The Revenue AI Report

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.

WORKFLOW1Pick the promptChatGPT2Pull the inputsLinkedIn3Generate & editChatGPT4Track & iterate weeklyGoogle Sheets
4 steps, in order, with the tool that owns each one.
Adoption ladderSix levels from Starter to Rebuilt. This item sits at level 1.L1 StarterOne tool, no workflow changeL2 AssistedAI drafts, humans approveL3 IntegratedWired into CRM and SlackL4 OrchestratedMulti-step, owned, measuredL5 AutonomousAgent runs, human auditsL6 RebuiltThe process itself changes
This playbook belongs at L1 Starter. Running it above your level is how pilots stall.
Measures of successreply rate; emails sent per rep per day; time per emailPROVE IT WORKEDreply rateemails sent per rep per daytime per email

The steps

  1. 01

    Pick the prompt

    Tool: ChatGPT

    Use a single shared prompt stored in Notion or a pinned Slack message. Template: "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." Owner: SDR manager (15 min to write, 10 min to socialize). Pitfall: don't let every rep invent their own prompt, drift kills measurement. DoD: prompt lives in one URL the whole team has bookmarked.

  2. 02

    Pull the inputs

    Tool: LinkedIn

    Rep opens the prospect's LinkedIn profile and copies: name, title, company, last 1,2 posts, and any obvious trigger (funding, hiring, new role). Owner: each SDR. Time: 2,3 minutes per prospect. Pitfall: skipping the trigger field, it's what makes the AI output non-generic. DoD: every send has at least one specific trigger referenced in the body.

  3. 03

    Generate & edit

    Tool: ChatGPT

    Paste the inputs into the shared prompt. ChatGPT returns 3 variants. Rep picks one, edits 1,2 lines to add a human detail (mutual connection, mispronounced last name joke, anything real). Owner: SDR. Time: 90 seconds per email. Pitfall: sending raw output verbatim, replies will mention "AI slop" and burn the domain. DoD: every email is edited by a human before sending; no exact ChatGPT phrases left in production sends.

  4. 04

    Track & iterate weekly

    Tool: Google Sheets

    One shared sheet: date, rep, prompt version, sent count, replies, positive replies. Manager reviews every Friday in a 15-minute standup. If a rep beats the team reply rate by 30%+, their tweak becomes the new shared prompt. Owner: SDR manager. DoD: prompt is versioned (v1, v2, v3) and the team can point to which version is current and why.

Tools in this playbook

Next playbooks

Unfamiliar terms are defined in the AI and Revenue Dictionary. Related frameworks live in the framework library.

Share this playbook

Posting to Instagram or TikTok? Copy the link, it carries the title, summary and share image.

Arrives weekly by email. Free. Unsubscribe anytime. By subscribing you agree to our Privacy policy and Terms. We never sell or share the list.