Prompt Engineering Guide: Improving the Prompts
We Use Regularly
Introduction: The Value of Better Prompts
Prompt engineering is the practice of crafting clear, structured instructions to
interact effectively with AI tools such as ChatGPT. While these tools are highly
capable, their usefulness depends heavily on the quality of the input they
receive. In many workplaces, prompts are used casually for tasks such as
drafting emails, generating content, or summarizing meetings. However, vague
or incomplete prompts often result in inconsistent or unhelpful responses.
This guide aims to improve the routine prompts we use in day-to-day tasks by
applying proven prompt engineering strategies. The goal is to help teams get
more accurate, relevant, and efficient outputs while improving understanding of
how to work collaboratively with AI.
Target Audience
This guide is designed for teams involved in content creation, communication,
support, education, or administrative operations—particularly those who are
regularly using AI to support tasks such as writing, summarizing, responding to
queries, or generating ideas.
Prompt Engineering Techniques Used
We will apply the following techniques to improve real-world prompts:
• Role-based Prompting: Framing the AI's perspective or role to match
the task.
• Few-shot Prompting: Providing examples to guide the output style or
structure.
• Chain-of-thought Prompting: Instructing the AI to generate step-by-
step or logically sequenced outputs.
• Clear Instruction Format: Structuring prompts with explicit tasks and
expected outputs.
Common Tasks and Optimized Prompt Examples
1. Writing a Weekly Update Email
Basic Prompt:
Write a weekly update email.
Improved Prompt:
You are a team lead preparing a weekly update for internal stakeholders. Keep
the tone professional and concise. Structure the email into three sections:
1. Completed tasks
2. Ongoing work
3. Next week’s focus
Techniques Used: Role-based, Clear Instruction
2. Summarizing a Meeting
Basic Prompt:
Summarize this meeting.
Improved Prompt:
Summarize the following meeting notes using this format:
• Topics discussed
• Key decisions made
• Action items (include responsible persons if available)
Techniques Used: Chain-of-thought, Clear Instruction
3. Generating Content Ideas
Basic Prompt:
Give blog ideas on remote work.
Improved Prompt:
Act as a content strategist planning a blog series on remote work. Two existing
blog titles are:
• “How to Run Better Zoom Meetings”
• “Avoiding Burnout When Working from Home”
Based on this tone and style, suggest five new titles.
Techniques Used: Role-based, Few-shot
4. Creating an FAQ Section
Basic Prompt:
Create an FAQ for a fitness app.
Improved Prompt:
Develop a beginner-friendly FAQ section for a mobile fitness tracking app.
Focus on five of the most common user questions. Keep the answers clear, brief,
and in plain language.
Techniques Used: Chain-of-thought, Clear Instruction
5. Responding to Customer Support Requests
Basic Prompt:
Write a response to a refund request.
Improved Prompt:
You are a customer support agent. Here is an example response tone:
“Thank you for reaching out. We apologize for the inconvenience...”
Now write a response to a customer who didn’t receive their download link and
is asking for a refund. Offer both a resolution and a refund option.
Techniques Used: Role-based, Few-shot
Before vs. After Comparison
Prompt Style Basic Prompt Improved Prompt
Clarity Vague and general Specific and structured
Role Context Not specified Clear perspective provided
Output Consistency Unreliable Aligned with task expectations
Best Practices for Prompting
Do:
• Be clear about the task and output format
• Include context, roles, or examples
• Break tasks into smaller parts if needed
• Test and revise based on results
Don’t:
• Assume the AI understands unstated goals
• Use vague or one-line instructions
• Ignore formatting or structure
• Expect consistent results without guidance
Conclusion
Prompt engineering is a practical skill that enhances how we interact with AI
tools in everyday workflows. By refining how we phrase and structure prompts,
we can unlock more useful, efficient, and reliable outputs. This guide provides a
foundation to improve commonly used prompts and encourages a more
intentional approach when working with AI.
By applying these techniques consistently, teams can integrate AI more
effectively into routine tasks, saving time and improving the quality of work.
