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?
In my experience, refining a prompt can significantly improve the quality of the output. Adding more context, clarifying the objective, specifying the desired format, and providing constraints helps the AI generate more accurate and relevant responses. Small changes to a prompt can make the results much more specific and aligned with the original goal.
Generative AI can produce more relevant, accurate, and effective results when prompts are i continuously refined. By improving the prompt’s clarity, specificity, and context through an iterative process, users can better guide the AI toward the desired outcome and enhance the overall quality of its responses. Saving Changes...
Generative AI can produce more relevant, accurate, and effective results when prompts are continuously refined. By improving the prompt’s clarity, specificity, and context through an iterative process, users can better guide the AI toward the desired outcome and enhance the overall quality of its responses.
Generative AI can produce more relevant, accurate, and effective results when prompts are continuously refined. By improving the prompt’s clarity, specificity, and context through an iterative process, users can better guide the AI toward the desired outcome and enhance the overall quality of its responses.
In my experience refining a prompt can change the output entirely. Hence, if we are looking to have specific information / output from AI then provided it with an accurate prompt helps a lot rather than putting in things vaguely.
Saving Changes...
Ghadeh AlsaifClient Relationship ManagerRiyadh, Saudi Arabia
When I first started using GenAI, I would write very general, open-ended prompts, and the results were often generic and not quite what I needed. Over time, I noticed a significant improvement when I started assigning the AI a specific role or persona before making my request. This alone made the responses much more relevant and focused. The next big shift came when I started clearly specifying exactly what I wanted, rather than leaving it vague. Results improved even further once I began requesting a specific output format, like tables, bullet points, or structured sections. This made the responses far more usable and easier to apply directly to my work. The biggest change, though, came when I started breaking down complex tasks into smaller, sequential steps instead of asking for everything at once. This made the whole workflow much smoother and easier to manage, and the quality of the output improved dramatically as a result. Overall, my takeaway is that combining a clear role, precise instructions, a defined format, and task breakdown leads to drastically better results. Saving Changes...
There are framewoks to create prompt. This is part of the Prompt Desing discipline. Those that gave me and the initiatives where I was included are:R-T-F (Role-Task-Format), T-A-G (Task, action, goal), B-A-B (Before, after, bridge), C-A-R-E (context, action, result, example), R-I-S-E (role, input, steps, expectations).
For a specific bid request we were awarded, the use of prompt creation was required. The initial input was directed into a what was required from a specific vendor. From the output it was then more refined into the specif labor and material would be needed to carry out the job. With the output received and verified another prompt was requested to create a short summary understandable for the end user and to provide a chart with tasks, quantities for esase of understanding. When providing this output to our vendor, the turn around time for a quote was less than 24hrs, rather than them havving to extract what was needed of them. Chain-of-thought was very useful in this instance. Saving Changes...
Fady HegazyProjects Manager| ZAD GULF FOR CONTRACTINGSJeddah, Saudi Arabia, Egypt
In my experience, prompt refinement can completely change the usefulness of a GenAI output sometimes more than changing the AI tool itself. One clear example from my work is reviewing technical information. A general prompt such as: “Review this document and tell me the issues.” usually produces a broad and somewhat generic response. But when I refine the prompt by defining the role, objective, context, evaluation criteria, constraints, and expected output format, the result becomes much more valuable. For example, I may ask the AI to act as a project/technical manager, review the information against drawings, specifications, scope, constructability, coordination risks, and missing data, then present the findings in a structured table with: Issue | Impact | Required Clarification | Recommended Action | Priority The difference in output quality is significant. Another improvement I often make is asking the AI to identify assumptions and ask clarifying questions before giving the final answer. This reduces incorrect assumptions and helps keep the response aligned with the actual project objective. The refinements that have made the biggest difference for me are:
Giving the AI a clear professional role.
Explaining the real project context instead of only the task.
Defining exactly what “good output” should look like.
Providing constraints and reference criteria.
Asking it to separate facts, assumptions, risks, and recommendations.
Refining the result iteratively rather than relying on the first response.
I have found that a good prompt does not simply make the answer “longer” it makes it more relevant, structured, actionable, and easier to validate. For me, prompt engineering is less about finding a perfect sentence and more about translating professional intent into clear instructions that the AI can follow accurately. Saving Changes...
Anonymous
Refining the prompts helps to improve the quality of the result and achieve the initial objectives of the consultation.
PMI Team Member
