All posts
L1 Maturityoutbound 4 min read

ChatGPT for First-Draft Cold Emails: The L1 Sales Playbook

Learn the L1 Cold Email playbook: A low-tech, high-impact AI workflow using ChatGPT and LinkedIn to triple SDR output without losing the human touch.

Run the playbook

ChatGPT for first-draft cold emails (L1)

The simplest possible AI lift: reps paste a LinkedIn profile + one pain point into ChatGPT and get a 3-variant cold email draft. No integrations, no automation — just a shared prompt and a 10-minute Loom showing how to use it.

Why this matters

The "blank page" problem is the single greatest killer of outbound productivity. When an SDR faces a blank compose window, they spend 15 minutes over-researching a prospect, checking their Twitter, and second-guessing every word. The result: maybe 15 high-quality emails a day. On the flip side, "spray and pray" automation has been effectively neutralized by spam filters and buyer fatigue.

The middle ground is L1 AI Implementation. By using a standardized, manual prompt-and-paste method, you solve for both speed and quality.

The cost of doing nothing: If your 10-person SDR team spends 12 minutes writing one personalized email, they cost you roughly $32 per email in fully-loaded labor. This manual workflow compresses that 12-minute block into 4 minutes. Across 10 reps, that’s 40 extra hours of selling time recovered per week without spending a dime on complex API integrations or RevOps professional services.

How it works: The 4-Step SDR Workflow

This is not a "fire and forget" bot. It is a human-in-the-loop workflow designed to leverage ChatGPT as a junior copywriter.

1. Pick (and Lockdown) Your Shared Prompt

Scale fails when 10 reps use 10 different prompts. You cannot measure what is not standardized. The SDR Manager must own the "Golden Prompt" and store it in a centralized location like Notion or a pinned Slack channel.

The prompt architecture:

"You are an SDR writing to {role} at {company size} {industry}. Their pain: {pain}. Recent trigger: {trigger}. Write 3 cold email variants under 80 words each — direct, curious, and value-led. No exclamation marks, no 'hope this finds you well', no AI-tells."

Why this works: It bans the classic "AI-isms" that trigger a prospect's mental spam filter. By demanding three variants, you give the rep choice, which keeps them engaged in the quality control process.

2. Pull the Human Inputs

The rep opens LinkedIn and spends 2 minutes identifying the "Spark." AI can't feel empathy; it needs raw data. The rep copies:

  • Standard firmographics (Title/Company)
  • The "Trigger": A recent funding round, a specific hiring surge, or a comment the prospect made on a post.

Tip: If the prospect hasn't posted recently, the 'trigger' should be the company's 10-K report or a recent podcast appearance by their CEO.

3. Generate & Edit (The 90-Second Rule)

The rep pastes the inputs into ChatGPT and gets the three variants. The crucial step is the "Human Polish." We recommend the 80/20 Rule: ChatGPT does 80% of the heavy lifting; the rep adds the final 20% of humanity.

This might be a reference to a mutual connection or a specific localized joke. If a rep sends raw AI output, they risk "AI Slop" accusations that burn your domain's sender reputation.

4. Track & Iterate Weekly

Treat your prompts like code. Use a shared Google Sheet to track:

  • Prompt Version (v1, v1.2, etc.)
  • Reply Rates per Rep
  • Positive Sentiment Rate

If an SDR finds a specific tweak—like adding a "P.S." line about a specific competitor—that boosts replies by 30%, that tweak is immediately integrated into the Global Prompt for the following week.

Tools you need

  • ChatGPT (Plus/Teams): For the heavy lifting.
  • LinkedIn Sales Navigator: For accurate trigger identification.
  • Notion: To house the "Prompt Library."
  • Salesloft/Outreach: To execute the actual send and track data.
  • Loom: For the manager to record a 10-minute "how-to" so there’s no ambiguity on the workflow.

KPIs to track

  • Reply Rate: Target a 2-3x increase over "generic" templates.
  • Emails Sent per Rep/Day: Expect this to jump from ~20 to ~50-60 high-quality personalized sends.
  • Time per Email: Benchmark this at <4 minutes from research to send.

Common pitfalls

  • Prompt Drift: Reps think they are "prompt engineers" and change the secret sauce. Lock the prompt.
  • The "AI Tell": If an email starts with "In the ever-changing landscape of...", delete it.
  • Skipping the Trigger: A personalized email without a trigger is just a well-written cold call. It doesn't work.

When to graduate to the next level (L2)

You are ready for the next level when your manual workflow is hitting its ceiling. When your reps are consistently spending 2 hours a day just copy-pasting into ChatGPT, it’s time to move to L2: Automated Enrichment. This involves using tools like Clay or Lindy to automatically pull those LinkedIn triggers and feed them into the LLM via API, removing the manual "copy-paste" middle step entirely.

But don't rush. If you can't get replies with a manual prompt, automation will only help you fail faster. Master the L1 first. floor.

ChatGPTLinkedInpromptscold emailstarter

Ready to ship it? Open the playbook

ChatGPT for first-draft cold emails (L1)

Step-by-step instructions, the tools to use, and the KPIs to watch — already wired into the Revenue AI Strategy workspace.

Open playbook