Triggers: Data Added, Data Updated, Data Arrival Time
What each of the three triggers means, which column to point it at, which columns a trigger condition can reference, and how the before-and-after reminders on Data Arrival Time work.
Three triggers
Add Automation → Trigger When:
| Trigger | Fires when | Suits |
|---|---|---|
| Data Added | A row appears (import, paste, sync, form submission, formula) | A new-order alert, assigning a new customer |
| Data Updated | A row's value changes in the Field you chose | Status moving from pending to approved; a stock count edited |
| Data Arrival Time | The time in the chosen date column arrives (or a stretch either side of it) | Contract expiry, task deadlines |
Each of them wants a Field as its trigger data — Select Data Field, and Data Field Required if you skip it. That column is where the event has to happen.
Data Arrival Time
Pick a column whose field type is Date, then set the Reminder Time:
- On Arrival — at the moment itself.
- Before Arrival, N Minute / Hour / Day / Week / Month — early.
- After Arrival, N … — late.
Every row is counted on its own: 100 rows in a due-date column are 100 separate reminders. Rows whose time has already passed are not fired retroactively.
Trigger conditions: one more sieve
To fire only when "a new order is over 10,000", add a condition: a field, a comparator (>=, >, <, <=, =, !=, Contains, Does Not Contain) and a value, with And / Or between several. You can also switch to Using Formula and write a condition formula.
There is a limit on which columns a trigger condition can reference. Plain columns (with no formula) always work; a formula column works only if it is one the trigger column depends on. The reason is that the engine has to have a settled value at the instant the event happens, and a formula column may still be computing. So the most dependable pattern is to put the judgement into an if formula column first, then watch that column with Data Updated.
Choosing the trigger column
- Data Added: a column every row is guaranteed to have — the index, or an order number.
- Data Updated: the column whose change you actually care about. Don't pick one that changes whenever anything else does, or unrelated edits will fire it.
- Data Arrival Time: a Date column. A date written as text will not do; derive a real date column first.
Testing
Before you save there is Test Automation: it simulates a run against the current data so you can see whether the action fired and whether the email arrived. Test first, then flip the switch to Running.