ipIterPrompt

Cold Email That Gets Replies

Write a 90-word cold email built around the prospect's problem, not your product.

iterpromptUpdated 2026-06-105,240 copies

A cold outreach prompt with hard constraints that eliminate the tells of AI-written sales email: under 90 words, one specific observation about the prospect, one clear ask, no buzzwords, no 'I hope this finds you well'. Includes a follow-up sequence structure.

The prompt

Variables to fill in: {{product}}{{research}}

Write a cold email using these hard constraints:

- Under 90 words total.
- Open with a specific, verifiable observation about the prospect or their company (from the research notes below) — never with my company or "I hope this email finds you well."
- One sentence connecting their likely problem to what we do. No feature lists.
- One low-friction ask (a question they can answer in one line — NOT "do you have 30 minutes").
- No buzzwords: avoid "synergy", "streamline", "revolutionize", "reach out", "touch base", "solution".
- Write at an 8th-grade reading level. Short sentences.
- Subject line: 4 words max, lowercase, no clickbait.

Then write two follow-ups (day 4 and day 9), each under 50 words, each adding one NEW piece of value — never "just bumping this."

My product & the problem it solves: {{product}}
Prospect research notes: {{research}}

How to use

  1. 1Do real research first — the prompt is only as good as {{research}}. One specific fact (a job post, a launch, a tech-stack detail) beats a paragraph of generic firmographics.
  2. 2Generate 3 variants and pick the one that sounds most like you'd actually say it out loud.
  3. 3Keep the ask a question, not a meeting request — reply rate roughly doubles.
  4. 4Personalize the first line manually if the research notes were thin.

Examples

SaaS outreach to a hiring engineering leader

Input

Product: flaky-test detection for CI. Research: prospect is VP Eng at Series B fintech, posted on LinkedIn about slow CI, hiring 10 engineers this quarter.

Output

subject: your ci post Saw your post about 40-minute CI runs. With 10 engineers joining this quarter, that queue math gets ugly fast. We find and quarantine the flaky tests causing most re-runs — teams typically cut CI time ~30%. Worth asking: how many of your re-runs are flaky-test retries vs. real failures?

Pro tips

  • Read the output aloud. If any sentence sounds like a brochure, regenerate with 'more casual, like a Slack message to a colleague.'

Frequently asked questions

Why under 90 words?+

Reply-rate data across sales tools consistently favors emails readable in under 15 seconds on a phone. Length limits also force the model to cut the filler that pattern-matches to spam.

Can I use this for link-building or partnership outreach?+

Yes — swap {{product}} for what you're offering (a guest post, a partnership) and keep everything else. The constraints are the value.

Related