A spreadsheet is the right inventory tool for a small business until one of four things happens: a second person needs to edit it, stock starts changing without anyone typing (you make things, or you sell through a till), you need to know which lot something came from, or you need it on a phone where there's no signal. Until then, keep the spreadsheet and make it good. After one of those four, inventory software stops being a luxury, because the spreadsheet is no longer telling the truth on the day you open it. Here's the comparison, what each one costs, and a way to decide that isn't "we've grown".
The short version
- A spreadsheet is free, immediate, yours, and correct as long as one person keeps it and nothing changes stock without them knowing.
- Software costs a monthly fee and a week of setup, and its number is right because the work that changes stock (receiving, making, selling, counting) is the same work that updates it.
- The four moments above are the switch. Not a turnover figure, not a birthday.
- Compare on five questions, below. Price is the fifth.
- Whichever you choose, make sure everything exports. Always.
The honest side-by-side
| Spreadsheet | Inventory software | |
|---|---|---|
| Cost | Nothing, plus your time | A monthly fee; ours starts at $49 |
| Time to start | An hour | A week to import lists and learn it |
| Who can edit | One person safely, two with luck | Everyone who should, with roles, and a history of who changed what |
| On a phone in the walk-in | Awkward; offline takes setup | Should be built for it; test that it is |
| Stock goes down when you make something | No, you subtract by hand | Yes, from the recipe, by lot |
| Which lot, which date | One row per lot, kept by hand | Recorded as stock comes in and goes out |
| Mistakes | Formula errors and overwrites, found late | Fewer places to type; a record of who did what |
| Reports | You build them | Built in, if they're the ones you need |
| Your data | A file you own | Yours only if it exports; check before you buy |
The error problem, fairly stated
The best-known research on spreadsheet mistakes is Raymond Panko's at the University of Hawaiʻi. Summarizing seven field audits of real business spreadsheets, he found "errors in 24% of the 367 spreadsheets audited", and noted that the more recent audits, which used better methods, "found errors in at least 86% of the spreadsheets audited"; audits that went cell by cell found that 1.1% to 2.5% of cells held an original error (Panko, 2000). His point wasn't carelessness. People make undetected errors in a few percent of complex steps whatever the tool, and a spreadsheet gives you a great many steps and no one checking them.
Apply that to an inventory sheet with 200 rows and six formula columns, which is 1,200 formula cells. At one percent, a dozen of them are wrong, and the way you find out is a count that disagrees with the sheet by a number nobody can explain. Software has bugs too. The difference is that the same bug hits every customer and gets fixed once, and your spreadsheet's mistakes are yours alone, with nobody testing them. That's my reading of the research, and I'd hold it loosely.
What the time costs
Suppose keeping the sheet honest takes an hour a week: updating quantities, adding deliveries, subtracting batches, fixing the row someone typed over. That's 52 hours in a year. At $20 an hour it's $1,040. Our Starter plan is $490 for a year paid up front, which is two months free, or $588 paid monthly. On those numbers the software is cheaper in year one and the hour comes back.
But your hour may be free, because it's yours and it's 9 p.m. And software takes its own hour a week for the first month, while you learn it and import things. So I'd treat the cost comparison as a tie and decide on the four moments instead. Where the arithmetic changes is stock you throw away: one expired case a month that a system would have flagged costs more than any subscription, and the expiry math is where that shows up.
Five questions to compare on
- What changes your stock, and does the tool see it happen? Deliveries, batches, sales through a till, transfers to a market stall, waste. Every one the tool can't see is a number you type, and a number you type is a number you'll forget.
- Who touches it, and where? Two people at once, a phone in the freezer, a co-packer's building.
- What do you need answered in under a minute? What expires this week. Which lots used that delivery. What to order tomorrow. A sheet answers what it was built to answer; nothing else.
- What happens when you leave? A spreadsheet is a file. Software should export every list to CSV, and if it doesn't, the price is beside the point.
- What does it cost in the second year? The fee, plus the hour, minus the waste.
Make the spreadsheet good first
If you're staying with a sheet for now, and most businesses under a certain size should, these are free and they work: a bakery inventory list, a coffee roaster's sheet with roast loss built in, an inventory list with expiration dates, a recipe costing template, and a lot tracking spreadsheet. Three habits keep any of them honest: one row per lot for anything with a date, a "last counted" column, and a printed count sheet from the same file so the count happens on paper and gets typed once.
Where this stops working
The four moments, concretely. The second editor: two people open the sheet, one saves over the other, and the count from Saturday is gone. Production: Tuesday's batch used four pounds of butter and nothing subtracted it, so by Friday the sheet says there's a case that isn't there. Lots: the supplier's recall notice names a lot number and the sheet has a single row that says "purée, 18 lb". No signal: the count gets written on a pad in the walk-in and typed in later, or not. How to count inventory is the routine that holds a spreadsheet together, and it's also the routine that shows you which of the four has arrived.
How TaroStack does it
TaroStack starts from your spreadsheets: items, recipes, suppliers, customers, opening stock, price lists and open orders import from CSV, with a preview of every row before anything is saved, and every list exports back to CSV whenever you like. Stock changes when the work happens: receiving a delivery, recording a batch (which takes ingredients from the lots that expire first and notes which ones), picking an order, counting on a phone with no signal in the walk-in. Roles limit who can change what, and every change says who made it. Starter, at $49 a month, covers stock, counts, receiving, lots, expiry and recalls; Standard adds recipes and planning. If you'd like a hand setting up, ask, and we do it with you.