Qautomata

Case Study: Automated Cold Email System Using Self-Hosted n8n

Objective: Automate the client’s outbound cold emailing process.

The Challenge

Our client had a lean but ambitious team focused on scaling B2B lead gen campaigns, they were previously using a suite of cold email SaaS platforms, which came with steep monthly fees, API limitations, and vendor lock-in.

They approached us with a clear mission:

🎯 “We want to cut costs and own the process, we’ve already set up a self-hosted n8n server, is it possible for us to run the entire cold outreach system ourselves?”

The requirements:

  • Pull lead data directly from a shared Google Sheet
  • Send cold emails and automatic follow-ups using their own SMTP (or optionally gmail)
  • Use AI to classify replies into buckets like Interested, Not Interested, Unsubscribe
  • Get notifications when someone replies positively.

All without paying per campaign or per email.

Our Solution: A Cost-Efficient Cold Email Engine Built on n8n

We architected a scalable, cost-effective solution using their self-hosted n8n environment, leveraging only free or usage-based tools that fit within their budget.

1. Lead Pulling from Google Sheets

Every interval, n8n pulls fresh leads from a Google Sheet, filters out leads that have already been contacted using a status column in the sheet, and get specific info from the leads for dynamic personalization in outbound emails.

Google Sheets served as the central hub, keeping everything lightweight and accessible.

2. Cold Emails + Automated Follow-Ups

We integrated n8n with the client’s own SMTP for sending cold emails generated using customizable and AI powered templates with placeholders for values like name, company, etc.

Follow-ups are scheduled at configurable intervals (e.g., Day 3, Day 7) unless a reply is detected.

All send activity and status updates are logged back in Google Sheets.

3. AI-Powered Reply Classification

Replies and events like email-opened, unsubscribed, bounce, delivered are captured via webhook, replies are passed to OpenAI LLM using a tailored GPT prompt.

AI classifies responses into categories: Interested, Not Interested, Needs More Info, etc. then classification results are logged to Sheets and used to guide next steps.

4. Instant Alerts for Hot Leads

When a reply is marked as Interested, n8n sends a Slack alert (with lead details and full reply), and follow-up sequences for that lead are halted immediately.

The Results

  • Cost savings of over $1K/year by eliminating SaaS fees
  • 100% self-owned system, no reliance on 3rd-party email tools
  • Follow-ups are sent in case leads didn’t reply
  • Faster response times thanks to notifications

Their entire cold outreach engine now runs 24/7 fully automated.

Final Thoughts

By replacing a patchwork of costly SaaS products with a streamlined, self-hosted automation setup, our client not only saved money, but also gained full control of their outbound engine, now, they scale campaigns without scaling costs.


Need help building your own lean, AI-powered outreach machine?
Let’s do it, Get in touch with us!