Outbound webhooks now use exponential backoff with jitter for retries, replacing the previous fixed-interval retry schedule.
Before: Retries at 30s, 60s, 120s intervals
After: Exponential backoff starting at 30s, with ±20% jitter to prevent thundering herd on recovery
Failed webhooks are retried up to 8 times over ~24 hours before being marked as permanently failed. You'll receive a webhook.failed event at that point.
This change improves delivery reliability during your endpoint's brief downtime windows without hammering a recovering service.