Article · 6 min read

Low stock alerts on Shopify: why one threshold is wrong for every product, and what to reorder instead

22 October 2026 · by njX

The usual low-stock alert is a number: tell me when a product drops under 10. For the product that sells 8 a day it fires the afternoon before the stock-out; for the one that sells 2 a month it fires every quarter for nothing. Here is why a fixed threshold is wrong for almost every product, how to turn sales into days of stock, where the reorder point and the quantity come from, and what a weekly list is worth.

Why “under 10 units” fails

A threshold in units says nothing about time. Ten units is two days for a bestseller and a year for a slow seller. The useful question is not “how many are left” but “how many days are left, and how long does a delivery take”. The answer is different for every product and changes with the season, so it has to be computed, not typed.

Days of stock

Take the last 60 days of sales for the variant, divided by the days it was actually on the shelf, not the days it was sold out. That is the rate. Stock on hand divided by the rate is days left. Sales are a count, not a constant, so days left is a range: a product that sold 3, 0, 5 and 1 in four weeks does not have “27 days left”, it has somewhere between three and five weeks. A number that pretends otherwise is the second most common reason people stop trusting their alerts.

The reorder point and the quantity

Reorder point: the supplier’s lead time plus the days you need to notice and place the order. If a delivery takes 14 days, a product with 14 days left is not “low”, it is late.
Critical: a second line, closer to the stock-out, for the products that need a call rather than an e-mail.
Quantity: how many days you want to buy for, times the rate, minus what is on hand and on order. Buying for 60 days on a product that moves 2 a day is 120 minus the stock, not “a lot”.
No history: a new product has no rate. For those the alert has to be in units, once, and the rate takes over as sales arrive.

Alerts versus a list

An alert per product is noise after the first week. What a buyer actually uses is one list a week, grouped by supplier, with the products to order from each one and the quantity next to them, and an alert only when a product crosses the critical line between two lists. The list is also where the opposite problem shows up: the products with 400 days of stock, the money sitting on the shelf, which no low-stock alert will ever mention.

Doing it without a spreadsheet

njX Stock Alerts & Reorder reads your products, inventory by location and the last 60 days of orders, and answers for every product in plain days: how many are left and how much to order, as ranges, with the four steps of the arithmetic shown on the row so you can redo it by hand. You set two lines in days, reorder and critical, and a supplier lead time per vendor on Pro; pinned products are judged against their supplier’s lead time. One list a week by vendor goes to e-mail, Slack or Discord, an alert only on a crossed line, and six short lists show what earns the most, what is overstocked and what comes back. The app is read-only: it shows the numbers, you place the orders.