Building Cold Email System Sending 1 Million Emails/Day

General Overview

• Character or Personal Brand Name: Shog (BlackHatWorld member, anonymous, self-proclaimed technically experienced in email marketing).
• Implementation Timeline: Started January 2023, with updates through March 2023 (based on posts from 2/7/2023 to 3/13/2023).
• Implementation Country: Not specifically determined, but targeting major global email providers like Gmail, Yahoo, Microsoft, GMX, aiming to reach hundreds of millions of email users.

Revenue Model

Primary Income Method: Cold email marketing to promote offers, potentially affiliate marketing, eCommerce products, or running campaigns for clients.

How Revenue is Generated:

• Build system sending 1 million emails/day, targeting low click-through rates (0.1–0.3%), corresponding to 1,000–3,000 offer views/day.
• With 1% conversion rate, expect to achieve 10–30 orders/day (revenue depends on offer value, not specifically mentioned).
• Use large email lists (scraped or self-built), focus on specific niches to increase open and click rates.
• Low operating costs (~$50/month), but requires high time and technical investment.
• Notable Features: Model targets large scale, accepts low open rates (1–2%) and focuses on deliverability rather than inbox placement. Suitable for affiliate marketing campaigns or high-profit products, but high legal and technical risks.

AI Tools and Technology Used

ChatGPT (indirectly mentioned):

• Role: Create content for “white” campaigns (example: “5 home organization tips before Friday”).
• Notable Features: Free (basic version), helps create engaging non-promotional content to avoid spam flags.

Custom AI Content (future suggestion):

• Role: Personalize emails (example: use NLP to analyze recipient’s website and create relevant content).
• Notable Features: Not implemented due to slow APIs, but potential to increase CTR.

PowerMTA:

• Role: Main SMTP server for large-scale email sending, manages IPs and handles errors (backoff when encountering SMTP errors like 421 4.7.0).
• Notable Features: Allows complete control over sending system, avoids dependence on commercial SMTP services (like Amazon SES) which require identity verification.

Mailwizz:

• Role: Email campaign management, supports multiple sending domains, tracking, bounce handling, and spintax for random content generation.
• Notable Features: Supports warmup schedules, integrates with PowerMTA, automatically adds bounces to blacklist.

Postfix:

• Role: Receive emails (abuse@, postmaster@, return path) to handle spam complaints and automatic unsubscribes.
• Notable Features: Easy configuration, supports multiple domains, replaces services like Zoho to reduce costs and complexity.

DMARC Monitor:

• Role: Monitor DMARC reports from Google/Yahoo to check DKIM/SPF, ensure emails aren’t blocked or going to spam.
• Notable Features: Uses open-source script (dmarcts-report-viewer) to analyze DMARC reports.

Private Link Shortener:

• Role: Create private short URLs to avoid being flagged by public shortening services.
• Notable Features: Based on Django, customized for temporary redirects to avoid content checking bots.

VPS and IPs:

• Role: Provide multiple sending IPs (buy cheap VPS $2–5), use autossh and badvpn-tun2socks to create SMTP tunnels.
• Notable Features: Flexible to add/remove VPS, ensure system operates if one VPS is blocked.

Mail-tester.com:

• Role: Check email quality score to reduce risk of spam flagging.
• Notable Features: Free, provides detailed analysis of DKIM, SPF, content.

Implementation Process

Phase 1: Infrastructure Setup (January 2023):

• Action:
  • Configure PowerMTA, Mailwizz, Postfix on VPS (6 cheap VPS, 1 large VPS for Mailwizz, physical PC on-site).
  • Buy multiple IPs and domains, set up DKIM, SPF, DMARC, rDNS, and MX records.
  • Integrate DMARC monitor and private link shortener.
  • Start test sending with 25 emails/day/IP, gradually increase to 2,000/day/IP over 30 days.
• Details: Use educational content (non-promotional) to warm up IP/domain, ensure deliverability.
• Result: System operational, sending 3,000 emails/day after 2 weeks, 10–15% open rate.

