Prompt engineering is the skill of crafting inputs that get the best outputs from AI. It’s the difference between mediocre results and outputs that genuinely save you time and improve your work.
Core Techniques
Role Assignment
“Act as a [role] with expertise in [domain].” This frames the AI’s response style and knowledge level. “Act as a senior data analyst” produces different output than “Act as a marketing intern.”
Few-Shot Prompting
Show the AI examples of what you want before asking for new output. Provide 2-3 examples of input-output pairs, then give your actual input. The AI matches the pattern of your examples.
Chain of Thought
“Think through this step by step” or “Show your reasoning.” This forces the AI to work through problems methodically, producing more accurate and well-reasoned outputs.
Constraint Setting
Define what you want AND what you don’t want. “Write in active voice, under 200 words, no jargon, with specific examples” produces tighter, more useful output than open-ended requests.
For ready-to-use prompts, browse ChatGPT Prompts and Claude Prompts.