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 radically changes the quality of the results. A vague question often yields a generic answer. But as soon as you clarify the context, objectives, and constraints, the answers become: more relevant, more precise, and often immediately actionable. Saving Changes...
Fernando GarciaPMO Manager| TELUS Communications Inc.Coquitlam, British Columbia, Canada
Well, this is the typical 'garbage in, garbage out'. Something that people sometimes forget is that GenAI is not actual 'reasoning', or even 'thinking'; it is just a statistical approach to a result based on a whole lot of (generic) data, so taking the time to put some effort into the proper prompt has made a huge difference in my experience Saving Changes...
Jack Febrian RusdiResearcher and Lecturer| Universitas Teknologi BandungBandung, Indonesia
Prompt engineering is crucial for project managers as it determines AI output quality. Specific prompts (e.g., "Create a construction project risk register with 5 key risks, mitigation plans, and responsible persons") yield more actionable results than generic ones. By including project context, stakeholders, and desired format, AI can deliver relevant, ready-to-use recommendations. The keys are: details, clear frameworks, and iterative prompt refinement based on project needs. Saving Changes...
Prompt engineering has really helped me get relevant and optimized results. It has helped me more efficient. I like the prompt engineering template and find that documentation is really helpful. Saving Changes...
One key lesson I’ve learned—refining my prompt can dramatically improve the output quality.
Initially, I asked, “How to improve team performance?” and received a generic response. But when I rephrased it to, “How to improve performance in a remote Scrum team using Zoho Projects for a cloud migration project?”, the AI delivered specific, actionable insights—like sprint planning tweaks, Zoho dashboard customization, and communication cadence best practices tailored for distributed teams.
The more context I give, the more relevant and actionable the results. Just like with any stakeholder. Saving Changes...
I’ve seen firsthand how refining a GenAI prompt can drastically elevate the quality of results—especially in project environments where clarity, tone, or stakeholder alignment matters.
For example, I once asked ChatGPT to “create a stakeholder communication plan” and got a generic outline. But when I refined it to:
“Build a communication matrix for a utility infrastructure project with regulators, internal SMEs, and third-party vendors, highlighting risks of scope creep and change fatigue,”
—the result was surprisingly aligned with what I’d expect from a seasoned consultant.
The biggest takeaway? Prompting GenAI is less about asking questions and more about setting context like a PM.
The more you apply stakeholder language, risks, timeframes, or role-specific intent into the prompt—the more GenAI feels like a capable team member.
Curious how others are experimenting with prompt refinement in real project settings—if anyone’s still around.
In my experience with GenAI, refining a prompt has made a huge difference in the quality of the output. I transitioned from simply asking a question or giving a one word direction like "evaluate this" (duh!) to structuring the request to provide context and other requirements - "who you are, what your problem is, provide solution looking like a table with columns, rank the rows by level of complexity" etc, the quality and usefulness of output was VERY significant and the number of iterations I had to go through to receive what I am looking for reduced dramatically. Saving Changes...
David WoodPortfolio Manager| AbbVieLadera Ranch, CA, United States
Always keep refining your prompts. Document what works, and you will be much more efficient. Saving Changes...
Godswill EgegwuAgile Coach, Scrum Master®, Author & TechpreneurAix-en-Provence, France
In my experience working with different GenAI tools, I've noticed that well-structured prompts, especially when broken down into clear steps and supported with resources or examples can improve the quality of responses by up to 70%.
It’s a bit like how a good Work Breakdown Structure (WBS) as we already know, helps identify risks and improve project clarity, decision making and performance. Similarly, refined prompting leads to more accurate, relevant, resourceful, and quality AI-generated results.
- - -
If I’ve missed anything or if you’ve had different results, I’d love to learn from your experiences too!
Saving Changes...
Godswill EgegwuAgile Coach, Scrum Master®, Author & TechpreneurAix-en-Provence, France
In my experience working with different GenAI tools, I've noticed that well-structured prompts, especially when broken down into clear steps and supported with resources or examples can improve the quality of responses by up to 70%.
It’s a bit like how a good Work Breakdown Structure (WBS) helps identify risks and improve project clarity, decision making and performance. Similarly, refined and optimized prompting leads to more accurate, relevant/interest-based, and quality AI-generated responses/results.
If I’ve missed anything or if you’ve had different results, I’d love to learn from your experiences too!
PMI Team Member
