The Cold Email Problem

Everyone knows generic cold emails don't work. Everyone also knows personalized emails take 15 minutes per prospect. At scale, the math doesn't work either way.

Unless you change the equation.

The Setup

A BD lead at a mid-market SaaS company was sending 50 personalized emails per week. Good quality, decent response rates (12%). But 50 emails isn't enough to move the pipeline needle.

They wanted 200 per week without dropping quality. Here's what they built.

Step 1: Automated Research

Instead of manually reading each prospect's LinkedIn and company page, they built an AI research pipeline:

1. Feed a list of company names and contact titles
2. AI pulls public info — recent news, job postings, funding rounds, tech stack
3. Output: a 3-bullet "context card" per prospect

Time per prospect: from 10 minutes to 15 seconds.

Step 2: Personalized First Lines

The worst AI emails start with "I noticed your company is doing great things in [industry]." The best ones reference something specific and relevant.

The key insight: AI writes better personalization when you give it the research first. The context card from Step 1 becomes the input for Step 2.

Prompt pattern:

"Using this context about [prospect], write a first line that references a specific challenge they likely face given their recent [hiring/funding/product launch]. Keep it under 20 words. Don't be sycophantic."

Step 3: Template + Variables

The email body stays mostly consistent. Only three things change:

  • First line (personalized from Step 2)

  • Value prop angle (matched to prospect's industry)

  • CTA (varied across 3 options to avoid pattern detection)

Step 4: Smart Follow-up Sequences

The follow-ups aren't just "bumping this to the top of your inbox." Each follow-up adds new value:

  • Follow-up 1: Share a relevant case study

  • Follow-up 2: Ask a genuine question about their workflow

  • Follow-up 3: Short and direct — "Worth a 15-min call?"

The Results

  • Volume: 50 → 200 emails/week
  • Response rate: 12% → 14% (actually went up)
  • Time spent: 25 hours/week → 6 hours/week
  • Pipeline impact: 3.8x more qualified conversations

The Non-Obvious Lesson

The BD lead didn't automate everything. They still personally review every email before it sends. The AI does the research and drafting. The human does the judgment call.

That's the pattern. AI handles volume. Humans handle taste. Together, you get scale without sacrificing quality.