ShelfCount vs a spreadsheet: when the sheet is enough, and when it quietly isn't
Most people shopping for inventory software aren't leaving another product. They're leaving a Google Sheet, or a clipboard that gets typed into a Google Sheet on Monday. I ran my own shop that way for a long time. This is an honest account of what the sheet does fine and where it stops.
What the spreadsheet is good at
It’s free, it’s already open, and everyone on your team can use it without being taught. You can add a column at eleven at night because you thought of something. No vendor can change it, deprecate it, or raise the price on it. The file is yours and it will still open in ten years.
Those are real advantages and I’m not going to talk you out of them. If your sheet is working, keep it.
When a sheet is enough
- You track a few dozen items rather than a few hundred
- One person does the count, start to finish
- Your menu is small and you priced it once, and it hasn’t moved
- An order list gets you what you need, without a food-cost percentage
- Vendor prices are stable enough that stale costs don’t hurt
Where it stops
It can’t tell you what you should have used
This is the one that matters. A sheet knows two things: what you bought, from your invoices, and what’s on the shelf, from your count. Subtract and you get what disappeared. It has no way to know what should have disappeared, because that number lives in your sales.
Say you sold 900 lattes and each one takes ten ounces of milk. That’s 9,000 ounces, roughly 70 gallons. Your invoices say you bought 78. Those eight gallons are your variance, and that gap is the only thing that tells you whether you have a portioning problem, a waste problem, a comping problem, or somebody walking out with milk. Getting there in a spreadsheet means pasting item-level sales for the period, expanding every recipe to its ingredients, handling every modifier, converting every unit. Every week. By hand. And when a paste lands one row off, nothing warns you.
Nested recipes turn into a web nobody wants to touch
Vanilla syrup is sugar, water and vanilla. A batch of it has a cost per ounce. That per-ounce cost belongs in the price of every drink that uses it, so when the sugar price moves, sixteen drinks should move with it. In a spreadsheet that’s a lattice of cross-references exactly one person understands, and it breaks the first time somebody inserts a row.
The unit maths is the quieter version of the same problem. You buy sugar in fifty-pound bags and use it in grams. You buy milk by the gallon and pour it by the ounce. Going from millilitres to grams needs a density for that specific ingredient, and a spreadsheet will happily multiply by one and hand you a plate cost that looks perfectly reasonable and is wrong.
Two phones, one walk-in, bad wifi
A shared sheet on two phones in a cold room is how counts get lost. One person overwrites the other, or the tab drops and takes the last twenty minutes with it. Counting properly means each person picking up a section, every row landing against the location it was counted in, and the draft surviving a phone going to sleep in an apron pocket.
Invoice prices go stale without telling you
Costs in a spreadsheet are as current as the last time somebody typed them. The case price that went up fourteen percent three deliveries ago is still sitting in your sheet at the old number, quietly flattering every recipe it touches. Scanning invoices is less about saving typing than about the price history you get as a side effect.
Ordering stays a separate job
A count and an order are the same information looked at twice, but in a sheet they live in different places. You finish counting, then open another tab, compare against the pars you keep in your head, and type out an email per vendor. Which is fine on a quiet Tuesday and falls over the week your closer is on holiday and somebody else has to do it. Par levels that live in the same system as the count turn a finished count into a purchase order per vendor, ready to send, with nobody guessing at what “we usually get about four” means.
Side by side
| ShelfCount | A spreadsheet | |
|---|---|---|
| Cost | $50/mo flat | $0, plus your hours |
| Getting started | Upload your item list, or $299 to have it done for you | An afternoon building it, then maintaining it forever |
| Two people counting at once | Yes, and every row is attributed to the location it was counted in | Technically yes. In a walk-in on bad wifi, in practice no |
| Resuming an interrupted count | Drafts save as you go | Whatever survived the tab closing |
| Pack sizes and unit conversion | Set per item, including a density when you need to go from millilitres to grams | Whatever formula you typed, and silently wrong when the units drift |
| Nested prep recipes | Batch costs roll up into the drinks and dishes that use them | Cross-sheet references that exactly one person understands |
| When an ingredient price changes | Everything downstream recosts on its own | You update it, if you notice |
| Invoice entry | Photograph or forward it, then approve what it read | Typing |
| Vendor price history | Kept per item, automatically | Only if you built it and kept building it |
| Theoretical vs actual COGS | Sales sync from Square or Clover, run through your recipes | Not realistically, and this is the big one |
| Purchase orders | Generated from your par levels and emailed to the vendor | Copy, paste, write the email |
| Works with no internet | No | Yes, if it is Excel on a laptop |
| Who owns the data | You. Export to CSV whenever you want | You |
The honest cost comparison
$0 against $50 isn’t the real comparison. Count the hours the sheet actually takes each month: the count itself, the typing-up, the recipe maintenance, that Sunday night you spent working out why the number looked wrong. Put your own hourly figure against them, then decide. For some operations the sheet still wins that arithmetic, and that’s a fine answer.
What usually triggers the switch
In my experience it’s rarely a feature. It’s a moment. You finally calculate food cost percentage and can’t explain the number. Somebody else has to do the count now, and the sheet only makes sense to you. A vendor raised a price and you found out two months and forty cases later.
You can keep the spreadsheet
None of this is a trapdoor. Upload your item list to start, export everything to CSV whenever you like, and cancel in the app if it isn’t earning its $50. The trial runs 14 days and nothing is charged until it ends.
Questions people actually ask
- Is a spreadsheet good enough for restaurant inventory?
- For plenty of operations, yes. If you track a few dozen items, one person does the count, your menu is small and stable, and you mostly need an order list rather than a food-cost number, a sheet does that job and costs nothing. Software you resent is worse than a sheet you keep up.
- Can a spreadsheet calculate theoretical food cost?
- In principle. You'd need item-level sales for the period, every recipe expanded to its ingredients, every modifier accounted for, every unit converted correctly, then all of it reconciled against counts and invoices. For a menu with modifiers that's thousands of rows a week, entered by hand, and the errors are silent. Nothing tells you the number is wrong. That's the gap software closes.
- Do you have a free inventory spreadsheet template?
- Not yet. Free templates and calculators are on the list. In the meantime the trial is 14 days: you add a card at signup, nothing is charged until it ends, and you can export everything to CSV at any point, including on your way out.
- Can I import my spreadsheet into ShelfCount?
- Yes. Upload your item list as a CSV and it becomes your items, with pack sizes and vendors attached. If your sheet has grown strange over the years, and most of them have, $299 buys the done-for-you version: we set up items, vendors, par levels and recipes, then hand you a verification report to approve before go-live.
- What if I want to go back to a spreadsheet?
- Export to CSV and cancel in the app. It's month to month, there's no notice period, and your data comes out in the same shape it went in.
Compare something else
You don’t have to book a call to see this one. The demo is the real product with a cafe’s data already in it. Open it, poke around, close the tab.