In your experience with GenAI, how has refining a prompt drastically changed the output quality?

In my experience using GenAI tools, refining prompts has made a huge difference in the quality of output, especially in project management tasks like drafting status reports, risk registers, or communication plans.

I’ve learned not to expect perfect results on the first try. These tools work best when you treat the prompt like a brief you'd give a team member: it needs context, clear objectives, and sometimes even examples. The more I refine by adjusting the language, narrowing the focus, or adding background info, the more relevant and usable the output becomes.

For instance, I once needed a risk analysis draft for a cross-functional project. The first AI-generated version was too generic. But after refining the prompt with specific project details, team dependencies, and potential blockers, the output became much sharper, almost like a first draft I could actually work with.

In short, good prompting is a process. The better I guide the AI, the more value it brings to my day-to-day as a PM.

Refining prompts is something I’m learning for the first time through this course, and while I haven’t practiced it much yet, I’m already convinced of its value. I’ve seen how even small changes in how I prompt Microsoft Copilot across apps can influence the quality of results. I’m especially impressed by the idea of prompt chaining—it’s a powerful way to guide the AI step-by-step. Moving forward, I plan to apply these techniques to get more accurate and relevant outputs from GenAI.