Use AI to write a better cold email
A practical workflow for using AI to draft cold email angles, personalize without sounding fake, and review whether the email has a clear reason to exist.
Who it is for
- Founders doing early outreach.
- Marketers testing messaging angles.
- Sales teams drafting respectful first emails.
Who should skip it
- People planning mass spam.
- Teams without a clear offer.
- Anyone pretending to know the recipient personally.
Workflow
Step 1
Clarify the reason to write
Before drafting, define the recipient, the trigger, the offer, and the one action you want. AI cannot rescue an email with no reason to exist.
Example input
Recipient: seed-stage SaaS founder. Trigger: hiring support team. Offer: reduce repetitive replies.
Expected output
A focused email brief.
Common failure
The model writes generic flattery.
Human check
Ask whether the recipient can tell why they specifically received the email.
Step 2
Generate angles, not final copy
Ask for several strategic angles before drafting. This helps you choose the reason for the email instead of polishing the first generic version.
Example input
Give me five respectful cold email angles for this recipient and offer.
Expected output
A list of angle options with tradeoffs.
Common failure
The first angle sounds plausible but is not relevant.
Human check
Reject any angle that could be sent unchanged to 1,000 people.
Step 3
Draft short and specific
Have the AI draft a short email with a clear opening, one reason, one benefit, proof if available, and a low-friction call to action.
Example input
Write under 120 words. No hype. One CTA.
Expected output
A concise email draft.
Common failure
The draft becomes too formal, too long, or too salesy.
Human check
Read it aloud and remove every sentence that does not earn its space.
Step 4
Personalize without pretending
Use only public, relevant, and true details. The goal is context, not creepy imitation or false intimacy.
Example input
Personalize using only this public note: they are hiring support engineers.
Expected output
A relevant first line or reason.
Common failure
The email implies a relationship or knowledge you do not have.
Human check
Ask whether the recipient would feel misled if they asked how you knew the detail.
Step 5
Review deliverability and ethics
Before sending, check subject line, length, claims, unsubscribe expectations, and whether the email respects the recipient's time.
Example input
Review this cold email for spam signals, unsupported claims, and unclear CTA.
Expected output
A send/no-send checklist.
Common failure
The AI optimizes persuasion but ignores trust.
Human check
If you would not send it to a respected peer, rewrite it.
Human review checklist
- Check whether the AI output directly solves the original cold email drafting instead of drifting into a generic answer.
- Verify all factual claims, dates, names, numbers, links, and quoted material against the original source or a trusted reference.
- Remove unsupported claims, filler language, repetitive transitions, and confident statements that do not have evidence.
- Compare the output with the intended reader, channel, and format before using it in public or sending it to another person.
- Keep a short note of the prompt, tool, input material, manual edits, and final decision so the workflow can be repeated.
Mistakes to avoid
- Starting the cold email drafting workflow with a vague prompt and no acceptance criteria.
- Asking the model for a final answer before giving it source material, constraints, examples, or review rules.
- Treating a fluent answer as correct without checking source coverage, missing assumptions, and edge cases.
- Using the same prompt for research, writing, review, and final editing even though those are different jobs.
- Skipping the human review step because the first output looks polished.
Related prompts
Turn bullet points into a concise email
Transform messy bullets into a clear email with a specific ask and next step.
Write five cold email angles
Create distinct outreach angles without pretending the sender has facts they do not have.
Related AI skills
Build a presentation outline with AI
A workflow for using AI to turn notes into a presentation structure with audience logic, slide purpose, speaker intent, and review criteria before any design work begins.
Choose the right AI tool for a task
A decision workflow for choosing an AI tool by task, input type, risk, review needs, deployment requirements, and failure cost instead of popularity or marketing claims.
Compare AI answers with source verification
A research workflow for comparing two or more AI answers without rewarding confidence, fluency, or longer output. The workflow uses source mapping, claim extraction, and contradiction checks.