Why Your EPOS Data Is the Key to a Revenue Comeback

Your EPOS already knows what sells, when it sells, and what actually makes money. The problem is that most pubs stop at the Z report: total take, cash versus card, and off to bed.

Everything useful sits one menu deeper, in reports nobody opens. The gap between a pub that grows and a pub that drifts is rarely the system. It is whether anyone reads it on a Monday and changes something on a Tuesday.

This is the minimum dashboard and experiment loop that works, and it takes about an hour a week once it is set up.

Get the data clean before you trust it

Bad data is worse than no data, because you act on it with confidence. Before you build any dashboard, spend one shift auditing your till buttons.

Look for these five faults, all of which are common:

• A "miscellaneous" or "sundry" key absorbing sales that should be attributed to a product. If more than one or two per cent of your take goes through it, your product mix is fiction.
• Items sitting in the wrong category. A bottled cider filed under soft drinks will wreck your wet gross profit by category.
• Duplicate buttons for the same product at different prices, usually left over from a price rise.
• Staff drinks, comps and wastage rung through as sales, or worse, not rung through at all. Every one of those needs its own reason code.
• Cost prices never loaded. Many EPOS systems can hold a cost price against each product. If yours does and the field is blank, your margin reports are guesses.

Fix the buttons, tell the team what changed, and give it a fortnight before you draw conclusions. If your current system genuinely cannot produce item-level reporting, that is a real constraint and our guide to choosing a pub EPOS system covers what to look for.

The four core metrics that matter

Track these every week and nothing else, at least to begin with:

• Covers by day and by time band.
• Average spend per head, split wet and dry.
• Gross profit by category, calculated ex-VAT.
• Labour percentage by shift.

Two of those need a word on the arithmetic, because this is where most licensees go wrong.

Gross profit must come off the ex-VAT price. Take a main course at £14.95. At 20% VAT the ex-VAT selling price is £12.46. If the plate costs you £4.20 in ingredients, your gross profit is £8.26, which is roughly 66%. Run the same sum off the £14.95 and you would tell yourself you were making 72%, and you would price the rest of the menu off a lie. Most pubs aim somewhere in the 65% to 70% range on food, with draught beer typically landing in the high fifties to mid sixties and spirits and wine by the glass considerably higher.

Labour percentage is a per-shift number, not a monthly one. A monthly figure hides everything. Take a Tuesday evening doing £820 in sales with £340 of wages on the floor: that is 41%. Compare it with a Saturday doing £3,400 on £780 of wages, which is 23%. The monthly average of those two might look respectable while Tuesday quietly eats the profit Saturday made. UK pubs generally run labour somewhere between the mid twenties and the high thirties as a share of sales, higher where food is a bigger part of the mix.

Build a one-page weekly dashboard

Keep it to one page, same format, every week. A spreadsheet is fine. Print it and pin it in the back corridor where the team walks past.

That last row is the one that matters. A dashboard without an action is a scrapbook.

The Monday hour

Put it in the diary and defend it. Same day, same hour, ideally before the week's ordering goes in so what you learn can change what you buy.

Fifteen minutes to pull the reports and update the sheet. Fifteen minutes to read it and write down what surprised you. Fifteen minutes to decide the one action. Fifteen minutes to brief whoever has to deliver it.

The discipline is in doing this on a bad week as well as a good one. Bad weeks are when the data is most useful and when you are least inclined to look at it.

Run weekly experiments

Treat each week as a small test. Change one thing, hold everything else steady, and measure the same way you measured before.

Examples worth trying:

• Move a high-margin item to the top of the menu and track units against the previous fortnight.
• Bundle a drink with a midweek dish at a price that protects your cash margin per cover.
• Change one price point by a pound and watch units, not just revenue. If units hold, the pound is free.
• Brief the team on one specific upsell for seven days, such as a premium mixer or a dessert, and count attachment rate.

Here is why small tests compound. Say you serve 300 covers a week. Lifting average spend per head by £1.20 through better dessert and coffee attachment adds £360 a week, which is £18,720 across a year, and none of it requires a single extra customer through the door. That is an example rather than a promise, but the arithmetic is the point: spend per head is usually the cheapest lever you own.

At The Anchor, working this way is how we grew food revenue by 98% in three months. It was not one big idea. It was a run of small, measured changes where the data told us which ones to keep.

Segment your guests

A Friday night crowd is not a Tuesday lunch crowd, and averaging them together produces offers that suit nobody. Split your EPOS data by day and time band and you will usually find three or four distinct trading patterns under one roof.

Look at what each segment actually buys. A midweek lunch trade buying one main and a soft drink needs a different offer from a Sunday family group buying three courses and a bottle of wine. Build for the segment you want more of rather than for an imaginary average customer.

Time bands are worth setting deliberately: open to 3pm, 3pm to 6pm, 6pm to 9pm, 9pm to close. The 3pm to 6pm band is where most pubs discover they are paying for staff to watch an empty room.

Spot hidden profit

Most pubs have a quiet goldmine sitting in one category. Very often it is premium soft drinks, desserts, coffee, or a particular spirit range where the cash margin per serve is far better than the beer everyone is focused on.

Find it by ranking every product by total gross profit contribution, which is cash margin per unit multiplied by units sold. That ranking looks nothing like the ranking by units, and it is the one that should drive your specials board, your staff briefings and your stock investment.

Then train for it. A one-line script at the point of order shifts more product than a poster ever will, and our upselling scripts guide has wording your team can actually use without sounding pushy.

Share insights with the team

Numbers are useless if only one person sees them. Put the dashboard where the team walks past it, and open every weekly briefing with the same three lines: what we did, what happened, what we are changing.

Name the target out loud. "This week we are pushing the sticky toffee pudding, and I want dessert attachment above one in four covers." A specific number gives people something to hit. "Try to upsell more" gives them nothing.

Tell them the result the following week too, including when it did not work. Teams stop engaging with data the moment they suspect it only surfaces when something has gone wrong.

Common mistakes

• Tracking too many metrics. Twenty numbers means nobody looks at any of them. Four is enough for the first quarter.
• Making changes without measuring. Three changes in one week means you learn nothing about any of them.
• Calculating gross profit on inc-VAT prices. It flatters every figure you produce.
• Ignoring labour cost by shift. The monthly average hides the session that is losing money.
• Leaving slow movers on the menu for months. Every dead line ties up stock, adds prep, and lengthens the menu for no return. Pair the product mix with tighter stock and waste control and the effect on margin is immediate.
• Reacting to one day. One wet Wednesday is weather, not a trend. Compare like weeks.

Quick checklist

• Till buttons audited and cost prices loaded.
• One-page dashboard built and printed.
• Four core metrics tracked weekly, gross profit calculated ex-VAT.
• A fixed weekly hour in the diary.
• One experiment run per week, one thing changed at a time.
• Top profit items ranked by cash margin and promoted by staff.
• Slow movers reviewed and cut every quarter.

Mini FAQ

Do I need complex software? No. A spreadsheet and the exports from your existing EPOS system are enough for the first six months. Buy analytics tools only once you have proved you will use the data.

How fast can data improve revenue? Often within two weeks if you focus on one lever at a time. Spend per head usually moves quickest because it needs no extra customers, only a better ask at the point of order.

What if my EPOS cannot export the reports I need? Ask the supplier before you replace anything, because the report often exists and is simply buried. If it genuinely does not, note it as a requirement for your next system rather than changing mid-season.