Cold email remains one of the highest ROI channels in sales and business development — when done right. The problem: most cold emails are ignored because they're generic, self-focused, or too long. ChatGPT, given the right prompt, can write cold emails that feel personal, lead with value, and end with a clear, low-friction ask. This collection gives you 50 prompts that produce emails people actually respond to. How to use these prompts: paste the prompt into ChatGPT (GPT-4o or GPT-4.5 recommended for best results). Fill in the [BRACKETS] with specific information about your prospect and offer. The output is a draft — always personalise the final version with one specific detail about the prospect (a recent post, funding round, product launch, or press mention) before sending. The anatomy of a cold email that gets replies: subject line (curiosity gap or specific benefit, under 6 words), opening line (about THEM, not you), one sentence of credibility, one sentence of the value proposition, one specific and easy ask (15-minute call, a yes/no question), one-line sign-off. ChatGPT follows this structure perfectly when prompted correctly. Prompt 1 — SaaS sales outreach: Write a cold email to a [job title] at [company name], a [industry] company with [X employees]. We offer [your product], which helps [specific problem your product solves]. Their likely pain point is [specific pain point]. The email should be under 150 words, open with a specific observation about their company (mention [recent news/achievement]), position our value as saving them [time/money/headcount], and end with asking if they have 15 minutes next week. Prompt 2 — Agency partnership outreach: Write a cold email to a [marketing director / CMO] at [company type]. I run a [your agency type] agency that [specific result you deliver, with a number]. I want to propose a referral partnership. The email should be friendly not formal, under 100 words, mention one specific type of client we could refer to each other, and end with asking if they'd be open to a brief intro call. Prompt 3 — Job application cold email: Write a cold email to the hiring manager (name unknown) at [company name] for a [job title] role. My background: [2-sentence summary of experience]. I'm not applying through the job board — I'm reaching out directly because [one specific reason this company interests you]. The email should be 4 sentences maximum, demonstrate knowledge of the company's recent [product launch/initiative], and end with a specific ask (portfolio link + 10-minute call). Prompt 4 — Podcast guest pitch: Write a pitch email to [podcast host name], host of [podcast name], a podcast about [topic] with [approximate listeners]. I want to pitch myself as a guest. My background: [your expertise in 1 sentence]. My suggested topic: [specific episode topic that serves their audience]. The email should reference a specific episode of theirs I can cite, be under 120 words, and not be over-eager. Prompt 5 — Investor outreach: Write a cold email to [investor name], a partner at [VC fund] who invests in [stage/sector]. We are [company name], a [one-sentence description]. We're raising a [amount] [seed/Series A] round. Our traction: [specific metric]. The email should be factual not salesy, under 100 words, mention a portfolio company of theirs that's relevant to our space, and end with asking to send a brief deck. Subject line formulas that consistently get opens: '[Their company name] + [specific result]' e.g., 'Acme Corp's churn reduction'; '[Mutual connection] suggested I reach out'; 'Question about [specific thing they published/built]'; '[Number] [specific result] for [similar company]'; 'Congrats on [recent achievement]'. ChatGPT can generate 10 subject line variations from your email — prompt: 'Give me 10 subject line options for this email, each under 6 words, using different psychological triggers'. Personalisation multipliers: ChatGPT can personalise at scale when given data. Feed it a CSV with columns: name, company, recent_news, job_title, and prompt: 'Using this data about [name], personalise this email template. Use recent_news in the opening line'. This works for 50-500 prospects if you have a VA to compile the data. The personalisation dramatically improves reply rates vs fully templated blasts. Follow-up email prompts: the first email isn't enough. A 3-email sequence (Day 1, Day 4, Day 9) doubles reply rates. Prompt: 'Write a 3-email follow-up sequence for this cold email. Email 2 should add new value (share a relevant resource). Email 3 should be a short breakup email that uses a soft close. Each email under 80 words.' What makes ChatGPT cold emails fail: prompting ChatGPT without specifying word limits produces wordy emails nobody reads. Not personalising the output (using the raw ChatGPT draft without editing) produces generic-sounding copy. Using ChatGPT for the subject line without testing alternatives. Always A/B test subject lines and opening lines using a tool like Lemlist or Reply.io. More ChatGPT prompts for sales and business at https://www.promptspace.in. Related: ChatGPT Content Writing Prompts at https://www.promptspace.in/blog/chatgpt-content-writing-prompts and ChatGPT Resume Prompts at https://www.promptspace.in/blog/chatgpt-resume-prompts-guide.