Director, Learning Design & Development| PMIAsheville, NC, United States
With Generative AI, iteratively refining and optimizing prompts can lead to better AI-generated results. This may involve adjusting the specificity or clarity of the prompt to increase relevance and accuracy of results.
What examples do you have of how improving a prompt drastically changed the output quality? What specific changes did you make that led to the improvement?
Carolyn WongPM Consultant| Project Controls TechnologyHouston, Tx, United States
Thank you. I find the chaining of the prompts to be useful. In the past using AI, I've already experienced asking too many questions on the initial prompt and having to refine the results and correcting assumptions. Chaining of prompts gives a clear way to get better quality results.
In my experience, refining prompts significantly changes the quality of the result—so much so that without that refinement, the response would make no sense. It is undoubtedly very important.
In my experience, refining a prompt can dramatically improve output quality by adding context, constraints, and a clear objective. For example, instead of asking, "Create a project status report," I might specify the audience, project phase, key risks, desired format, and level of detail. The initial response is often generic, while the refined prompt produces a tailored report that is more actionable and relevant. I've found that iterative prompting—reviewing the output and then adding clarifications—often leads to significantly better results, saving time while improving accuracy and usefulness. This demonstrates that effective prompt engineering is less about asking a single question and more about guiding the AI toward the desired outcome through structured refinement.
In my experience, refining a prompt can dramatically improve output quality by adding context, constraints, and a clear objective. For example, instead of asking, "Create a project status report," I might specify the audience, project phase, key risks, desired format, and level of detail. The initial response is often generic, while the refined prompt produces a tailored report that is more actionable and relevant. I've found that iterative prompting—reviewing the output and then adding clarifications—often leads to significantly better results, saving time while improving accuracy and usefulness. This demonstrates that effective prompt engineering is less about asking a single question and more about guiding the AI toward the desired outcome through structured refinement.
In my experience, refining a prompt can dramatically improve output quality by adding context, constraints, and a clear objective. For example, instead of asking, "Create a project status report," I might specify the audience, project phase, key risks, desired format, and level of detail. The initial response is often generic, while the refined prompt produces a tailored report that is more actionable and relevant. I've found that iterative prompting—reviewing the output and then adding clarifications—often leads to significantly better results, saving time while improving accuracy and usefulness. This demonstrates that effective prompt engineering is less about asking a single question and more about guiding the AI toward the desired outcome through structured refinement. Saving Changes...
In my working experience with GenAi, refining prompts enhance the accuracy and the relevance of the output, and provided me with a significantly valuable output.
The difference between a vague prompt and a precise one is dramatic. When someone asks "help me with my project," the output is generic and often useless. When someone says "I'm a Senior PM at a telecom company deploying a CPaaS platform, write a concise weekly status report for a steering committee highlighting three risks and one key milestone" ; the output is immediately structured, relevant, and actionable. Same tool, completely different result. Refinement compounds that effect. Adding constraints like tone, format, audience, and length transforms a decent response into something genuinely useful. The prompt is essentially the quality control mechanism; the AI will always produce something, but whether that something is valuable depends almost entirely on how well the human frames the request Saving Changes...
PMI Team Member
