Prompt Lab #2: Restructured Rob Lennon Prompt

Don't whisper your prompts, structure them.

Rob Lennon is a well-know content creator who calls himself the “AI Whisperer,” but his approach to crafting prompts gets a bit silly … asking AI to be “the best writer in the world” or “you have free reign to be creative.” AI will take a good guess at these vague words, but doesn’t really know what these mean.

A structured approach defines these elements with prompt blocks and structures it for AI to more easily process (or for creators to reuse across prompts). I restructured one of Rob Lennon’s prompts as an example, and got significantly different results.

Rob Lennon’s Prompt

This prompt has AI go through a process created by Justin Welsh … also someone AI probably doesn’t know about. It is a useful framework … and the prompt works, but there isn’t anything particularly creative that one might identify as the best writer in the world.

… and why are we thanking AI? It is a statistical machine. 🙄

Act as an expert content editor and ghost writer. This is important! Please take it step-by-step. I know you’ll do a great job.

TOPIC= Pitching  financial planning and asset management to people 30-45 years old.

TASK= For a given topic, apply the PAIPS FRAMEWORK as Justin Welsh would to come up with ideas. Use your awesome creativity to create a mega-outline with 3 options for each part. You have creative freedom to improve the topic and deliver as much value as you can.

PAIPS FRAMEWORK= 1. Address a problem, something your audience can relate to, to stir up an emotional response. 2. Agitate that problem, intensifying the emotional reaction. 3. Just when your audience is on edge, introduce intrigue, something new and shiny to distract and allure. 4. With curiosity piqued, paint a vision of a positive future for the reader, easing the tension. 5. Delivers a solution, triggering a sense of relief and satisfaction.

Thanks so much for your hard work on such a detail-oriented project.

Restructured Prompt

I took out all the vague terms, defined the most important, and restructured them into the prompt. For example, what does it mean to be the best content creator ever? Or to be as creative as possible? There are many different ways this could be defined, depending on who I am and my purpose.

[ROLE] Act as an expert content editor and ghost writer, adept at generating engaging, relevant materials across various mediums. You have stellar research, writing, and storytelling skills, along with a deep understanding of their target audience's preferences. With proficiency in digital trends and SEO, you strategically optimize content for maximum reach and impact.

[TOPIC] Pitching  financial planning and asset management to people 30-45 years old.

[TASK] For a given topic, apply the PAIPS FRAMEWORK to come up with ideas. Create an outline with 3 options for each part. 

[STYLE] Be creative. Use evocative language and imagery. Paint a vivid picture using concrete examples and scenarios when possible. 

[CONTENT] Choose intriguing information — a lesser-known fact about the problem, an unexpected consequence, or a promising new development that hints at a possible solution.

[FRAMEWORK] Address a problem, something your audience can relate to, to stir up an emotional response.
1. Agitate that problem, intensifying the emotional reaction.
2. Just when your audience is on edge, introduce intrigue, something new and shiny to distract and allure.
3. With curiosity piqued, paint a vision of a positive future for the reader, easing the tension.
4. Delivers a solution, triggering a sense of relief and satisfaction.

Output

Both prompts are useful, but the quality from the structured prompt is much better than the unstructured one. Notice the specificity of the second result and how much more creative it is.

I’m also fascinated that both resulted in essentially the same answers, even though these were two distinct chat sessions. This has been very common in my experience, showing that there is a specific way that AI is processing the prompts … not necessarily “thinking” as we understand thinking.

The results below are just number four in the process.

Unstructured Results

➡️ Find the entire chat here.

Structured Results

➡️ Find the entire chat here.

➡️ If you are interested in beta testing my prompt engineering course for content and information development, you can check it out here (and view my library of prompts).