Switching from a spreadsheet to stocktaking software: what breaks, what it costs, and when not to bother

I ran my stocktake on a spreadsheet I built myself. For years. So this isn't a vendor telling you spreadsheets are dead, it's a licensee telling you exactly which four jobs mine couldn't do, what switching actually costs in time and money, and the honest cases where the spreadsheet should stay.

Every software company writing about this has the same problem. They've never done a Sunday night count with a biro behind their ear, so the article is always "spreadsheets bad, our dashboard good" with a testimonial from a 16-site restaurant group underneath. That's not your pub and it wasn't mine.

My spreadsheet wasn't even a bad one. Formulas on every line, a fresh sheet for every count. I was proud of it. It also hid problems from me for years, and I only found out how much when I started weighing bottles instead of guessing them.

What a spreadsheet does well, and keep doing

Fair's fair first. A spreadsheet counts packaged stock perfectly. Bottled beer, cans, post-mix boxes, crisps if you're tracking them. Count the units, type the number, the sum works. It costs nothing you haven't already paid, and nobody can put the subscription up on you.

If you run a small site with a handful of lines, you count everything yourself, and you're happy judging the business on stock value alone, you can stop reading. Honestly. Do the count on paper, type it up, and put the software money behind the bar. My DIY stocktake guide works fine with a spreadsheet and I wrote it knowing that.

The switch is worth it when you need answers a grid of typed numbers can't give. There are four of them.

The four jobs my spreadsheet couldn't do

1. Tell me what's in a part-used bottle

The spreadsheet takes whatever number you type. And the number you type for an open bottle of vodka is a guess, because your eye is reading "about four tenths" off a bottle that's actually nearer three. Multiply that kind flattering guess across forty open spirits and the spreadsheet's valuation is out by more than you'd think, in the direction that hides loss.

The fix isn't a better spreadsheet, it's a different measurement. Bottle on a scale, software that knows the empty and full weights, contents in millilitres. A spreadsheet can't hold a tare table for every product and do that conversion at the speed of a count. This one job is most of the reason I stopped.

2. Tell me whether draught loss is froth or a leak

British Beer and Pub Association guidance says a pint served with a head only needs to be 95% liquid. So up to 5% of every draught pint can leave the keg without ever ringing through the till as volume, legally. Add line cleaning and pull-through. A spreadsheet doesn't know any of that, so every count it shows draught "missing" and you're left staring at good staff wondering.

Software with proper keg and cask calibration separates the froth from the real variance. The calibration guide covers how that works. Without it, your draught numbers are noise.

3. Do the GP maths ex-VAT without me getting it wrong

The classic spreadsheet mistake, and I made it: calculating GP% on till prices with the VAT still in. It flatters your margin by around seven points, and you feel great about a number that was never real. The formula fix is one cell, but you have to know to do it, remember you did it, and never break it when you copy the sheet for next month. I've written the full working here and the same trap wrecks stock valuations too.

Which points at the quieter problem: spreadsheet formulas break silently. One sorted column or one dragged cell and the total still looks like a total. Nothing flags it, because nothing checks a spreadsheet except you, at midnight, when you're least likely to catch anything.

4. Tell me which products the money is leaving through

A spreadsheet can tell you stock value went down. What I actually needed was variance against the till, in pounds, per product: expected sales against counted usage, worst lines first. That means holding sales data, cost prices, sell prices and counts together and reconciling them every time. You can build that in a spreadsheet. I tried. It becomes a second job, and knowing what a normal gap looks like is hard enough without also maintaining the machine that measures it.

What switching actually costs

Here's the part the vendor articles skip, because it's the unsexy bit.

The subscription is the visible cost. UK-relevant apps run from under £20 a month to about £139 per site, and I've listed every published price I could find in what pub stocktaking software costs. Against that, remember your spreadsheet isn't free either. The typing-up stage after a paper count is an hour or more you don't get back, and at the £12.71 living wage even your own time counting and reconciling is real money every month. The payback maths guide works that through properly.

The real cost is setup. Nobody escapes it. Every product needs its size, cost price and sell price loaded, and for weighing to work the software needs bottle weights. For a typical 200-line pub that's an afternoon of solid graft. Some of it transfers straight from your spreadsheet, which is one more reason not to delete it. Budget the afternoon honestly and it stops being a nasty surprise.

And your first count on the new system will be slower than your hundredth on the spreadsheet. You know your spreadsheet. You don't know the app yet. From the second count it flips, and it keeps flipping: there's no typing-up hour, the maths is done when the last line is counted, and the report is waiting before you've locked up. Real per-line timings are in how long a stocktake takes.

Spreadsheet versus software, job by job

The honest version of this table. Where the spreadsheet wins, it says so.

The job	Spreadsheet	Software
Counting packaged stock	Fine	Fine, slightly faster
Part-used spirits	Eyeballed tenths, flattering	Weighed to the millilitre
Draught, froth and line cleans	Can't separate them	Keg/cask calibration
GP% ex-VAT	Right only if you built it right	Done for you, every time
Variance in £ per product	A second job to build	The core report
Formula errors	Silent until they cost you	Not your problem
Counting in a no-signal cellar	Paper, then typing up	Offline on your phone, syncs after
Anyone else can count	Only you understand the sheet	Staff count on their own phones
Monthly cost	£0 plus your hours	Under £20 to £139 a site
Ongoing setup	None	An afternoon, once

How to switch without gambling on it

Don't take my word or anyone else's. Run both, once.

Export your spreadsheet and keep it forever. It's your history, and a decent app will let you use an old count as a baseline to compare against. Then load your top thirty lines by value into whatever you're trialling, not all 200, and do one real count both ways. Weigh the spirits in the app, eyeball them for the sheet, and look at the two answers side by side. The gap between the weighed number and your guessed number is the argument, one way or the other, made with your own stock.

If the software doesn't beat the spreadsheet on that count, in your cellar, on your phone, don't switch. Cancel the trial and keep the sheet with a clear conscience. A trial that wants card details before you can do that comparison is answering a different question for you, and the free routes worth trying first are in what free stocktaking software actually gets you.

One thing I'd say from the other side of it. The spreadsheet made me feel in control because I'd built it. The scales made me actually in control, because for the first time the numbers weren't my own guesses typed into my own grid, marking my own homework. That difference took me years to see. It shouldn't take you that long.