How to see what runs in the background

Automations shows every background job your workspace runs as a stack of cards, one per job. This page is the tour of what they are. See What are automations? for what an automation is.

Which jobs alert you?

These are the ones worth configuring, because they are the ones that speak.

• Low stock alerts: tells you when a product's shippable stock drops to five or fewer.
• Fulfillment backlog alerts: tells you when unfulfilled paid orders pile up past a number you set.
• Notable order alerts: tells you when an unusually large order arrives.
• Stale-data watchdog: tells you when your store data stops refreshing.
• Task due reminders: buzzes an assignee as their task's due date approaches.
• Daily analyst brief: writes the morning briefing.

Setting these up is How to set up the stock and order alerts and How to switch on and read the daily brief.

Which jobs just keep things fresh?

These need no configuration and have no threshold or recipient list, because nobody is being told anything.

• Customer sync: pulls customers and their stats across from Klaviyo.
• Shopify change watcher: notices when things change in your store.
• Daily metrics snapshot: pre-computes the dashboards so they open instantly.
• Calculated field refresh: recalculates the fields that work themselves out. See A field that works itself out.
• Outcome verdicts: measures committed outcomes when their date arrives. See How to commit to an outcome and have it measured.

Leave these on. When a dashboard looks stale or a calculated field has not updated, this is where to look first.

What does the line under each title tell you?

A one-line health summary, which is the fastest read on the page:

• Never run: it has not run yet, usually because it was only just switched on.
• Last run 12 Aug, 3 found, 3 notified: it worked and told three people.
• Last run 12 Aug, no alert: it ran and there was nothing to say. This is the healthy everyday state, not a fault.
• Last run 12 Aug, failed with a reason: it went wrong, and the reason is named.

Found, notified and emailed are three different numbers: what the job saw, how many in-app alerts it sent, and how many emails went out.

How often does each one run?

Each job has its own fixed cadence, from every few minutes to once a day. There is no schedule picker, because the right frequency for watching stock is not the right frequency for writing a morning brief.

Run now at the foot of a card makes a job due immediately rather than waiting. See How to turn an automation on, off, or run it now.

Why does a card show a name with no description?

A handful of jobs display their raw internal name with no explanation and no controls. Those are upkeep lanes rather than anything you configure, and there is nothing to set on them.

Ignore them. They are working if the things they feed are working.

Why is there an amber banner across the top?

Because automations are switched off for the whole environment. The banner says everything below is configured but nothing will run.

That is a deployment-level setting rather than anything about your workspace, so switching individual cards on will not change it. It is the first thing to check when nothing has fired at all.

Why did a job record a skipped run?

Because it had nothing to work with, and skipping is the honest outcome.

The two common reasons are named in the run itself. A module is switched off, so its automations are dormant, per How to switch parts of Tideswell on and off. Or the data source is not connected yet, meaning Shopify or Klaviyo, per How to manage integrations, your workspace name and the audit log.

A skipped run is not a failure, and it is not lost work either. See How to check whether an automation worked.