Phase 2: Warmup and Testing (February 2023):

• Action:
  • Gradually increase sending volume (3,000–10,000 emails/day), use 6 IPs and 2 main domains, over 12 subsidiary domains sending trickle (5–10 emails/day).
  • Deploy “white” campaigns (educational newsletters), “grey” (links to reputable sites like CNN), and “black” (direct offers).
  • Monitor bounces, SMTP errors, and open rates, adjust campaigns based on data.
• Details:
  • PowerMTA automatically backs off 90 minutes if encountering errors like “421 4.7.0 Messages temporarily deferred”.
  • Open rates dropped to 1–2%, indicating emails going to spam folder.
• Result: Sending 10,000 emails/day, but encountered issues with Gmail (blocking SMTP for main domain).

Phase 3: Dealing with Blocks and Adjustments (Late February–March 2023):

• Action:
  • Identified Gmail as more sensitive to domain reputation, blocking SMTP when reputation “Bad” (according to Google Postmaster).
  • Tried rotating new domains, but got blocked after few hundred emails.
  • Paused sending to reassess strategy, considering rotating both domains and IPs, or simulating interaction with Gmail accounts.
• Details:
  • Proposed solutions: Automate domain/IP rotation, use commercial SMTP services (SES, Mailgun), or build legitimate lists.
  • Challenges: Configuration takes many hours, haven’t found way to rewrite domains in PowerMTA queue.
• Result: Project lost momentum, haven’t reached 1 million emails/day goal, but maintained low costs (~$50/month).

Results Achieved

Revenue: None yet (haven’t tested offers, only sent test content like “daily inspiration”).

Sending Volume:

• Reached 10,000 emails/day after 1 month, using 6 IPs and 2 main domains.
• Open rate 1–2% (mostly to spam), but stable deliverability (SMTP 250 OK).

Other Achievements:

• Built complex technical system (PowerMTA, Mailwizz, Postfix, DMARC monitor) at low cost.
• Identified Gmail as more sensitive to domain reputation compared to Yahoo/Microsoft, providing valuable data for optimization.
• Created foundation to test “white”, “grey”, “black” campaigns, with potential application for affiliate marketing or eCommerce.

Challenges:

• Gmail blocks SMTP for main domain after 100,000 emails, disrupting warmup.
• Low open rates (1–2%), showing difficulty achieving inbox placement without quality lists or real interaction.

Lessons Learned

Key Success Factors:

• Controlling infrastructure (PowerMTA, Mailwizz, Postfix) helps avoid dependence on commercial services and reduces risk of account blocking.
• Slow warmup (25–2,000 emails/day/IP over 30 days) ensures deliverability, avoids initial blacklisting.
• Using multiple IPs/domains and private link shortener reduces risk of spam flagging.

Creative Highlights:

• Categorized campaigns into “white” (educational newsletters), “grey” (reputable links), and “black” (direct offers) to build funnel and optimize interaction.
• Integrated DMARC monitor with open-source script to track DKIM/SPF performance, rarely used at this scale.
• Configured SMTP tunnels via autossh and badvpn-tun2socks, allowing one PowerMTA instance to send through hundreds of IPs.

Challenges and Solutions:

• Challenge: Gmail blocks SMTP due to low domain reputation. Solution: Rotate domains frequently, experiment with simulating Gmail account interaction, or build legitimate lists.
• Challenge: Slow warmup, loss of patience. Solution: Use Mailwizz to automate warmup, monitor data to adjust sending speed.
• Challenge: Legal risks (CAN-SPAM Act). Solution: Comply with 1-click unsubscribe, PO Box address, and don’t spoof headers.

Application Suggestions for Beginners:

• Start Small: Test with 1 VPS ($5/month), PowerMTA, and Mailwizz (~$99 license), send 100–500 emails/day to get familiar.
• Build System:
  • Configure DKIM, SPF, DMARC for 1–2 domains, use mail-tester.com to check.
  • Use “white” content (educational newsletters) from ChatGPT, send to small list (buy from Zoominfo or self-scrape).
  • Monitor bounces and open rates with Mailwizz, adjust campaigns if open rate below 15%.
• Marketing: Start with “white” campaigns to build reputation, then try “grey” with reputable links, only use “black” when having interactive lists.
• Initial Cost: ~$50 (VPS, domain, Mailwizz license), or ~$200 if buying quality lists.
• Learn Skills: Read PowerMTA/Mailwizz documentation, watch tutorials on BlackHatWorld or YouTube (search “cold email setup”).
• Note: Comply with CAN-SPAM Act (unsubscribe, real address), avoid lists containing spam traps, and accept risk of IP/domain blocking.

Source/Links:

• Original Link: Journey: Build a cold email system that can send 1M messages a day from scratch