Ever go back-and-forth with ChatGPT, adding more and more clarifications… until the poor thing is confused and your output is worse than where you started? 🙃
There’s a smarter way: the Prompt Edit technique
Here’s how it works:
1️⃣ Write your first (basic) prompt. Example: “Write a short summary of our new product feature.”
2️⃣ Run it. Look at the output. Meh.
3️⃣ Instead of explaining in the thread → go back, click 📝 (pencil icon), Edit message, and refine the original prompt:
“Write a short summary of our new product feature for our Q4 launch deck. Audience is internal, non-technical stakeholders. Tone: clear and confident.”
4️⃣ Run it again. Notice how the output is sharper, cleaner, and more on target.
💡 When to use Prompt Edit:
✔️ Trying different prompt styles
✔️ Experimenting with tone
✔️ Refining an output for side-by-side comparisons
✔️ Testing the difference between garbage vs. CRISP prompts (👀 looking at you, GIGO 😅)
Credit: 100 School 🙌.