The Inventory Nobody Wanted to Count

Picture a mid-sized distributor in the Quebec City region. Industrial parts, a solid customer base, a warehouse out back. And a warehouse nobody wanted to walk into with a clipboard.

Not because stock was missing. The opposite, really. There were boxes with labels from three years ago, parts kept “just in case,” and a back corner everyone jokingly called “the graveyard.” The operations manager knew it. The controller knew it too. Neither of them wanted to be the first to say out loud how much money was sleeping back there.

If that sounds familiar, you’re not alone. In a lot of small and medium-sized businesses, inventory isn’t really a spreadsheet problem. It’s a fear problem: fear of finding out you over-ordered, fear of an awkward conversation with a supplier, fear of your bank asking why the “inventory” line on your balance sheet never seems to turn into cash.

I learned this far from a typical small business, in large mining and energy operations at companies like Rio Tinto and Ecopetrol. There, the wrong spare part or a pile of obsolete stock doesn’t just hurt your margin. It can shut down a line. The lesson I brought with me applies just as well to a team of 15 or 150 people: what you don’t sort, you can’t decide on. And what you don’t decide on charges you rent every single month.

The real cost (no drama)

Let’s use a simple example. Say you carry CAD $80,000 in inventory. If 20% of it is dead or nearly dead, that’s CAD $16,000 that isn’t paying wages, isn’t funding marketing, and isn’t paying down your line of credit. At a cost of capital of 8%, that corner costs you about CAD $1,280 a year… just for not looking at it. Plug in your own numbers; the point is the habit, not my example.

Worse, it steals your attention. Every month-end, someone wrestles with numbers nobody fully trusts.

What AI does here (and what it doesn’t)

AI won’t magically clean out your warehouse. It won’t replace your business judgment either. What it does well are three things that take a person hours:

Sort quickly. ABC by value (your few big-ticket items versus the long tail) and XYZ by how predictable demand is.
Flag likely dead stock using clear rules that you set.
Let you ask questions in plain English, like: “Which items haven’t moved in 12 months and are worth more than $200?”

You still decide what to clear out, what to reclassify, and what to bring up with your supplier.

Your first step, in 15 minutes

All you need is an inventory export (Excel or CSV) from whatever you already use: QuickBooks Online, Sage 50, Acomba, your ERP, or a plain spreadsheet. At a minimum, include: item code, description, quantity on hand, unit cost, and the date of the last movement. If you don’t have that date, use “last sale” or “last shipment” instead.

Open ChatGPT, Claude, Copilot, or whichever assistant you already use, and paste this (adjust the thresholds to fit your business):

You are an inventory analyst for a small Canadian business. Below is a table with these columns: sku, description, quantity, unit_cost (CAD), last_movement (YYYY-MM-DD).
Today is 2026-10-13. Please do the following, in plain English, and do not invent any rows:
1) Calculate total_value = quantity * unit_cost.
2) Classify each item A, B or C by cumulative total_value (A ≈ 80%, B ≈ 15%, C ≈ 5%).
3) Flag "possibly_obsolete" if last_movement is more than 365 days ago or is blank.
4) Return a table with: sku, abc_class, total_value, days_without_movement, possibly_obsolete, suggested_action (review / clear out / keep / fill in missing data).
5) End with a 5-line summary for the owner: how much money (CAD) is tied up in possibly obsolete items, and the 3 actions to take this week.

If your file is large, upload the CSV and ask for the same analysis. If you’re uneasy about sharing real data, rename the item codes and use a sample of 30 rows. You’ll build the same muscle either way.

How you should feel when you’re done

Not “I’ve solved my inventory.” More like: “I now have a short list of decisions.” Fear shrinks when the problem stops being a fog and becomes three rows on a page.

The companion model (coming soon)

Smart Inventory (ABC/XYZ + dead stock): an Excel/Google Sheets template with colour-coded input cells, rules you can edit, a dashboard of items to clear out, and ready-to-use prompts. Planned price: CAD $49. In the meantime, the exercise above is free, and it’s yours.

Note: this doesn’t replace advice from your accountant or tax professional, or a proper inventory audit. It’s a way to think clearly.

For the community

If you try the prompt, reply with just one thing: how much did your thermometer drop, from 1 to 10? That’s how we’re building the FluxLogic SI supporters’ wall: not with likes, but with courage you can measure.