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?
Rod PancineLead UX/UI Designer| KIAI Agency IncMaple Ridge, BC, Canada
In my experience, refining a prompt can be like turning a sketch into a full painting. With GenAI, I’ve seen how a vague blog idea turned into a structured book outline — just by iterating and sharpening the prompt. Sometimes it feels like I wrote a whole book when all I wanted was a post lol. Saving Changes...
Albert GisoreProgram and WASH Advisor| Malteser InternationalNairobi, Kenya
Refining a prompt as enabled a better and quality access of needed feedback or output based on the case at hand. Additionally, the clear summarized version of the output , depicts and includes very necessary story inputs that i was seeking and researching for. Continuous use and application is much needed majorly for complex tasks. Saving Changes...
Eric BrooksConsultant| Michigan MedicineMichigan, United States
Once I feed a prompt with granular details, my output quality outcomes show a favorable distinction between iterations. As I become more detailed in my prompts, my output quality outcomes become more and more useful. Saving Changes...
Refining a prompt can dramatically transform the quality, relevance, and usefulness of GenAI outputs it's like tuning a radio from static to crystal-clear sound. Prompt engineering isn’t just tweaking words it’s shaping the AI’s lens Saving Changes...
Refining a prompt can dramatically transform the quality, relevance, and usefulness of GenAI outputs it's like tuning a radio from static to crystal-clear sound. Prompt engineering isn’t just tweaking words it’s shaping the AI’s lens Saving Changes...
Guillermo Vazquez-ToroProgram Director| Medical College of WisconsinGreendale, Wi, United States
In my GenAI work as a healthcare PM, refining prompts from vague asks to structured briefs consistently turns “nice prose” into decision-ready output: a loose “compare RTLS vendors” became “produce a 1-page, 7-column comparison for a 400-bed Epic hospital with FDA status, integration method, 3-year TCO, timeline, and ≤18-month sources,” which cut hallucinations and enabled apples-to-apples steering-committee review; “summarize CMS telehealth rules” became “summarize 2025 behavioral health reimbursement changes with CPTs, POS, audio-only allowances, effective dates, and cite CMS by rule name + publication date,” which removed generic noise and made compliance sign-off feasible; and “clean up meeting notes” became “extract action items as JSON with owner, task, due date, dependency, risk, status, inferring dates from sprint end and flagging assumptions,” which produced import-ready tasks. The pattern that drives the jump in quality is simple: Goal + Context + Constraints + Evidence + Format + Checks (ask for dated citations, define acceptance criteria, provide schemas/examples, keep temperature low for factual work, and require an assumptions list).
Saving Changes...
David Gonzalez De LaraPROCESS OPTIMIZATION MANAGER| Stevin Rock LLCRas Al Khaimah, United Arab Emirates
Significantly refining a prompt can greatly impact the quality of the output; however, the degree of impact depends on the context. For instance, crafting a prompt to evaluate a pumping system project requires a high level of specificity, where each adjustment provides more detailed and actionable insights. Conversely, when creating a prompt for organizing a project matrix, the changes tend to have a less pronounced effect. Understanding this distinction is crucial for applying prompt engineering effectively and achieving optimal results across various use cases. Saving Changes...
Roland SilvaBusiness Operations Project Mangaer| Texas Department of TransportationElgin, Tx, United States
I've been working with Gen AI for not too long but have certainly learned through practical excellence, before taking PMI's courses, that one can certainly overload a prompt. But one can also refine responses through prompt chaining and or using prompt patterns like chain of thought and chain feedback. AI 'has a great memory' and adapts well to question refinement and ReAct. Saving Changes...
Anonymous
improved specificity and quality of delivery Saving Changes...
PMI Team Member
