C.A.R.E. Framework

The C.A.R.E. framework—Context, Action, Result, Example—provides a comprehensive structure for crafting effective AI prompts that lead to specific, actionable outputs tailored to your needs.

Introduction

The C.A.R.E. Framework—Context, Action, Result, Example—is a structured approach to prompt engineering designed to produce highly specific, practical outputs from AI systems. This framework excels at generating results that are:

• Contextually Relevant – Grounded in the specific situation or environment
• Action-Oriented – Focused on clearly defined tasks
• Results-Driven – Delivered in formats aligned with precise needs
• Example-Guided – Informed by reference points that shape the output

• Business and professional communications
• Strategic planning and analysis
• Content creation with specific goals
• Technical documentation and guides
• Project planning and proposals
• Any task requiring highly specific outputs

C.A.R.E. Framework Structure

1. Context

The Context component provides critical framing information that helps the AI understand the specific circumstances surrounding your request. Good context focuses on relevant details that directly influence how the task should be approached.

• "We're a B2B SaaS platform targeting enterprise customers with 1000+ employees in regulated industries"
• "I'm preparing for a job interview at a management consulting firm after transitioning from a technical role"
• "Our mobile app has experienced a 40% drop in retention after our recent UI redesign"

• "Our company sells products" (too vague)
• "I need help with my project" (lacks specifics)
• "We have a website" (insufficient context)

2. Action

The Action component clearly articulates what you want the AI to do. This should be specific and directive, using precise verbs that indicate both the task and the type of output you expect.

• "Analyze these user testing findings and identify the top 3 UX issues affecting conversion"
• "Create a 30-day content calendar for our product launch targeting financial professionals"
• "Draft a project proposal template with sections for objectives, scope, timeline, and budget"

• "Help me" (undefined action)
• "Write something" (unclear deliverable)
• "Make it better" (non-specific direction)

3. Result

The Result component specifies exactly what the final output should look like and what qualities it should have. This includes format, style, length, and any specific requirements or constraints.

• "Deliver a 5-point action plan with specific steps, rationale, and expected outcomes for each recommendation"
• "Create a comparison table with features in rows, competitors in columns, and a clear visual hierarchy highlighting our advantages"
• "Produce a conversational script with branching dialogue options for common customer objections, keeping responses under 30 seconds"

• "Make it good" (subjective and unmeasurable)
• "Write a lot" (imprecise quantity)
• "Do a thorough job" (undefined standard)

4. Example

The Example component gives the AI a concrete reference for the style, format, or approach you want. This can be a brief sample, a comparison to a known format, or specific elements you want incorporated.

• "Format similar to McKinsey's situation-complication-resolution structure with clear headings and bullet points"
• "Use the tone and storytelling approach of Malcolm Gladwell—accessible explanations of complex ideas with illustrative anecdotes"
• "Include data visualization styles similar to those on Our World in Data—clean, minimalist graphs with clear annotations"

• "Make it professional" (subjective style)
• "Like a good article" (undefined standard)
• "Use the best format" (non-specific reference)

Example Prompts Using the C.A.R.E. Framework

Example 1: Business Communication

C.A.R.E. Breakdown:

• Context: SaaS platform raising prices (15%), first increase in 3 years, cost increases (23%), new features (37)
• Action: Draft an email announcement explaining the price change
• Result: 300-word email emphasizing value, explaining factors, highlighting features, with timeline and FAQ
• Example: Reference to Slack's transparent 2021 price increase announcement

Example 2: Product Development

C.A.R.E. Breakdown:

• Context: Fitness wearable company, new smartwatch for runners, priorities based on user research
• Action: Create a product requirements document for core running features
• Result: Structured PRD with sections for user stories, requirements, metrics, priorities, specific features
• Example: Reference to Garmin's Forerunner documentation style and categorization

Best Use Cases for the C.A.R.E. Framework

1. Business Strategy and Planning

• Strategic plans
• Competitive analyses
• Business proposals
• Market research summaries

2. Content Marketing

• Case studies
• White papers
• Industry reports
• Blog posts

3. Technical Documentation

• User guides
• Technical specifications
• API documentation
• Implementation guides

4. Customer Communications

• Email sequences
• Support responses
• Crisis communications
• Update announcements

Bonus Tips for Using C.A.R.E. Effectively

💡 Prioritize relevant context: Include only background information that directly impacts how the task should be approached

🎯 Use action verbs: Begin with clear, specific verbs that indicate exactly what you want done

📏 Be specific about metrics and constraints: Numbers, quantities, and limits help define clear results

🔍 Reference familiar examples: Choose examples the AI can understand and emulate

⚙️ Adjust based on results: If the output isn't quite right, refine your context or be more specific about results

Conclusion

The C.A.R.E. Framework provides a structured approach to prompt engineering that consistently delivers highly relevant, actionable outputs. By systematically addressing context, action, results, and examples, you create prompts that provide all the necessary information for the AI to generate exactly what you need.

The framework can be extended with additional components for specialized needs:

• Audience: Defining who will consume the output
• Timeline: Specifying progression or sequence requirements
• Constraints: Adding explicit limitations or boundaries
• Evaluation: Establishing criteria for success