What your inventory spreadsheet actually costs

The spreadsheet is free, and that's the whole problem with arguing about it. Anyone who wants to sell you something has to claim it's costing you money, which is exactly what a vendor would say, so the claim gets discounted to zero and the conversation ends. Fair enough.

So here's the arithmetic instead. It costs nothing to run, we're not going to put a number in it for you, and if it comes out small then it comes out small — that's a real answer and you should keep the spreadsheet.

Why we think this is the right question

Of the 482 posts in our corpus that described a real, unresolved inventory problem, 65 named a spreadsheet as the thing they were running on. That's the single most common answer to "what are you using now?" — more common than any software product.

Be careful with that 65, including with how we use it. It's a keyword match over free text, so it only catches people who happened to write the word. It undercounts, certainly. Treat it as a floor.

The four costs, and only one of them is time

Most attempts to cost a spreadsheet stop at data entry, which is the smallest of the four and the easiest to dismiss — everyone knows typing is fast.

• Searching. Time spent walking to a shelf that turns out to be empty, or looking for something the sheet says you have.
• Reconciling. The count at month end, plus every mid-month argument about whether the number is right.
• Errors that reach a customer. Short shipments, wrong items, the credit and the reship.
• The one person. What it costs that only one person knows where things really are — and what happens on the days they aren't in.

Measure the first one this week

Pick a normal week. Every time anyone goes looking for something and it isn't where they expected, they make a mark on a sheet taped to the wall. One mark. Nothing else — no form, no app, no explaining.

At the end of the week, count the marks and multiply by five minutes, which is conservative. Multiply by the loaded hourly cost of whoever was walking. That's your searching cost, per week, in your own money.

Nobody needs a study to know this number is bigger than zero. The point of measuring it is that you stop guessing which direction it's moving.

The second one you already have

Whoever does your month-end count knows how long it takes and how many lines came out wrong. You don't need a project for this — you need to write down two numbers you already know and stop treating them as the weather.

The third one is the one that actually decides it

Pull last quarter's short shipments and wrong-item credits. For each one, ask a single question: did the system say we had it in a place where it wasn't?

That subset is the cost of not knowing where stock is. Everything else — supplier problems, damage, genuinely being out of stock — belongs to a different problem and shouldn't be counted here. Counting it would inflate the number in our favour, which is precisely what makes this kind of arithmetic easy to dismiss.

And the fourth one has no formula

If one person is the index, the cost isn't their salary. It's the fact that the operation slows down measurably when they're on holiday, and that you can't hire into it — a new person can't learn a system that only exists in someone's head.

That one is a judgment call, and we're not going to pretend otherwise by putting a multiplier on it.

What to do with the total

Compare it to what a fix would cost, and be honest that a fix is not only software. Labeling your locations properly is free, takes a weekend, and removes a large share of the searching cost on its own. We've written that up separately, and you can do it without buying anything from anyone.

If your number comes out at a couple of hundred dollars a month, the spreadsheet is winning and you should leave it alone. Under about 300 SKUs it usually does — one person can genuinely hold that much in their head, and software adds a step without removing a problem.

Our bias, stated plainly

We sell warehouse software, so we went looking for a problem software solves. That's the bias we're least able to correct for ourselves.

The correction we can offer is that this method works entirely without us, and that two of the four costs are reduced most cheaply by labeling rather than buying. If you run the numbers and the spreadsheet wins, we'd rather you knew that than found out in month two.

One more thing worth knowing before you ask anyone for advice: when we tagged who was talking in these threads, 59% were vendors or consultants rather than people who pick orders. Including, to be fair, us.