n8n Automation: When to Use It and When to Go Straight to Custom Code

Before tools like n8n, Make, and Zapier, automating a process meant hiring a developer for every little task. "When an email arrives in this inbox, copy the data to the CRM." That used to take two weeks of development. Now, it’s a 15-minute drag-and-drop workflow. The real question is: when does n8n (or similar tools) do the job, and when does it make sense to build custom code instead?

What is n8n and Why It Stands Out

n8n is a visual workflow automation platform: you drag and drop nodes (triggers, actions, conditionals) and connect them to build automation flows. Conceptually, it’s similar to Zapier and Make, but with two key advantages:

1. Open source and self-hostable. You can run it on your own server without per-execution fees. Platforms like Zapier charge per trigger or task, while n8n’s costs scale with your infrastructure needs.
2. More technical. n8n lets you embed JavaScript directly into nodes, handle complex transformations, and call custom APIs. This makes it far more powerful than Zapier for non-trivial use cases.

If your team has technical expertise to maintain it, n8n is often significantly cheaper and more flexible than its no-code alternatives.

Where Visual Automation Shines

Visual automation tools like n8n excel in these scenarios:

• Linear processes: Step 1, Step 2, Step 3. No complex branching or nested logic.
• Integrations with popular SaaS tools: Gmail, Slack, Notion, Google Sheets, HubSpot. Pre-built connectors make setup a breeze.
• Moderate volume: Up to a few thousand executions per month works perfectly.
• Frequently changing processes: Since flows are visual, non-developers can tweak them without waiting for a dev ticket.

A classic example: "When a lead submits a form, copy their data to HubSpot, notify the team in Slack, and add a row to a tracking spreadsheet." This takes 10 minutes to set up in n8n.

When n8n Becomes More Pain Than Gain

There’s a breaking point where n8n starts to struggle:

When logic grows complex. A 30-node workflow with 5 nested conditionals and parallel loops becomes unreadable in n8n. The same logic in code? ~80 clean, testable lines.

When performance matters. n8n adds latency between nodes. If your process needs to run in seconds (not minutes), raw code is far more efficient.

When you need automated testing. Manually testing a 30-node workflow after every change is error-prone. Code can be tested automatically with CI/CD pipelines.

When volume explodes. Beyond tens of thousands of monthly executions, the operational overhead of n8n (server maintenance, observability) starts to rival custom solutions.

When integrating with proprietary software. If your business logic lives in a custom app and automation is core to the user experience, embedding it in an external tool adds unnecessary complexity. Keep critical logic in your codebase.

Case Study: Mi Seguro de Auto – Custom Code Wins

Mi Seguro de Auto handles workflows like:

• When a quote is submitted: validate data, query multiple insurers in parallel, calculate commissions, persist results, and notify the user.
• If the quote is accepted: generate the policy, issue a receipt, email the client, log the transaction, and trigger a follow-up for payment.

This requires more than 15 steps with parallel execution, validations, error handling, and fallbacks. We built this directly in code (TypeScript + Next.js API routes) because:

• The logic is tightly coupled to the product, not peripheral.
• We needed automated testing for every step.
• Performance matters: users expect to see their quote in real time.
• When an insurer changes their API, we version control updates in Git.

Using n8n would have been possible, but far harder to maintain.

Where n8n Does Make Sense

n8n is ideal for internal, back-office processes that aren’t part of your core product:

• Daily reports generated at 8 AM and emailed to stakeholders.
• Syncing tools your team uses internally (CRM, calendar, spreadsheets).
• Operational alerts: "If metric X drops below Y, ping Slack."
• Client onboarding: When someone signs up, create their workspace, send a welcome email, and schedule a call.

For these use cases, n8n delivers speed, flexibility, and low costs.

The Golden Rule for Decision-Making

A simple heuristic: If the process is internal and operational, use n8n. If it’s part of your product and users interact with it directly, use code.

Another way to frame it: If the process is primarily for your operations team, n8n is a great fit (since they can adjust workflows themselves). If it’s for end users or customers, build it in code (for reliability, testing, and SLA guarantees).

Best Practices for Adopting n8n

If you’re moving forward with n8n, follow these tips to avoid future headaches:

1. Self-host from day one. It’s cheaper and gives you full control. Vercel isn’t suitable for n8n—opt for a VPS or Docker on a cloud provider.
2. Version your workflows. n8n lets you export workflows as JSON. Store them in Git like any other code.
3. Use environment variables for credentials. Never hardcode secrets in nodes.
4. Implement logging and alerts. If a critical workflow fails, someone needs to know before the client does.
5. Keep n8n separate from production databases. A hung n8n instance shouldn’t take down your app.

Final Thoughts

n8n and similar tools aren’t magic—they don’t replace code. They’re the right tool for a specific category of problems (linear, operational processes) and the wrong tool for another (critical product logic). Identifying which is which can save you time, money, and frustration.

If your business still relies on manual processes and you’re unsure which to automate with n8n or build in code, reach out to us. We’ll map out your workflows in a discovery session and build an automation roadmap with estimated ROI for each.

---

By Esteban Aleart, Founder & Lead Engineer at Pair Programming.