Most people think the problem with AI is learning how to prompt. It’s not. It’s learning how to think.
See the comparison in the image. Both prompts use the same solid structure (in this case, the GRACE framework: Goal, Role, Action, Context, Example). Both have the same data.
But the thinking behind the prompt is fundamentally different.
➡︎ Approach 1 – “Help me reduce our churn rate.”
Result: Standard retention playbooks. Email sequences, loyalty programs, health scores. Useful, but expected.
➡︎ Approach 2 – “What if we’re thinking about churn all wrong? What if losing customers is revealing an opportunity?”
Result: A game-changing insight. The discovery that your “churned” customers are actually your most successful users outgrowing the product. Churn is reframed from a retention problem into a qualified lead signal for an enterprise upsell.
You get the most from AI when it can show what you’re missing about your own business.
This represents a fresh perspective that gives teams an edge.
Instead of focusing only on prompt engineering, the real advantage lies in better framing problems.
It doesn’t matter which prompt framework you use. What matters is the depth of the thinking before you even start writing.
I go deeper on how to develop this strategic thinking in my latest newsletter (link in comments: https://lnkd.in/d7jkBF7h).
What’s one core assumption your team isn’t questioning right now?
For more examples and insights on this topic: https://www.linkedin.com/pulse/why-smart-teams-still-get-mediocre-ai-results-liza-adams-adwgc