Check ATV trend month on month. If it’s flat or falling while footfall is stable, your upselling or bundling isn’t working and needs a rethink.
- Step 5: Run the void report and read every line for one full week, not just the total. Patterns show up in the detail, never the summary number.
- Step 6: Pick one change based on what you found, run it for a minimum of four weeks, and measure the same reports again before making a second change.
That last step matters more than people think. Changing three things at once means you’ll never know which one moved the number.
Where AI fits into all this
This is where things have shifted in the last couple of years. A lot of EPOS platforms now bolt on basic AI forecasting, predicting next week’s busy periods from historical patterns, flagging stock that’s likely to run out, even suggesting reorder quantities. Used well, this saves hours of manual spreadsheet work every week. I’ve written before about how AI helps small businesses grow smarter and faster, and stock and sales forecasting is one of the strongest, most immediately practical uses of it for a business with physical tills, far more useful right now than chatbots or content generation for most retail and hospitality owners.
But the forecasting is only as good as the data feeding it, which means the same discipline applies: someone still has to check the outputs against reality every few weeks, because AI forecasting trained on eight weeks of a struggling summer won’t handle a sudden autumn rush well without a human sense-checking it.
Connecting your EPOS data to your marketing
This is the bit most EPOS guides completely skip, and it’s where the real growth usually sits. Your sales by hour report should be directly shaping your social media posting schedule. If your data shows a spike in walk-ins between 5pm and 7pm on Thursdays, that’s precisely when you want a promotional post landing, not whenever happens to be convenient for you. I’ve covered this timing question in more depth around the best time to upload Reels for maximum reach, and pairing that timing data with your own till patterns, rather than generic advice, is where the real advantage sits.
Your bestseller report should also inform what you’re promoting. I’ve seen businesses spend a marketing budget pushing a product that, according to their own EPOS data, sells three units a week, while their genuine top performer gets no promotion at all because nobody bothered to check. And don’t dismiss why engagement signals like likes still matter here either; a product getting strong social engagement but weak till sales is telling you something specific, usually that people love the idea of it but the price point or in-store placement is putting them off at the point of purchase.
If you’re running any outbound sales alongside your retail or hospitality operation, business customers, corporate orders, catering, that kind of thing, your EPOS data on which products and price points convert best should feed directly into how you approach it. There’s a useful roundup of sales prospecting tools worth looking at if you want to build that outbound side, using your till data as the evidence for what to pitch.
The number everyone chases and the one they should
Almost every small business owner I’ve worked with obsesses over total revenue and footfall. Those matter, but here’s the uncomfortable truth that gets left out of most EPOS advice: total revenue can be climbing steadily while your business is quietly getting less profitable, because nobody’s watching margin by product or labour cost as a percentage of sales. I’ve watched businesses celebrate a “record month” that was, once you accounted for a discounting push and extra weekend staffing, less profitable than the quieter month before it. The top-line number felt good. The report, read, said something else entirely.
If you take one thing from your EPOS system this month, make it labour cost as a percentage of revenue by day, not just by week. That single ratio, tracked daily, will show you more about the health of your business than a month of watching the total climb.
Frequently asked questions
What’s the difference between a Z-report and an X-report on my EPOS system?
An X-report is a snapshot you can pull mid-shift without closing the till, useful for checking progress during the day. A Z-report is the end-of-day close, and it resets the daily totals, so it’s your official record for that trading day. Always keep your Z-reports; they’re what your accountant and, if it ever comes to it, HMRC will want to see.
How often should I review my EPOS reports?
Daily for the total takings and any obvious void or refund spikes, weekly for sales by hour and product performance, and monthly for the bigger trend view: margin, ATV, and labour cost as a percentage of revenue. Reviewing everything daily leads to burnout and you’ll stop looking altogether within a fortnight.
My EPOS dashboard shows dozens of reports. Which ones can I ignore?
Ignore anything that duplicates a view you already check elsewhere and any “insight” widget that just restates your total revenue in a different chart style. Focus on sales by hour, sales by product with margin attached, ATV, void/refund, and staff performance. Those five cover almost everything that changes a decision.
Can EPOS data really help with marketing decisions, or is that a stretch?
It’s not a stretch, it’s underused. Your till data tells you exactly when customers show up, what they buy versus what you assume they buy, and which promotions moved units rather than just felt successful. That’s more reliable than guesswork or generic best-practice timing advice, because it’s your customers, not an average.
Related reading: How to Influence Customers and Grow Sales and Using Sales Route Planner Technology to Grow Your Business.
