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?
Grant DeCeccoPrincipal| FisherPeak ManagementNorth Vancouver, British Columbia, Canada
Aug 08, 2025 3:11 AM
Replying to James Cody
...
Project managers are all too familiar with 'quality in' relating closely to 'quality out'. This is particularly true with GenAI. Without background, persona, target audience and task, among other inputs, the output is seldom of much use.
agreed. I think knowing what prompt framework to use for what situation is helpful as well. Saving Changes...
Grant DeCeccoPrincipal| FisherPeak ManagementNorth Vancouver, British Columbia, Canada
Aug 06, 2025 3:02 PM
Replying to Jaymesh Shah
...
The most dramatic improvements came from teaching AI to think like my stakeholders rather than like a generic consultant. When I started including stakeholder psychology, organizational culture, and specific business pressures in my prompts, AI outputs became not just accurate—they became strategically valuable.
The 10x improvement factor: Moving from generic prompts to psychologically-aware, context-rich strategic prompts didn't just make AI better—it made AI outputs superior to what I could have created manually, because AI maintained perfect consistency with all the contextual requirements I gave it.
Yes, this is where using a "Project" or "Custom GPT" can be very useful. You can give specific information about the situation and stakeholders as context to be used with all future prompts. This saves a lot of time and gets your replies more tailored to your needs. Saving Changes...
Grant DeCeccoPrincipal| FisherPeak ManagementNorth Vancouver, British Columbia, Canada
Aug 01, 2025 10:18 AM
Replying to Adeboriota Ajulo
...
Is there a negative effect in using Gen AI ?
There can be many. AI tends to hallucinate meaning that it will makeup information, there are ethical considerations where the model could be bias, there are legal issues with data confidentiality and liability for AI's answers used in communicated to customers or the general public. Saving Changes...
Grant DeCeccoPrincipal| FisherPeak ManagementNorth Vancouver, British Columbia, Canada
Aug 01, 2025 10:18 AM
Replying to Adeboriota Ajulo
...
Is there a negative effect in using Gen AI ?
There can be many. AI tends to hallucinate meaning that it will makeup information, there are ethical considerations where the model could be bias, there are legal issues with data confidentiality and liability for AI's answers used in communicated to customers or the general public. Saving Changes...
Grant DeCeccoPrincipal| FisherPeak ManagementNorth Vancouver, British Columbia, Canada
Jul 29, 2025 11:26 AM
Replying to Sanjay Srivastava
...
yes it improves, but make sure you keep it grounded so that it doesn't start hallucinating; and more important you keep it validating towards logical and realistic outcome.
Always check and valid the output you get from AI. Saving Changes...
Multi-shot prompting works well in my case.
As it's well said in the AI ethics, give the output a check with your own eyes and if you're not satisfied with the answers, have a short chat with your pal for some feedback and redo the output to get better answers.
I have a few AI buddies with different names as my personal secretary, mentor and private diary.
I guess this is working well because none of the subjects gets messed up across the chats if you ask them to behave as a consistent individual. Ultimately not having to repeat your self providing the persona each time the conversation gets off track. Saving Changes...
Hi! In my experince, refining a prompt is often the turning point between an average output and an outsanding one. Even small adjustments, like adding context, specifying tone, or defining structure can completely transform the quality, mking the Ai's response far more relevant and actonable. Saving Changes...
Be more precise and give more detail to help the AI with what we want to output to be. Saving Changes...
Nagendra ReddyProgram Manager| KyndrylHyderabad, Telangana, India
Refining the Prompt usually helps to tailor to the needs and fitment for the relevant task and situation.
conversation and refining is a good way to make the output useable to the various situations and adapt and modify in due course of time to accomodate the things that are relevant in the context. Saving Changes...
In my experience, refining a prompt has significantly improved output quality by making the AI’s responses more precise, relevant, and aligned with my needs. I’ve learned that clearly defining the role I want it to take, the objective I want to achieve, and the context of the request sets a strong foundation. From there, I iterate—clarifying details, requesting refinements, or specifying the desired format—until the response fully meets the intended purpose. This approach consistently transforms generic outputs into highly valuable results.
PMI Team Member
