Scanning delivery notes: how a pub catches price rises before they eat the GP

Your GP report is only as honest as your cost prices, and cost prices go stale the moment a supplier reprices. Heineken's February 2026 rise averaged 2.7% before duty and it moved line by line, some products up, Foster's actually down, packaged frozen. If your stock system is still carrying last year's cost price, your GP report smiles at you while the margin bleeds. Scanning the delivery note is the thirty-second habit that keeps the numbers true, and this is what it does, what it does not do, and the two checks that matter at the cellar door.

I will say the quiet part first: this is a guide about a feature StockTap has, so read it knowing that. But the problem is real whatever software you run, and the paper trail works the same on a clipboard.

Two documents, two different checks

Every delivery generates two pieces of paper, and they answer different questions.

The delivery note arrives with the dray. Its question is: did what is written here actually come off the wagon? Count the kegs on the ground, not the kegs on the sheet, before you sign. Short deliveries and substitutions are normal life, and a signed note for eleven kegs when ten landed is a variance you created yourself. I have covered what delivery problems do to a stocktake before; the cellar-door count is the fix.

The invoice comes later. Its question is: what did each line actually cost this time? This is the one nearly everyone skips. The delivery gets checked because missing kegs are visible. A price that crept 40p a case is invisible, it just turns up weeks later as a GP report that will not add up.

Why stale cost prices wreck the GP report

The sum is blunt. Say a supplier rise adds 2.7% to a £700-a-week wholesale bill. That is £18.90 a week of extra cost, about £245 over a quarter, coming straight off gross profit until you either reprice or at least know you have chosen not to. Miss it for a quarter and your GP report has been overstating you the whole time, because it is still costing stock at the old prices. Those figures are my own illustration, but the 2.7% is Heineken's own announced average, and the point of the line-by-line detail is that an average tells you nothing about your bar. You have to see your lines.

And the repricing arithmetic is harsher than most people think. To hold your GP percentage, a cost rise has to go on the shelf multiplied, not matched. At 65% GP, a 5p cost rise on a pint needs about 17.5p on the selling price once VAT is on. The full workings are in the price rise tracking guide, along with what Heineken's rise did line by line. The short version: a pub that does not read its invoices gives itself a silent pay cut every February.

What scanning a delivery note actually does

In StockTap the flow is a photo. You take a picture of the delivery note or invoice, the software reads the lines and drafts an entry: products matched to your library, quantities, prices. You check the draft, fix anything it misread, and approve it. Nothing posts without your say so, and that is deliberate. A crumpled note photographed in a dark cellar is exactly the kind of thing automated reading gets wrong, so the design is AI drafts, human approves.

On approval, three useful things happen at once. The delivery lands in your stock as purchases, so your next count's expected-stock sum is right. Any changed cost price updates on the product, so the GP report is costing at what you actually pay now. And a finance entry posts, so the money side and the stock side stop being two separate arguments. The point is not saving the typing, though it does. The point is that cost price changes get seen the week they happen instead of surfacing as a mystery in the quarterly numbers. The draft shows the new unit cost next to what you have been paying, so a changed price is in your face on delivery day rather than left to your memory.

What it will not do, and who does more

Being straight about the limits. StockTap's scanning is built for a pub keeping its stock and cost prices true. It is not an accounts-payable suite. There is no purchase-order matching, no approval chains, no recipe costing to allocate that invoice across a menu. If you are food-led or multi-site and you want full invoice processing wired into recipes and a finance system, that is the territory where Jelly and MarketMan earn their much bigger price tags, and I have written both up straight, credit where it is due.

For a wet-led pub, though, most of that machinery is solving problems you do not have. Your job is narrower: get the quantities in and keep the cost prices current, so the GP report tells the truth. That is a thirty-second photo per delivery, not a finance project.

Tied pubs: the invoice trail is your evidence

If you are tied, there is a second reason to capture every invoice, and it is the Pubs Code. A significant tied price rise, measured against the same four weeks last year and the thresholds of CPI plus 3% for beer and CPI plus 8% for other alcohol, can open the market rent only route, and the window to act is 21 days. You cannot spot a trigger you have no record of. A dated, line-by-line history of what your pubco charged you is exactly what that conversation runs on, and it builds itself if the invoices get scanned as they arrive. The mechanics are in the tied price rises guide.

The weekly habit

Here is the whole discipline, and it fits inside the delivery itself. Count the physical stock against the note before signing, and mark what is short. Photograph the note there and then, because paper that goes upstairs to be dealt with later has a way of reaching March. When the invoice lands, scan that too and let the draft show you any line where the price moved. Once a week, glance at what changed and decide, on purpose, whether the shelf price moves with it. That last step is the one with money in it. A price rise you know about and absorb is a decision. A price rise you never saw is just a leak, and my 30-day GP plan starts with exactly this kind of leak because it is the easiest one to stop.