Understand automation structure
Understand how triggers, list steps, and action steps process repeated work.
An automation selects data at a configured time and applies work to each result. The dashboard calls the automation definition a workflow. A workflow has a name, trigger, and ordered steps.
Business value
Automation is valuable because it applies the same action under the same conditions. It handles the normal path and leaves failed runs for review. Staff can spend more time on customer service and exception decisions.
| Operating goal | Automation role | Measure |
|---|---|---|
| Reduce repeated work | Start work from a time or data change. | Manual changes and automated results |
| Apply one operating standard | Reuse the same query and action steps. | Branch differences and exceptions |
| Find missed or delayed work | Record each run and step result. | Failed, delayed, and repeated runs |
| Focus staff on exceptions | Automate normal cases and keep failures for review. | Review time and manual handling rate |
Examples
Example: Mark a no-show 30 minutes after a reservation starts. Select only confirmed reservations past their start time. Change each result to no-show. Review processed and failed counts. Confirm that the next run does not select the same reservation.
Example: Retire ended promotions. Select only active promotions past their end time. Deactivate each result. Confirm that the promotion leaves the public website. Record automated results and exceptions that needed manual work.
Three parts of an automation
- The trigger controls when the workflow starts.
- A list step selects the rows to process.
- An action step applies a feature to each selected row.
For example, a heartbeat can find confirmed reservations that started over 30 minutes ago. The
next step changes each selected reservation to no_show.
A safe automation removes processed rows from the next query. Changing confirmed to
no_show removes the row from the same condition. Read Runs and repeat safety
for the full model.