Replay failed Zap runs

If you're on a paid plan, you can manually replay any Zap run with steps that Halted or Errored.

• Go to Zap History.
• Select Zap runs to replay individually or in bulk. You can use the dropdown menus to filter Zap runs by status, date range, owner, folder, app, or Zap name.
  • Individual: select the checkbox next to the Zap runs you want to replay.
  • Bulk: in the upper left, click the checkbox dropdown menu and select Select shown to select all the Zap runs on the page, or Select all (5,000 at a time) to select up to 5,000 Zap runs.
• In the upper left, click Play [x].
• In the upper right, the notifications menu will open to indicate that the Zap run is replaying. Here, you can monitor the status of your replay job. Once it's complete, you'll see a summary of the results.

If you're on a Professional plan or higher and enable Autoreplay, Zapier will attempt to retry any steps in a Zap run that fail due to temporary errors or downtime. Autoreplay will retry the step again immediately, and then a few more times if there is still an issue.

• Go to Zap History.
• In the upper right, click to toggle the Autoreplay switch on.

How Autoreplay works

• If a step in a Zap run fails, Zapier will retry the step up to 5 times. While Autoreplay is active:
  • No error emails will be sent.
  • No Zapier Manager Zaps will be triggered.
  • You cannot manually replay the Zap run.
• Zap runs that are being replayed (or scheduled to be replayed) show up as "Scheduled" in Zap History.
• Zapier backs off after each successive failed retry attempt on this schedule: 5 minutes, 30 minutes, 1 hour, 3 hours, 6 hours. This means that if the Zap errors at 1:00 PM, Autoreplay will try again at 1:05, 1:35, 2:35, 5:35, and 11:35. The back-off schedule implies the final retry happens approximately 10 hours, 35 minutes after the first error.
• If you edit your Zap after the original error occurs, any future Autoreplay attempts will consider those changes. This gives you the chance to update your Zap to help with preventable errors.
• When Autoreplay attempts to retry a step, it checks the state of the Zap (this is how it considers Zap changes). If the Zap is off, or if Autoreplay itself has been switched off, the retry attempt will not run and all remaining retry attempts will be cancelled.
• If you turn your Zap off between retries, it won't affect the retry schedule as long as the Zap is turned back on before the next retry attempt.
• Only Errored steps can start Autoreplay, but any Halted or Errored steps in the Zap will be replayed when the attempt occurs.
• If Zapier cannot successfully replay the Zap run after the last attempt (approximately 10 hours later):
  • You'll be sent an email notifying you of the error.
  • Zapier will trigger any Zapier Manager Zaps you might have.
  • This is the only notification about the error Zapier will send if you enable Autoreplay.

Zapier may not be able to replay a step in a Zap run for any of the following reasons:

• The Zap is off: turn on your Zap to replay these steps.
• The Zap has been deleted: steps associated with a deleted Zap will not replay.
• The Zap has changed too much: steps associated with a Zap that has changed significantly will not replay.
• The step is too old: steps that are older than 2 months will not replay.
• The step has already run successfully: steps that have already run successfully will not replay.
• Filter and path steps are never replayed. The original data used to evaluate whether filter or path conditions were met will remain the same even if data from previous steps changes after replay.
• The step is already scheduled to replay: steps that are already scheduled to replay cannot be manually replayed.
• The Zap runs are marked as Held: the Autoreplay feature won't replay Zap runs with a Held status. This status flags potential issues like webhooks breaking or sending more data than expected, and will need to be manually replayed.
• Filter and path steps are never replayed. The original data used to evaluate whether filter or path conditions were met will remain the same even if data from previous steps changes after replay.