A polling trigger is the part of an automation that decides when to start, by checking a source on a repeating schedule. Every few minutes it asks ‘is there anything new?’, and the moment the answer is yes, the rest of the workflow runs. It is the most common kind of automation trigger in tools like Zapier.
Imagine a postman who walks past your mailbox every 15 minutes to see if you have an outgoing letter. If there is one, he takes it. If not, he keeps walking and comes back later. The schedule is fixed, so a letter you drop in just after he passes waits until his next round. That waiting time is the catch with polling: it is reliable but never instant.
To know what counts as “new”, a polling trigger remembers where it left off, usually by storing the ID or timestamp of the last item it saw. That bookkeeping is what stops it processing the same record twice, and it is also where bugs hide if a source reorders items or reuses IDs. A common failure looks harmless: a new row gets backdated, so it lands before the last marker the trigger checked, and the workflow quietly skips it for good.
The faster the trigger checks, the fresher your results, but very frequent checks use more resources and can hit rate limits on the source. When you need true real-time reactions and the source supports them, a webhook is usually the better choice.
At TopDevs we pick the polling interval to fit each workflow, so an order update lands fast while a nightly report does not waste checks all day.
