Why Automations Fail (And How Often)
API rate limits, expired tokens, schema changes, timeout errors, and data format mismatches are the top causes of automation failures. Even well-built automations fail 2–5% of the time on average. At 1,000 tasks/month, that is 20–50 failed executions you might never see without proper monitoring.
Built-In Monitoring by Platform
| Feature | Zapier | Make | n8n |
|---|---|---|---|
| Error notifications | Email alerts on all plans (Free+) | Email alerts on all plans (Free+) | Email alerts (cloud Starter $20/mo+). Self-hosted: configure manually. |
| Error logs / history | Task history: Free 7 days, Pro $29.99/mo+ 30+ days | Execution log: Free 5 days, Core $10.59/mo 30 days, Pro $18.82/mo 60 days | Full execution log. Self-hosted: unlimited retention. Cloud: varies by plan. |
| Auto-retry | Automatic retries on errors (all plans) | Configurable retry per scenario (all plans) | Retry node available (all plans). Configurable attempts and delays. |
| Error routing | Paths with error conditions (Pro $29.99/mo+) | Error handler routes (all plans) | Error workflow node (all plans) |
| Webhook/Slack alerts | Build a monitoring Zap (uses tasks) | Error handler → Slack/webhook (built-in) | Error workflow → Slack/webhook (built-in) |
Zapier: Error Monitoring Setup
Step 1: Enable email notifications. Go to Settings → Notifications. Ensure "Zap errors" is enabled. Zapier sends email alerts when a Zap fails. Available on all plans including Free.
Step 2: Check Task History regularly. Go to Zap History to see all executions. Free plans retain 7 days. Professional ($29.99/mo) and above retain 30+ days.
Step 3: Build a monitoring Zap (advanced). Create a Zap that triggers on errors and sends a Slack message or webhook. Note: this uses tasks from your plan quota. On Professional (750 tasks/mo), a monitoring Zap that fires frequently can eat into your allowance.
Limitation: Zapier does not have dedicated error routing on Free. Paths (conditional logic including error handling) require Professional at $29.99/mo. On Free (100 tasks/mo), you get email alerts only.
Make: Error Monitoring Setup
Step 1: Enable email notifications. Each scenario has notification settings. Enable error alerts per scenario. Available on all plans including Free.
Step 2: Add Error Handler routes. Make's error handling is a standout feature. You can add an error handler to any module that catches failures and routes them to a Slack notification, email, or data store. This works on all plans, including Free (1,000 ops/mo, 2 active scenarios).
Step 3: Use the Incomplete Executions feature. Make stores failed executions and lets you replay them after fixing the issue. This is far more powerful than Zapier's replay, which only retries the same data.
Execution log retention: Free: 5 days. Core ($10.59/mo): 30 days. Pro ($18.82/mo): 60 days. Teams ($34.12/mo): 60 days.
n8n: Error Monitoring Setup
Step 1: Set up Error Workflows. n8n has a dedicated Error Workflow feature. When any workflow fails, it triggers a separate error workflow. Use this to send Slack messages, emails, or POST to a monitoring webhook. Available on all plans including Community (self-hosted, free).
Step 2: Use the Retry node. n8n's Error Trigger node captures errors with full execution data. The Wait node can add delays between retries. You have complete control over retry logic.
Step 3: Self-hosted monitoring. If you self-host n8n (Community edition, free), pair it with Uptime Kuma or similar for infrastructure-level monitoring. n8n cloud (Starter $20/mo, Pro $50/mo) handles infrastructure monitoring for you.
Advantage: n8n gives you the most control. Because you can write JavaScript/Python in any node, you can build sophisticated error handling that checks error types, escalates based on severity, and integrates with any monitoring tool via HTTP.
When to Add External Monitoring
Built-in monitoring is enough for most teams. Consider external monitoring when:
- You have 20+ active automations across platforms
- You need a centralized dashboard for all automation health
- You need to track automation SLAs (e.g., lead must be processed within 5 minutes)
- You self-host n8n and need infrastructure monitoring
Free options: Uptime Kuma (self-hosted, free), Healthchecks.io (free tier: 20 checks). Paid options: Better Stack (from $29/mo), Datadog (enterprise).