Cold outreach, newsletters and customer replies.
Email prompts help you produce email faster using ChatGPT, Claude, Gemini and more. Browse optimized prompts for writing workflows — copy, tweak the placeholders, and ship.
Editorial
This is the Email hub, part of our broader Writing collection. You'll find 13 email prompts written specifically for this workflow — each one structured to give the AI the role, audience, format and constraints it needs to produce output you can use immediately.
These prompts are built for writers, editors, marketers, founders and anyone who ships words for a living. They work across the major AI models — ChatGPT, Claude, Gemini and the leading image and video generators — and they hold up whether you're prompting in a chat window, an editor extension or a workflow tool. Reach for these when you need a strong first draft fast — blog posts, emails, newsletters, scripts or social posts.
Why these prompts work: each one starts with a clear role, names the audience or scene, defines the output format and adds 2–3 hard constraints. That structure is what separates a vague AI answer from a usable one. The library is curated so you don't have to test 30 versions of the same prompt to find the one that actually performs — that work is already done.
Expect a clean, on-tone first draft that needs light editing — not a starting point that needs rewriting. You'll occasionally want to regenerate or refine, especially the first time you use a prompt with your own context. Treat each prompt as a starting structure: the variables and constraints stay, the topic and tone become yours.
Read each prompt before you copy it. Swap the placeholders for your real audience, product and constraints, and add one or two writing-specific details from your own brief. The prompts here are designed to be edited — the more context you bring, the stronger the output.
If this is your first time using AI for email, start with the featured prompt below and adapt it to your context. Then explore the related use cases for adjacent workflows that often pair well with this one.
⭐ Editor's Choice
One prompt we'd ship today. Read why it works, see a preview, and copy it in a click.
Why it works · It pins the role, audience and format up front, so the output reads like a finished draft instead of a brainstorm.
Best use case
Best for email workflows where a short, opinionated prompt that gets a great first draft.
Expected output
A draft you can edit lightly and ship — not a starting point that needs rewriting.
A short, opinionated prompt that gets a great first draft.
The complete guide to writing emails with ChatGPT, Claude and Gemini. 30+ tested prompts for cold outreach, customer replies, newsletters and follow-ups — with templates, before/after rewrites and the mistakes that get emails ignored.
Beginner's guide
If you're brand new to using AI for email, work through these five fundamentals first. Skip them and every prompt below will sound generic.
Most AI-written emails fail for the same reason: the prompt skipped context. The model doesn't know who you are, who you're emailing, what you're selling, or what success looks like. So it writes the average of every email it has ever seen — which is exactly what makes those emails get ignored.
A great email prompt does five things on purpose. It assigns a role, names the recipient, sets the objective, defines the constraints, and gives the model a voice to imitate. Once you internalise this pattern, you can write a publishable email in a single attempt.
Voice-matching beats every other technique
Paste two short emails you've actually sent (with names redacted) and ask the model to match the cadence, sentence length and word choice. This single addition turns generic AI output into something that sounds like you wrote it.
Avoid these
The single biggest mistake
Sending the first draft. Every great AI email needs one human pass: cut the throat-clearing sentence, add one specific detail, and rewrite the CTA in your own voice.
Prompt templates
Five battle-tested prompt skeletons. Replace the bracketed fields and paste into ChatGPT, Claude or Gemini.
You are a [role] at [company], a [one-sentence product description].
You are emailing [recipient persona] at [recipient company type].
Write a cold email that:
- Opens with a specific, researched observation about [recipient company / industry trend]
- Connects that observation to a measurable pain point
- Proposes one concrete next step (a 15-minute call)
- Stays under 90 words, no buzzwords, no "I hope this finds you well"
- Ends with a P.S. line that adds a credibility signal
Voice reference — match the cadence and word choice of these two emails:
"""
[paste 2 of your real cold emails here]
"""You are a senior customer support lead at [company].
The customer is frustrated because [exact issue, in their words].
Account context: [tier, tenure, recent activity, anything relevant].
