Guides › Alerts that get acted on
Making inventory alerts get acted on
Most inventory alerting does not fail because the software was wrong. It fails because the emails kept arriving and people stopped reading them. That is worse than having no alerts, because you believe you are covered.
Last reviewed 16 August 2026
On this page
How alerts die
The pattern is consistent and it is always the same shape.
- Alerts are switched on with one threshold across the catalogue.
- Slow moving products sit under it permanently and email every day.
- People learn that most alerts need no action, which is a correct inference from the evidence.
- A filter gets created. The folder stops being opened.
- The one alert that mattered arrives into that folder and something runs out anyway.
The rule. Every alert that needed no action is a small withdrawal from the credibility of every future alert. Precision is not a nicety here, it is what keeps the system alive.
Precision beats coverage
The instinct when setting up alerting is to cover everything. It is the wrong instinct at the start.
Begin with the products where running out genuinely hurts. Best sellers, long lead time imports and anything with committed customers. Set those thresholds carefully. Leave the long tail unmonitored for now.
A system that alerts on 40 products and is right almost every time gets read. One that alerts on 4,000 and is mostly noise gets filtered within a fortnight. You can always widen coverage once the alerts have earned trust. Regaining trust after it is lost is much harder.
Who should receive them
Two failure modes, opposite directions, both common.
Only one person. The system has a hole in it the shape of that person's annual leave and peak season and holidays overlap more often than anyone plans for.
Everybody. Diffusion of responsibility. An alert that arrives for eight people is an alert nobody specifically owns. It is remarkable how reliably all eight assume somebody else has it.
The workable shape:
- One clear owner who is expected to act.
- One backup who covers absence.
- A summary to whoever carries the commercial consequence, as a digest rather than as interrupts.
Where alerts differ by location, route each location to whoever can actually act on it. See alerts across every location.
What the alert should contain
A good alert is actionable without opening anything else. It should carry:
- Product and variant name, not only a SKU code
- Current quantity and the threshold it crossed
- The location, by name rather than an ID
- A direct link into the product in Shopify admin
The test is whether somebody reading it on a phone can decide what to do without logging in. Alerts that require investigation before a decision get postponed, which is functionally the same as ignored.
Instant alerts and digests do different jobs
Instant alerts catch sudden drops, which is the case where hours matter. Reserve them for products where they genuinely do.
A daily summary catches slow drift and gives one view of everything currently low. It is also the format that survives a busy week, because it is one predictable message rather than an unpredictable stream.
Most stores want instant alerts on a narrow set of critical products and a digest for everything else. Instant alerts on the whole catalogue is how the filter gets created.
Review what the alerts told you
Once a quarter, look back at two things.
- Stockouts that happened without an alert. The threshold was too low or the product was not monitored.
- Products that alerted repeatedly without ever running out. The threshold is too high and it is spending your credibility.
Adjusting on that evidence converges on a genuinely good configuration within a couple of quarters. Nobody gets thresholds right at the first attempt and the setups that work are the ones that got corrected.
Alerts built to stay credible
Stockwell holds a threshold per SKU so slow movers stop crying wolf, names the product and the location in the email and offers a daily summary alongside instant alerts.
Get Stockwell on the Shopify App Store Free plan covers 50 SKUs with a daily summary. Paid plans add instant alerts and more recipients. No card needed to start.Common questions
Who should receive low stock alerts?
One clear owner expected to act, one backup for absence and a digest to whoever carries the commercial consequence. Alerts to a single person fail during leave. Alerts to everybody mean nobody specifically owns them.
Why does my team ignore inventory alerts?
Almost always because too many of them needed no action, usually from one threshold applied across the whole catalogue so slow movers alert permanently. People correctly infer that most alerts are not urgent, then filter all of them.
Should I use instant alerts or a daily summary?
Both, for different products. Instant alerts on the narrow set where hours matter, a daily digest for everything else. Instant alerts across the whole catalogue is what causes people to filter them.