Just launched: 50 AI prompts for B2B sales reps (lessons from an AI engineer)

Background: I'm an AI engineer — I benchmark LLMs for a living.

I kept watching sales-rep friends use ChatGPT in ways guaranteed

to fail: pasting confidential prospect data into free chatbots,

trusting hallucinated company facts, sending emails that read

like AI parodies.

So I wrote a 47-page playbook with 50 tested prompts for finding,

researching, and emailing B2B prospects. Each prompt has:

– The exact text to copy-paste

– What good output looks like

– What to do when output is bad

– A real example

The angle most sales-prompt content misses: most of it is written

by salespeople who don't actually understand how LLMs fail.

I tried to write the version someone with technical depth

would write — covering hallucinations, model selection

(Claude vs ChatGPT vs Gemini for which task), and data

safety honestly.

A few honest things I'm still figuring out:

1. Distribution. I'm in Azerbaijan, so most Western platforms

(Etsy, Stripe-based tools, etc.) are limited for me.

Currently selling via Gumroad and looking at Amazon KDP next.

2. Pricing. Set at $29 — uncertain if that's right for an

ebook in this category. Open to feedback.

3. Audience. The product is for B2B sales reps specifically,

but my own network is mostly AI engineers. Still working

out how to reach the actual buyer audience.

If anyone's gone through this kind of "expert in adjacent

field tries to sell to specific buyer audience" problem

before, I'd love to hear what worked for you.

Happy to share what I learned about prompt engineering for

non-image LLM tasks if anyone's curious.