The Basics of Prompt Writing

Prompt engineering is the process of designing and refining input so that ChatGPT gives more accurate answers to what you want. The key is how you ask, and whether the result is a summary, report, or analysis, clear instructions greatly determine performance.

There is no single perfect prompt. It's important to change your wording as if talking to a person, adjust the tone, and iterate to refine it into the most useful form.

To write a good prompt, you should first clearly state what ChatGPT should do, for whom, and why. Where possible, start the request with verbs like plan, draft, or research, and it's better to present a concrete task right away, such as a schedule or sales report.

Providing background information and materials together further improves answer quality. If there is context that's needed, such as travel companions, preferences, or attached files and documents, include them together, and when you need to use external materials, you can also use file and app integrations.

The response format should also be specified in advance. Adding constraints such as tone, length, target audience, and output format results in more tailored outcomes.

• You can request a table format, like a 7-day schedule table.
• If it's for an executive report, you can specify the style, like a formal executive summary.
• If there are multiple conditions, it's advantageous to break a large task down into smaller steps.
• Specifying priorities like accuracy, creativity, or speed helps better guide the direction of the answer.

Even for the same question, the more context and constraints you add, the more specific and useful the answer becomes. Rather than simply asking a question and stopping there, continuously revising the prompt until you get the desired result is the most practical approach.