Write a reply that:
1. Acknowledges the specific impact on the customer (not a generic "sorry")
2. Owns the cause without blaming systems or other teams
3. States exactly what we are doing, by when, and how they will know
4. Offers one fair gesture appropriate to the situation
5. Closes with a direct line back to a human
Tone: warm, decisive, never defensive. Max 130 words.You are the editor of [newsletter name], a weekly newsletter for [audience].
This week's lead story: [topic + the contrarian angle].
Write the first 120 words of the issue:
- A hook that pays off the subject line "[subject line]"
- One specific scene, number, or quote in the first three sentences
- A bridge into the rest of the issue
- Voice: [paste a paragraph from a past issue]
Avoid: "In today's edition…", "Welcome back…", any throat-clearing.Before & after
Same goal, two prompts, very different results. Notice what the second one adds — that's where the lift comes from.
Before
Write a cold email to a marketing director offering our SEO tool.
After
You are a founder at Rankly, an SEO platform that helps in-house teams cut agency spend by 40%. You are emailing the VP of Marketing at a Series B SaaS company that recently posted a job for an SEO Lead. Write a 75-word cold email that references the open role, names the agency-spend pain point, and asks for a 15-minute call next week. No buzzwords. Match the voice of these two emails: [paste examples].
Why it works — The second prompt names a role, a real trigger event, a measurable pain point, a single CTA, a word limit and a voice reference. The model now has everything it needs to produce a publishable first draft.
Before
Write a follow-up email after a sales call.
After
You are an account executive at Notion. You just finished a discovery call with the Head of Product at a 200-person startup. Their main pain: their PRDs live in five tools and engineers ignore them. They mentioned a deadline of end of Q3 and three skeptical engineers. Write a 110-word follow-up that recaps the pain in their language, addresses the engineer skepticism with one concrete proof point, and proposes a 30-day pilot with a clear success metric.
Why it works — Specificity is the unlock. Names, numbers, exact objections — these are the inputs an AI model cannot invent on its own.
Advanced
Workflow
The exact loop high-volume operators use to produce 20+ on-brand emails per day.
| Step | Tool | Output |
|---|---|---|
| 1. Research | ChatGPT + web search | 5 bullet points on the recipient and their company |
| 2. Draft | Claude (long context) | First-pass email using a saved template |
| 3. Voice match | Claude | Rewrite in your voice using 3 reference emails |
| 4. Pressure-test | ChatGPT | Predict objections + tighten the CTA |
| 5. Human pass | You | Cut one sentence, add one specific detail, send |
Why two different models?
Claude tends to produce warmer, more idiomatic prose. ChatGPT is better at structured research and adversarial review. Using them in sequence avoids the 'one-voice' problem of pipelines built on a single model.
By industry
Lead with a trigger event — funding round, new hire, a post the recipient published. The trigger event is what makes the email feel one-to-one. Without it, even a perfectly worded prompt produces an email that reads like a campaign.
Specificity over warmth. A customer who reports a bug doesn't want apologies — they want a timestamp, a root cause and a fix. Tell the model to lead with the facts and let empathy come through structure.
Reference the candidate's actual work — a project, a talk, a repo. Give the model 2–3 links and ask it to extract a genuine reason this person is a fit. Generic "I came across your profile" emails have a sub-1% reply rate.
Scenarios
Increase the volume of voice samples (5+ short emails, not 1 long one) and add a negative example. Then run a second pass: "Rewrite — shorter sentences, more contractions, one personal aside."
Paste your one-pager directly into the prompt and say: "Use only facts present in the source above. If you need a stat that isn't there, leave a [TODO: stat] placeholder." This eliminates 95% of hallucinations.
Set a hard word limit and a hard sentence limit ("under 90 words, no sentence longer than 18 words"). Constraints produce better writing than instructions to "be concise."
Continue exploring
Hand-picked next steps based on what you just read.
Featured Prompts
Eight prompts to copy, tweak and ship right now.
A short, opinionated prompt that gets a great first draft.
A structured prompt with role, audience and output constraints.
Expanded prompt for in-depth, polished output.
Refines or rewrites an existing draft toward a target tone.
Generates a wide list of angles or variants to pick from.
Reviews work and returns prioritized, actionable feedback.
Builds a structured outline you can hand to a writer or another model.
Takes a rough draft and ships a final, ready-to-use version.
