Prompting ChatGPT Is Not a Revenue-Generating Activity

A founder once sent me a sales playbook he was genuinely proud of. He’d put a basic prompt into ChatGPT, gotten back several confident pages, and believed he’d found something close to genius. It wasn’t. Buried in the “research” was a claim sourced from what read like a random Reddit comment, presented with the same authority as everything else. The rest wasn’t much better — generic, internally inconsistent, built on assumptions about his market that were simply wrong. Reading it closely enough to explain why, without wasting an hour being unkind about it, was genuinely hard. And then came the part that actually mattered: he had no real intention of doing any of the actions in it anyway.

Think of it as a team member, not an oracle

If you handed a brand-new hire with zero knowledge of your market a two-line prompt and told them to write your sales strategy, you’d expect exactly what you got: their best guess, delivered with whatever confidence they happen to have, none of it validated against reality. That’s what happened here — except the “hire” writes fluently enough that the guess reads like expertise, and the founder had no way to tell the difference because he didn’t know the subject either. That’s the actual failure mode: not that the tool is bad, but that giving vague direction to something that can’t tell you what it doesn’t know produces confident nonsense at a speed no human intern ever could.

The output was never the point

Here’s what got missed entirely: even a mediocre, half-wrong plan is worth more executed than a “perfect” one left in a doc. If he’d taken the worst version of that playbook and actually called three prospects with it, he’d have learned more in a week than another round of prompting would ever tell him. Instead, the plan became the deliverable — something to feel good about having produced — rather than a starting point for the only activity that actually moves revenue: talking to a real customer or prospect.

What to do instead

Before you put another prompt into ChatGPT expecting strategy back, ask whether you already know enough about the subject to catch it when the answer is wrong. If you don’t, the fix isn’t a better prompt — it’s finding someone who does know, even for twenty minutes, before you act on anything it hands you. And whatever you do with the answer, don’t let producing it substitute for doing something with it. Call the customer. That’s the part AI can’t do for you, and it’s the only part that was ever going to generate revenue.

This is the exact gap between using AI and leveraging it responsibly — which is the whole premise behind the AI Revenue Leverage Intensive: judgment first, tools second.