Inventory is one of those things nobody thinks about until it bites them. A product goes viral, you sell 60 units of something you had 40 of, and now you are emailing apologies and refunding orders on your best sales day of the year. Or the opposite happens: you tied up eight thousand dollars in a "sure thing" that is still sitting in the back room ten months later.
Both problems come from the same root cause. You did not have a clear, trusted picture of what you actually own and how fast it moves.
Here is the setup I walk merchants through, roughly in order.
Get your counts honest before anything else
Every inventory tool, forecast, and automation is built on top of your on-hand numbers. If those numbers are wrong, everything downstream is wrong too, just with more confidence.
So start with a real count. Not a guess, not "roughly what the system says." Pick a quiet day, count what is on the shelf, and reconcile it against Shopify (Products, then the Inventory tab). You will almost always find discrepancies: damaged units that were never written off, returns that went back into stock but not into the system, a bundle that was quietly eating components.
Fix the counts once, properly, and then protect them. Which brings us to the settings.
Turn on the guardrails Shopify already gives you
A lot of oversells happen because people never touch the defaults.
In your product and variant settings, "Track quantity" needs to be on for anything you actually count. And unless you genuinely sell made-to-order, leave "Continue selling when out of stock" switched off. That one checkbox is the difference between a clean "sold out" label and a cart full of orders you cannot fulfill.
If you sell in more than one place (a physical shop, a market stall, wholesale, plus the website), set up your locations properly under Settings, then Locations. Shopify can then track stock per location instead of pretending everything lives in one magic pile.
The two numbers that actually run the show
Once your counts are trustworthy, inventory stops being about "how many do I have" and starts being about "when do I run out."
Two numbers answer that:
Sell-through rate. How many units you move per week for a given product. Not the average across your whole catalog, which is meaningless. Per product, or at least per key product.
Lead time. How long it takes from the moment you place a reorder to the moment it is back on your shelf and sellable. Include everything: supplier turnaround, shipping, customs if you import, and the day it sits in a box before someone puts it away.
Multiply weekly sell-through by lead time in weeks, add a little safety buffer, and you have your reorder point. When stock drops to that number, you order. That is the whole game. Everything fancier is a refinement of this.
When spreadsheets stop being enough
For a small catalog, a spreadsheet genuinely works. I have seen stores run to a few hundred thousand a year on a well-kept sheet and a calendar reminder.
The wheels come off when a few things stack up at once:
- You cross a couple hundred SKUs and cannot eyeball them anymore.
- You have variants and bundles, so one sale should decrement several component counts.
- You reorder from multiple suppliers with different lead times and minimums.
- You have more than one stock location.
- More than one person touches inventory, so the sheet is always slightly out of date.
At that point the spreadsheet is not saving you money, it is quietly costing you money in oversells, rushed reorders, and cash parked in the wrong products.
Where a dedicated tool earns its place
This is where inventory software stops being a nice-to-have. A proper tool ties the sell-through and lead-time math together automatically, tells you what to reorder and when, generates purchase orders you can send straight to suppliers, and keeps counts synced across locations without anyone updating a sheet by hand.
If you are shopping for one, Stockra is built for exactly this on Shopify: demand forecasting, purchase orders, and multi-location stock in one place. Whatever you pick, the questions to ask are the same. Does it forecast per product or just show current stock? Can it handle your bundles and variants correctly? Does it generate real purchase orders, or just alerts? And how painful is getting your data in and out?
A weekly routine that keeps it boring
Boring is the goal. Fifteen minutes, once a week:
- Review anything below its reorder point and place the orders.
- Skim your slowest movers. If something has not sold in 60 to 90 days, decide now: discount it, bundle it, or stop reordering it.
- Spot-check a handful of counts against the shelf so small errors never grow into big ones.
Do that consistently and inventory becomes the thing you never think about, which is exactly what it should be. The stores that get burned are almost never the ones with a fancy system. They are the ones with no system at all.
