99% Of People STILL Don't Know The Basics Of Prompting (ChatGPT, Gemini, Claude)

Chat GPT Prompt Engineering: Getting the Best Results

Key Concepts: Goal-oriented prompting, role assignment, context provision, constraints and outputs, example-based learning, iterative refinement, prompt generation using AI, custom GPTs for prompt reusability.

1. Starting with a Goal in Mind

• Problem: Generic prompts like "Please write me an email" yield generic and unhelpful responses.
• Solution: Define a specific goal or action you want Chat GPT to accomplish.
• Example 1 (YouTube Titles): "Give me three YouTube titles that would probably get more than 10% click-through rate on my video about Chat GPT prompt engineering."
  • Goal: Generate titles to increase click-through rate.
  • Specifics: Requesting three titles for more options.
• Example 2 (Blog Intro): "Write me a 100-word blog intro for the blog post with a keyword, how much does plumbing services cost?"
  • Goal: Create a blog introduction.
  • Specifics: 100-word limit, keyword inclusion.

2. Providing a Role

• Concept: Tell Chat GPT who it is acting on behalf of; define its behavior.
• Importance: Aligns Chat GPT's thinking with the desired persona.
• Example 1 (YouTube Titles): Adding "You're an intelligent AI automation expert that understands psychology, persuasion, and how to get people to take action" to the title generation prompt.
  • Result: Titles become more persuasive and action-oriented.
• Example 2 (Blog Intro): Adding "Please act as a comedian and create a hilarious blog post" to the plumbing blog intro prompt.
  • Result: Introduction becomes humorous and engaging, contrasting with a typical dry blog post.
  • Quote: "Ever googled, how much does a plumbing service cost and felt like you needed a second mortgage just to unclog a toilet?" - Example of humorous output.

3. Providing Context

• Concept: The more details you provide about the situation, the better Chat GPT can understand your needs.
• Importance: Context is a crucial ingredient for tailored results.
• Example 1 (YouTube Titles): Adding "You are in the AI automation space creating YouTube videos for people looking to learn more about Chat GPT, NAN, Make.com, etc. Your target demographic is 25 to 55 year old males and their interests are in business and entrepreneurship."
  • Result: Titles become more targeted towards AI entrepreneurs and business owners.
• Example 2 (Blog Intro): Adding "We've been running a family-based plumbing business for 35 years, have a five-star rating, have served over 10,000 customers in the New York area" and providing source material with a specific tone of voice.
  • Source Material Example: Passage from a humorous article about writing with humor.
  • Result: Introduction incorporates the desired tone and context, making it more engaging and personalized.
  • Quote: "We're a family-owned plumbing biz that's been keeping New York's pipes happier than a pigeon with a bagel for 35 years." - Example of context-aware and humorous output.

4. Constraints and Outputs

• Concept: Define the desired output format and any limitations (e.g., character limits).
• Importance: Ensures the output conforms to your specific requirements.
• Example 1 (YouTube Titles): "Please output the data in a CSV file with the title on one column and the CTR or click-through rate in another column."
  • Alternative: Output in a table format.
• Example 2 (Blog Intro): "Please output the data in less than 100 words and make sure that it's in HTML format."
  • Alternative: Output in Markdown format.
  • Technical Terms: CSV (Comma Separated Values), HTML (HyperText Markup Language), Markdown (lightweight markup language).

5. Providing Examples

• Concept: Give Chat GPT real examples of past work that performed well.
• Importance: Guides Chat GPT towards the desired style and quality.
• Analogy: Like studying for a test by reviewing the answers.
• Example 1 (YouTube Titles): Providing two examples of successful YouTube titles.
  • Result: Generated titles are more likely to perform well, leveraging proven strategies.
• Example 2 (Blog Intro): Providing an excerpt from a successful blog post in the wedding industry.
  • Result: Introduction incorporates elements from the example, such as humor and engaging language.

6. Follow-Up Questions (Iterative Refinement)

• Concept: Ask follow-up questions to refine the output and improve results.
• Importance: Allows for continuous improvement and fine-tuning.
• Example (YouTube Titles): "Can you please rate each of those titles out of 10 and provide ways that each one could be improved?"
  • Follow-up: "Please go ahead and incorporate your suggestions."
  • Technical Term: Iterative (involving repetition to improve).

7. Bonus Tip 1: AI-Generated Prompts

• Concept: Use Chat GPT to generate prompts for you, saving time and effort.
• Process: Ask Chat GPT to write a prompt that would generate the best YouTube titles and rank them out of 10, including a goal, context, role, and output.
• Key Ingredients for Prompts: Goal, context, role, and output.

8. Bonus Tip 2: Custom GPTs for Prompt Reusability

• Concept: Create custom GPTs to store and reuse successful prompts.
• Process:
  1. Create a new GPT.
  2. Define its description and instructions (paste the successful prompt).
  3. Save the GPT for personal use.
• Benefit: Bypasses the need to rebuild the same prompt every time, saving significant time.

Synthesis/Conclusion

The key to effective Chat GPT prompt engineering lies in providing clear goals, defining roles, offering relevant context, specifying output formats, giving examples of successful work, and iteratively refining the results through follow-up questions. Furthermore, leveraging Chat GPT to generate prompts and creating custom GPTs for prompt reusability can significantly enhance efficiency and productivity. By implementing these strategies, users can unlock the full potential of Chat GPT and achieve optimal results in a fraction of the time.