Model comparison
A ranked shortlist of the models PromptInFlow users reach for first.
Prompt Writing Guide
Six habits that consistently produce stronger AI output. Apply them to any prompt on this page.
Tell the model who it is — "You are a senior B2B content editor" — so it picks the right register, vocabulary and structure on the first try.
Name your audience, their job title and what they already know. The same prompt produces very different writing for a CFO versus a junior developer.
Ask for the exact deliverable: a 700-word blog draft with H2s, a 4-line cold email, a Twitter thread of 6 tweets. Vague requests produce vague drafts.
Use 2–3 tone descriptors plus an example sentence the AI should match. "Confident, warm, no jargon" beats "professional".
Cap sentence length, ban filler phrases ("in today's fast-paced world"), require at least one concrete number per paragraph. Constraints raise quality fast.
Reply with surgical edits — "Tighten paragraph 3, keep the rest" — instead of rewriting the prompt. You'll converge in two rounds.
Best Practices
Apply these in every prompt on this page. Small habits, outsized improvements in output quality.
Paste two sentences in your real voice. Style transfer beats adjectives every time.
Asking for sentences under 22 words quietly fixes pace and rhythm.
Numbers anchor abstract claims. This single rule sharpens most drafts.
"In today's fast-paced world", "navigating the landscape", "unlocking value" — name them and ban them.
Avoid These
The same handful of mistakes are responsible for most weak AI output. Catch them before you hit send.
No topic, no audience, no length. You'll get filler. State the angle, reader and word count.
If you don't ask for H2s, meta description and a CTA, you won't get them — and you'll do the structural work by hand.
Write the draft first, then edit in a second prompt. Mixing them produces a draft that's half written and half rewritten.
"Make it engaging" means nothing. Use two adjectives and a sample sentence.
One example of the output you want replaces a paragraph of instructions.
Pro Tips
Side-by-side rewrites that show what separates a weak prompt from a great one.
Instead of
“Write a blog post about productivity.”
Use
“Write a 900-word blog post for engineering managers titled "The 3 meetings every senior IC should decline." Use a confident, direct tone, three H2 sections, and end with a 2-sentence summary plus 5 takeaways.”
Why it works: Audience, length, structure, tone and deliverables are all pinned.
Instead of
“Make this email better.”
Use
“Rewrite this cold email to a VP of Engineering. Keep it under 90 words, open with a personalized line referencing their recent hiring post, and end with a 15-minute meeting CTA.”
Why it works: Edits become measurable when you state the constraints.
Instead of
“Summarise this article.”
Use
“Summarise this article for a busy CEO in 5 bullets, each under 18 words, and end with one decision the CEO should make based on the content.”
Why it works: Format, length and downstream use are all explicit.
Instead of
“Write a newsletter intro.”
Use
“Write a 120-word newsletter intro in the voice of a candid solo founder. Open with a surprising stat, link it to this week's main story (link below), and close with a single question to the reader.”
Why it works: Voice, structure and intent are locked.
Instead of
“Rewrite this paragraph.”
Use
“Tighten this paragraph by 30%, ban filler phrases, and keep the original meaning. Return only the edited version.”
Why it works: Targeted edits beat generic rewrites.
Instead of
“Brainstorm content ideas.”
Use
“Generate 10 article ideas for an audience of B2B marketing leaders at $5M–$50M ARR companies, each idea framed as a question the reader is already asking.”
Why it works: Audience-led ideation produces ideas worth writing.
Browse by Model
Jump into the model hub to see how each one handles this workflow.
Parent category
Email is one of 5 writing workflows on PromptInFlow. Open the full hub for the complete library, FAQs and supported AI models.
Visit Writing hubCollections
Curated groups for different skill levels and goals.
Easy starters with clear placeholders — copy, paste, done.
Multi-step prompts with role, constraints and structured output.
Conversion-focused angles built on proven frameworks.
Polished, client-ready outputs for day-to-day work.
FAQ
Common questions about these prompts.
From the blog
Guides, playbooks and case studies from the Prompt InFlow blog.
Related searches
Popular searches related to this page. Each one links to the closest matching prompt library.
Browse optimized prompts for modern AI models.
