Trigger steps are the part of your automations that tell your Zaps when to run. When you are building a Zap, Step 1 is your trigger. There are two types of triggers, and they can help you to determine how and when your Zaps will run.

Polling Triggers#

Zaps operate on a “polling” basis. This means we check in with the application's server on a regular interval, and that interval is determined by your Zapier subscription plan (this will be either every 5 minutes or every 15 minutes). All the updates we receive within that 5- or 15-minute window will be processed the next time we poll for updates.

Timing and precision of Polling Triggers#

Your Zaps will poll approximately every 5 or 15 minutes. Polling is distributed across our servers in a queue, and polling times are not always exact. This means if you try to time multiple Zaps to happen in a specific order closely together, it may not happen as you expected.

You might also not be able to time your Zaps to happen at an exact minute. For instance, if you use Schedule by Zapier to schedule a Zap to run at 5 PM daily, it will run the first time your Zap polls after 5 PM, which could be anywhere between 5:00 and 5:15 depending on your Zap's polling cycle.

Can I control when my polling Zap runs?#

It is not possible to control exactly when your Zap runs due to the factors outlined above.

Instant Triggers#

You may see that some Zaps have an “Instant” badge in your dashboard that looks like this:

Example instant badge

When your Zap says it is Instant, this means it uses a webhook-based trigger. Webhooks are a way to immediately, pro-actively receive notifications. When we receive a webhook from another application, we process it immediately. Most often, your other service provider will send these webhooks to us promptly. (This isn't always guaranteed.)

Timing and precision of Webhook Triggers#

Since webhook-based triggers are not on a cycle, their timing should be fairly predictable. All the same, if you are sending a large number of webhooks within a few seconds of each other for example, we do distribute the load across multiple servers. If you expect webhooks that very closely follow each other will be processed exactly in the order in which they are sent, between network and processing factors this might not always be reliable.