
Achieve CLEAR FOCUS: The Ultimate Guide to Smart Prompt Engineering
Nov 20, 2024
5 min read
1
16
0


CLEAR FOCUS is a framework for prompt engineering, designed to guide you in crafting precise and effective prompts to get accurate, relevant, and insightful responses from generative AI.
The acronym stands for:

Details on how to use the CLEAR FOCUS method:
Clarity
Do: Use precise language to communicate exactly what you want. For example, instead of “tell me about cats,” try “describe the personality traits of Siamese cats.”
Avoid: Vague or overly broad requests, as they may lead to generalized or unrelated responses.
Leverage Context
Do: Provide background information to help AI understand the scenario. For instance, “Suggest team-building activities for a remote marketing team.”
Avoid: Requests without context, especially if your query could be interpreted in multiple ways.
Examples
Do: Provide examples to clarify your intent. For example, “Suggest fun icebreaker questions like ‘What’s your favorite travel destination?’”
Avoid: Assuming the AI understands niche terms without any guidance.
Ask for Alternatives
Do: Ask for different options to explore various angles. E.g., “Suggest three different strategies for increasing productivity in the workplace.”
Avoid: Relying on a single suggestion if you’re looking for a variety of ideas.
Refine Iteratively
Do: Experiment with variations of your prompt. If the first response isn’t exactly right, adjust your language, add specificity, or clarify the output format.
Avoid: Expecting perfect results from a single prompt. Iteration often leads to better responses.
Format
Do: Specify the format if you need a list, paragraph, table, or step-by-step instructions. For instance, “List three benefits of exercise in bullet points.”
Avoid: Open-ended prompts if you want a structured answer.
Output Constraints
Do: Mention any constraints directly, such as time, word count, or exclusions. E.g., “Write a 50-word summary of sustainable energy that excludes solar power.”
Avoid: Leaving out constraints if they are essential to your purpose.
Chunk Complex
Do: Split complicated prompts into smaller, manageable parts. For example, “List popular digital marketing strategies. Then explain how each is applied.”
Avoid: Long, complex prompts in a single line, as they may lead to incomplete answers.
Use Follow Ups
Do: Engage with the AI to build on initial answers. If a response is too general, follow up with a more targeted question, like “Can you expand on how this applies to remote teams?
Avoid: Moving on if the answer isn’t complete; follow-up prompts can yield more comprehensive responses.
Start General
Do: Begin with a general query and progressively refine. For example, “Suggest a healthy dinner recipe,” then add, “Make it low-carb and under 500 calories.”
Avoid: Starting with excessive details that might limit the AI’s creativity on the first pass.
By following each step, you ensure that your prompts are precise, well-structured, and aligned with your specific needs. This method empowers you to craft smarter prompts, which in turn leads to richer, more useful AI responses, ultimately saving time, enhancing productivity, and enabling you to harness Generative AI's full potential effectively.

Welcome to CLEAR FOCUS, the ultimate method for crafting powerful prompts that unlock the full potential of Generative AI.
Want to know how it works? Watch this video to discover the secret to getting accurate, relevant, and game-changing responses every time you engage with Generative AI!

Let's have some fun implementing the CLEAR FOCUS process.
If you need assistance logging into a Generative AI application, click here.
Follow the below steps to experience how CLEAR FOCUS will enable better Generative AI results.
Objective:
I'm hosting a pizza party for friends who are food enthusiasts with a global palate—they've sampled pizzas from around the world. I want to "wow" them with something extraordinary, using Generative AI to craft a truly unique pizza. I'll apply the CLEAR FOCUS method to refine my prompts and achieve the best results.
Note: I’ll be using Microsoft’s CoPilot, so your AI responses may vary slightly.
Initial Prompt:
Type: What are good pizza toppings?
Result: Bland and uninspiring. Time to use CLEAR FOCUS to improve!

Step-by-Step CLEAR FOCUS Process
1. Clarify: To set clearer expectations, I refined the prompt:
Type: What are unique and flavorful pizza toppings?

2. Leverage Context: Adding context ensures the AI understands the scenario better.
Type: What are unique and flavorful pizza toppings for people who have tried pizzas from around the world?

3. Examples: Including preferences helps tailor the response.
Type: What are unique and flavorful pizza toppings for people who have tried pizzas from around the world and prefer foods from Japan, Mexico, and Singapore?

4. Ask for Alternatives: Diverse options add creative possibilities.
Type: What are unique and flavorful pizza toppings for people who have tried pizzas from around the world and prefer foods from Japan, Mexico, and Singapore. Provide options to blend flavors from the three countries.

5. Refine Iteratively: I noticed the results lacked fun and creativity, so I added that to the prompt.
Type: What are unique and flavorful pizza toppings for people who have tried pizzas from around the world and prefer foods from Japan, Mexico, and Singapore. Provide options to blend flavors from the three countries. Add some fun to the process.

6. Format: I need recipes suitable for an amateur chef.
Type: What are unique and flavorful pizza toppings for people who have tried pizzas from around the world and prefer foods from Japan, Mexico, and Singapore. Provide options to blend flavors from the three countries. Add some fun to the process and ensure the recipes are simple enough for someone with basic cooking skills.

7. Output Constraints: I included dietary restrictions to refine the results further.
Type: What are unique and flavorful pizza toppings for people who have tried pizzas from around the world and prefer foods from Japan, Mexico, and Singapore. Provide options to blend flavors from the three countries. Add some fun to the process and ensure the recipes are simple enough for someone with basic cooking skills. Do not include recipes with spicy ingredients or beef.

8. Chunk Complex: To address soggy pizza crusts, I added a new question to tackle this issue.
Type: What are unique and flavorful pizza toppings for people who have tried pizzas from around the world and prefer foods from Japan, Mexico, and Singapore. Provide options to blend flavors from the three countries. Add some fun to the process and ensure the recipes are simple enough for someone with basic cooking skills. Do not include recipes with spicy ingredients or beef. Explain how I can ensure my pizza crust is not soggy, including oven temperature and baking time.

9. Use Follow-Ups: I asked follow-up questions to dig deeper into ensuring a crispy crust.
Type: What are unique and flavorful pizza toppings for people who have tried pizzas from around the world and prefer foods from Japan, Mexico, and Singapore. Provide options to blend flavors from the three countries. Add some fun to the process and ensure the recipes are simple enough for someone with basic cooking skills. Do not include recipes with spicy ingredients or beef. Explain how I can ensure my pizza crust is not soggy, including oven temperature and baking time. What else can I do to ensure a crispy pizza crust?


10. Start General:
We began with a broad question and gradually refined it using the CLEAR FOCUS process. The prompts became clearer, more targeted, and aligned with the specific needs. This structured approach led to richer, more useful AI responses, saving time and enhancing productivity.
Pro Tip: Instead of appending details with every step, you can "converse" with Generative AI, refining the same prompt dynamically:
Type: What are unique and flavorful pizza toppings for people who have tried pizzas from around the world and prefer foods from Japan, Mexico, and Singapore.
Type: Blend the flavors from the three countries.
Type prompt: Make sure the recipes are doable for a person with basic cooking skills
This conversational refinement simplifies the process while maintaining the clarity and precision of your prompts.


