Output Requirements
Defining Your Desired Result
Take precise control over tone, style, length, and structure to transform raw AI output into polished deliverables.
What You'll Learn
- • Format specifications that work
- • Tone and style instructions
- • Length and structure control
- • Template-based output design
Why Output Requirements Matter
Even with perfect persona and context, if you do not specify how you want the output formatted, you will get generic responses that require significant editing. Output requirements transform raw AI output into polished, ready-to-use deliverables.
Format Specifications
Specify the exact format you want:
- Structure: Bullet points, numbered list, table, prose, JSON, markdown
- Sections: Headers, subheaders, introduction, body, conclusion
- Elements: Code blocks, quotes, callouts, examples
Format Examples
Blog Post:
"Format as a blog post with an engaging headline, 3-4 paragraph introduction, three main sections with subheadings, and a conclusion with a call-to-action."
Report:
"Format as a structured report with: Executive Summary, Key Findings (bulleted), Analysis (3 paragraphs), Recommendations (numbered list), and Next Steps."
Tone and Style Instructions
Define the voice and style you want:
- Tone: Professional, friendly, authoritative, conversational, technical
- Style: Formal, casual, academic, journalistic, marketing
- Audience: Who is reading this? What is their level of expertise?
Length and Structure Control
Be specific about length:
- Word count: "Under 200 words" or "Approximately 500 words"
- Paragraph count: "3-4 paragraphs" or "5 paragraphs maximum"
- Section count: "Three main sections" or "Five key points"
Template-Based Output Design
For consistent outputs, provide a template structure:
Format:
- Title: [Engaging headline]
- Introduction: [Hook + problem statement]
- Section 1: [First key point]
- Section 2: [Second key point]
- Section 3: [Third key point]
- Conclusion: [Summary + call-to-action]Moving Forward
Output requirements are the third pillar. Combined with persona and context, you now have three of the four core techniques. Next, you will learn few-shot examples, the final pillar that shows the AI exactly what good output looks like.