Plaid automatically retries failed webhook deliveries, but does not offer on-demand replay. Monitor webhooks are not special-cased: if your endpoint returns a non-200 status or doesn't respond within 10 seconds, Plaid keeps retrying delivery for up to 24 hours using exponential backoff (starting at 30 seconds, with each delay 4× the previous; see Webhook retries). However, there is no mechanism to request an on-demand replay of past webhook events, and webhooks will be lost if your endpoint is unavailable for longer than this 24-hour retry window. If that happens, reconcile the affected screenings via the API — all data delivered by webhooks is also available through Plaid's APIs.
To avoid this issue:
- Ensure your webhook endpoint returns a 2xx response quickly (before any long-running processing).
- Implement idempotent event handling so that re-processing the same event is safe.
- Periodically poll
/watchlist_screening/individual/listto catch any screenings that may have been missed by your webhook handler.