Understanding Exponential Backoff for Webhook Retries

When a webhook delivery fails, you need a strategy to retry it. Exponential backoff is the industry-standard approach that balances persistence with server health.

What is Exponential Backoff?

Instead of retrying immediately or at fixed intervals, exponential backoff increases the delay between each retry attempt:

• 1st retry: 1 minute
• 2nd retry: 2 minutes
• 3rd retry: 4 minutes
• 4th retry: 8 minutes
• 5th retry: 16 minutes
• And so on...

Why Use Exponential Backoff?

Prevents Overwhelming Servers

If a server is down due to high load, hammering it with retries makes things worse. Exponential backoff gives the server time to recover.

Handles Temporary Issues

Many failures are temporary—network blips, brief maintenance windows, or rate limiting. Spreading out retries increases the chance of eventual success.

Respects Rate Limits

APIs often have rate limits. Exponential backoff naturally reduces request frequency, helping you stay within limits.

Adding Jitter

Pure exponential backoff can cause "thundering herd" problems—many retries happening at the same time. Adding random jitter prevents this:

const baseDelay = Math.pow(2, attemptNumber) * 1000;
const jitter = Math.random() * 1000;
const delay = baseDelay + jitter;

How HookWatch Handles Retries

HookWatch uses intelligent exponential backoff with:

• Up to 10 retry attempts over 24 hours
• Random jitter to prevent clustering
• Configurable retry schedules per endpoint
• Alerts when retries are exhausted

When to Stop Retrying

Eventually, you need to give up. HookWatch:

1. Marks the webhook as "failed" after all retries
2. Sends you an alert
3. Keeps the full payload so you can manually replay it

Never lose a webhook again—let HookWatch handle your retry logic.