Write Prompts Like A Pro (without overthinking it)

Key Concepts

• RICECO Framework: A six-step framework for improving AI prompt effectiveness: Role, Instruction, Context, Examples, Constraints, Output Format.
• Prompt Engineering: The process of crafting effective prompts to elicit desired responses from AI models.
• Few-Shot Prompting: Providing examples within a prompt to guide the AI model's output.
• IC Framework: A condensed three-step version of RICECO (Instruction, Context, Constraints) for common use cases.
• Evaluate, Iterate, Optimize (EIO): A process for refining AI outputs and prompts for continuous improvement.

RICECO Framework Breakdown

1. Role (Importance: Medium)

• Definition: Assigning the AI model a specific persona or role to influence the tone, perspective, and depth of the response.
• Example: Instead of "Give advice on how to get better sleep," use "You are a board-certified sleep doctor. Give me advice on how to get better sleep." or "You are a sleep-deprived parent writing to other parents, give advice on how to get better sleep."
• Impact: Significantly alters the output's tone, priorities, and level of detail.
• Application: Useful for tailoring responses to specific audiences or scenarios.
• Example: For Skill Leap AI, different roles (founder pitching to investors vs. AI YouTuber recommending to audience) resulted in vastly different explanations and value propositions.

2. Instruction (Importance: High)

• Definition: The core task or request you are giving the AI.
• Key Point: Specificity is crucial. Vague instructions lead to generic outputs.
• Example: Instead of "Write me an engaging YouTube short about prompting tips," use "Write a 60-second YouTube short script about prompting tips using a curiosity gap hook and a scroll-stopping visual anchor."
• Actionable Advice: Replace vague terms like "engaging" or "interesting" with concrete details. For example, instead of "make it interesting," try "start with a surprising stat and end with a thought-provoking question."

3. Context (Importance: High)

• Definition: Providing background information to make the output relevant, useful, and aligned with your goals.
• Elements of Context:
  • Audience: Who is this for?
  • Background: What's the business, platform, or scenario?
  • Purpose: What are you trying to accomplish, and why does it matter?
  • Tone/Perspective: What kind of voice or framing should it use?
• Example: Instead of "Write a 500-word blog post about AI video tools," use "Write a 500-word blog post about AI video tools for small business owners who want to turn podcast episodes into YouTube shorts. The goal is to explain the value in simple, practical terms."
• Key Point: More context is generally better, as long as the core instruction remains clear.
• Application: For complex tasks, provide a "context dump" of multiple pages.

4. Examples (Importance: Medium)

• Definition: Showing the model exactly what kind of output you're looking for in terms of structure, tone, formatting, or logic.
• Technique: Few-shot prompting (including examples in the prompt).
• Example: When drafting a newsletter, include a real newsletter you've written before as a reference for tone and format.
• Application: Useful for writing prompts, data formatting, and even multi-art logic.
• Example: Using JSON prompting for video models by providing examples of desired output formats.
• Impact: Provides an "anchor" for the model to imitate or remix.

5. Constraints (Importance: High)

• Definition: Defining limits, rules, or must-haves that the model should follow.
• Purpose: Tightens the output and eliminates bad habits (e.g., being too wordy, vague, repetitive, or safe).
• Examples: "Keep it under 100 words," "Avoid buzzwords," "Don't use hashtags," "Use a warm conversational tone," "Include a quote and a stat in the first paragraph."
• Key Point: Different from context; context provides background, while constraints provide boundaries.
• Application: Especially useful for long-term AI collaboration, content workflows, and automation.
• Implementation: Add constraints to custom instructions or projects in tools like ChatGPT.

6. Output Format (Importance: Low)

• Definition: Specifying how you want the response structured.
• Purpose: Makes outputs cleaner and more usable.
• Examples: Bullet points, table, JSON, markdown with headers, tweet thread, three-act script, carousel slides, flowchart.
• Application: Useful for comparisons, data analysis, and generating code snippets.
• Example: Instead of asking for a comparison of three tools, specify "Give this to me as a three-column table with columns for tool name, key features, and ideal use cases."

RICECO in Action

• Bad Prompt Example: "How can I implement AI and automations into my real estate business?" (Vague, lacks context)
• RICECO Prompt Example:
  • Role: "You are a business growth strategist specializing in AI adoption for businesses."
  • Instruction: "Identify the top AI and automation opportunities to help streamline my business and produce a prioritized action plan that maximizes time savings and revenue growth."
  • Context: (Detailed description of the real estate business, lead sources, and time allocation)
  • Examples: (Potential areas to automate)
  • Constraints: "$400 budget, must be less than 3 hours per week to maintain, not overly technical."
  • Output Format: "Present as a prioritized AI and automation playbook with quick wins, core systems, and long-term growth."
• Result: A customized, actionable plan tailored to the specific business, within budget and time constraints, with projected time savings and revenue growth.

Condensed IC Framework

• Components: Instruction, Context, Constraints.
• Application: Covers 80% of use cases.
• Example:
  • Vague Prompt: "Write a Twitter post about ChatGPT tips."
  • IC Prompt: "Write a Twitter thread with five lessons I've learned from using ChatGPT every day. The audience is indie creators and solo entrepreneurs trying to use AI more practically. Make it clear, non-technical, and casually insightful. Avoid hype or buzzwords. No hashtags."

Evaluate, Iterate, Optimize (EIO)

• Purpose: Turns a good output into a great one and creates reusable prompts.
• Steps:
  1. Evaluate: Interrogate the AI's response. Ask questions like, "What assumptions did you make? What might be missing? Is there anything that feels off? Critique your own output."
  2. Iterate: Treat the AI like a creative partner. Prompt it again with instructions like, "Rewrite this to be more concise. Add a bit of humor. Give me three variations with different pacing."
  3. Optimize: Refine the prompt itself. Cut unnecessary words, clarify the instruction, tighten the context. Ask the model to rewrite the prompt to be more concise and reliable.

Conclusion

The RICECO framework provides a structured approach to crafting effective AI prompts, leading to faster workflows, fewer rewrites, and better results. While all six components can be powerful, the condensed IC framework (Instruction, Context, Constraints) covers the most essential elements for common use cases. The EIO process (Evaluate, Iterate, Optimize) enables continuous improvement of both AI outputs and the prompts themselves, leading to repeatable and reliable systems.