Back to Blog

10 Best ChatGPT Prompts for B2B Sales & Cold Emails (2026)

MB
Muhammad Bilal Khan — Founder
August 26, 2026SEO Optimized

I’ve sat on both sides — sending cold emails and receiving them. Most end up in the trash for the same reasons: all about the sender, too long, same template as everyone else.

These prompts are the ones I keep coming back to when I need a first draft fast. They’re starting points. Always add one detail only you would know.

1. Prospect research

Role: Sales researcher.
Context: Company is [Company]. My product helps with [problem].
Task: 5 specific observations I could mention in a cold email (news, hiring, product gaps). No generic praise.
Output: Bullets only.

2. First cold email

Role: B2B salesperson who writes like a peer.
Context: Product [X]. Target [title] at [company type]. Pain [Y]. Proof [one line].
Task: One email under 100 words. Lead with them. Soft CTA. No "I hope this finds you well."
Output: Subject + body.

3–4. Follow-ups

Follow-up #1 (3–4 days): Add one new angle. Under 60 words. No guilt.

Follow-up #2 (breakup): Polite close. Under 50 words. Door left open.

5–6. Objections

"Too expensive": Validate. Reframe cost of inaction or time saved. Soft next step. No defensiveness.

"We’re on [Competitor]": Don’t bash them. One smart question about a gap you solve. Under 80 words.

7–8. LinkedIn

Connection note: Under 300 characters. Specific reason. No pitch.

First message after accept: Helpful or insight-led. Soft ask only if natural. Under 80 words.

9–10. After they engage

Meeting confirm: 20-min call, 3-bullet agenda, clear and friendly.

Post-demo recap: 3 takeaways, agreed next step, one open question. Under 120 words.

Sequence I use

  1. Research → first email
  2. Follow-up → LinkedIn or breakup
  3. Objection reply if needed

Tip: Never send the first AI draft raw. Personalize one line. That’s usually enough to stop it feeling like spam.

Going deeper: what actually gets a reply in 2026

Generic outreach volume died a while ago. What still works is proof that you read something specific about the prospect before writing to them.

The signal I look for before writing

  • A recent LinkedIn post, not their job title — titles don't tell you what they care about this month
  • A trigger event (funding, a new hire, a product launch) that gives the email a real reason to exist right now
  • One thing I can reference specifically in the first line, so it's obvious this wasn't sent to 500 other people

The constraint that fixed my open rates

I cap every AI-drafted cold email at 60 words in the prompt itself. Left alone, the model over-explains, and over-explaining reads as automated the same way generic phrasing does — length is its own tell.


— Muhammad Bilal Khan, Founder at PromptHub. Built from outreach patterns that survived real inboxes, not theory decks.

\nB2B Sales and Emails

Read More Articles