Polling

Polling is an integration pattern where a system repeatedly checks a data source at fixed time intervals to see whether anything new or changed exists — "is there a new row yet? is there a new row yet? is there a new row yet?" — as opposed to the source system proactively pushing a notification the instant something happens (a webhook). In the no-code automation world, polling is the fallback trigger mechanism used whenever the source app doesn't support webhooks: "New Row in Google Sheets" triggers, for instance, are almost always polling-based, because Google Sheets has no built-in webhook system to notify external services of changes. Why it matters: polling has two direct, practical consequences for builders. First, latency — a polling trigger checking every 15 minutes means an automation can lag up to 15 minutes behind the actual event, which is unacceptable for time-sensitive flows (e.g., "instantly notify sales when a hot lead submits a form") but perfectly fine for others (e.g., "sync yesterday's completed orders to the accounting system each morning"). Second, resource cost — polling consumes API rate limit quota and platform "task" allowances even when nothing has changed, since the check itself counts as an operation regardless of whether new data is found, whereas a webhook only fires (and only costs anything) when an actual event occurs. How it works: a polling-based trigger stores a "checkpoint" — typically the ID or timestamp of the last record it saw — and on each check, queries the source system for anything newer than that checkpoint (e.g., `SELECT * FROM records WHERE created_at > '2026-07-01T14:30:00Z'`), processes any new results, and advances the checkpoint. Worked example — comparing polling frequency trade-offs for a Zapier "New Row in Google Sheets" trigger on the free tier (15-minute polling interval) versus a paid tier (1-minute polling interval): a sales team logging leads into a shared spreadsheet on the free tier might see a lead sit unprocessed for up to 15 minutes before the automation notices it and creates a CRM record — during a live demo day with time-sensitive follow-up, that lag could be the difference between a lead getting a same-day callback or not; upgrading to a paid plan with 1-minute polling (or better, switching the data source to something with native webhook support, like Typeform instead of a raw spreadsheet) closes that gap. Builders should always check whether a "trigger" is documented as instant/webhook-based or polling-based before relying on it for time-sensitive automations.