Tag: Chain of Thought

  • From Bullet Points to Descriptions: Chain of Thought Prompts in Action

    From Bullet Points to Descriptions: Chain of Thought Prompts in Action

    Introduction

    So, what is this “Chain of Thought Prompting”? It’s a fancy term for breaking down a complex task into smaller, manageable ones and guiding the AI step by step along the way. Think about how this blog post is being written—you brainstorm a title first, then outline the structure, and finally, you write each section piece by piece. That’s chain of thought prompting. You’re using the context of what’s already done to guide what comes next, ensuring the AI delivers the output you’re aiming for.

    Writing product descriptions is something I deal with regularly, and it can be tricky. So today, we’re going to learn together how to use chain of thought prompting to create perfect Amazon Bullet Points, craft compelling descriptions, and even brainstorm ideas for info-graphics that tie everything together seamlessly.

    Amazon is just the example for today—you can use this approach for any content, text, or project that feels a bit complex. And since I recently bought myself a Garmin Epix Pro Smartwatch, that’s the product we’ll be using as our example.

    Let’s get rolling.

    Starting Small: Creating Amazon Bullet Points

    Alright, first things first—we need some bullet points. These are the main selling points of our product, and they’re one of the most critical parts of a listing.

    Before jumping into writing, let’s take a step back and collect the data. What are the most important parts of a listing? When I start writing a new description, I like to prepare a quick outline of key points that will provide the AI with context about what we’re writing. Think of this as a cheat sheet for the AI.

    Disclaimer: I’m keeping it simple here, and I might miss a unique selling point (USP) or two. In a real-world scenario, I’d spend more time researching to ensure every detail is covered. Remember, the more information we give the AI, the better the output.

    To make this reusable and easy to understand, I’ll divide the prompt into headings. (And I’ll share a Google Doc with all of this later, so you can adapt it for your own product, service, or whatever complex task you’re tackling.)

    Product Info

    • Name: Garmin Epix Pro Gen2 47mm Smartwatch
    • Brand: Garmin
    • Category: Smartwatch
    • Description: High-quality smartwatch
    • Unique Selling Points (USPs):
      • Up to 16 days of battery life in smartwatch mode
      • Up to 30 hours of battery life in GPS mode
      • 24/7 health and wellness monitoring
      • AMOLED Display
      • SatIQ technology and multi-band GPS
      • Bright LED flashlight
      • Additional Features: Touchscreen, Time Display, Bluetooth, Gesture Control, Multisport Tracker, Text Messaging, Activity Tracker, Alarm Clock, GPS, Notifications, Stress Tracking, Contactless Payments
    • Dimensions: Screen size 47mm

    Instructions

    1. Each bullet point should be a maximum of 250 characters.
    2. Use an engaging and professional tone that resonates with the target audience.
    3. Highlight the benefits and unique selling points (USPs) of the product.
    4. Address specific problems and needs of the target audience.
    5. Structure the bullet points so that the most important information comes first.

    Task

    Generate five bullet points for Amazon that promote a high conversion rate. These bullet points should:

    • Highlight the specific benefits, technical features, and unique selling points of the Garmin Epix Pro.
    • Use persuasive language that appeals to both emotional and rational buying motives.
    • Be clear, precise, and sales-oriented.

    AI Persona

    You are a skilled E-Commerce Marketing Expert specializing in crafting persuasive, high-converting content for online product listings and promotional materials. Your expertise includes:

    • Writing engaging product descriptions and bullet points optimized for platforms like Amazon.
    • Highlighting unique selling points (USPs), key benefits, and technical features that resonate with target audiences.
    • Brainstorming and creating visually appealing and compelling content ideas for info-graphics and digital promotional campaigns.
    • Using a blend of emotional appeal and rational persuasion to address customer needs and pain points.
    • Ensuring all content aligns with the product’s brand voice and marketing goals while adhering to platform-specific guidelines.

    That’s our prompt! Just copy-paste it into an AI with a big enough context window (like Gemini or ChatGPT), and let it rip. Watch as it generates persuasive, high-converting bullet points in seconds.

    Now go on with evaluating and iterating. Check if everything is covered—if something doesn’t sit right, ask the AI to change it until you’re happy with the result.

    Now, here’s the fun part: I didn’t come up with the persona description myself—I asked the AI to create it for itself. If you’re unsure how to write a prompt, just tell the AI what you want it to prompt about, or paste your version and ask the AI to make it better.

    Expanding to the Product Description

    Alright, now that we have our bullet points, we can keep the chat open and move on to the next step: crafting the product description.

    If you’ve read my previous post, you’ll notice that we’re following the same pattern as always: Task, Context, References (not needed in this case), Evaluate, and Iterate. This pattern keeps things simple and consistent, and it’s exactly what we’ll use for the next part.

    For the description, we want to keep the context open so that it feels like a continuation of the bullet points. This ensures everything fits together neatly and flows naturally. In practical terms, this means we don’t need to rehash details the AI already knows—such as the AI Persona or product information—because, if the AI’s context window is big enough, it will still “remember” those details.

    Here’s an example of what our prompt might look like:

    Instructions

    1. Length: Approximately 2000 characters.
    2. Structure (not headings): Introduction, Benefits, Target Audience, Details, Call-to-Action.
    3. Tone: Clear, friendly, and direct, with a focus on problem-solving and building trust.
    4. Call-to-Action: End the text with a strong purchase appeal emphasizing product benefits.
    5. Use vivid language to highlight practical benefits and advantages.
    6. Incorporate keywords to optimize the text for SEO.

    Prompt:

    Based on all the information you have—such as Product Info, Bullet Points, and general style—generate a description for Amazon aimed at increasing sales by improving the conversion rate.

    Keyword Phrases:

    • “Best GPS smartwatch for adventure.”
    • “Premium fitness tracker with AMOLED display.”
    • “Smartwatch with long battery life for outdoor enthusiasts.”

    If we’ve done a good job with the previous step, this part should flow seamlessly. The AI can reuse the knowledge we’ve provided, maintain the context, and deliver a result that aligns perfectly with the bullet points.

    Quick Reminder: The keywords and information here are just placeholders—I didn’t do detailed research on this product or conduct any keyword analysis. If this is the product you’re actually trying to sell, do your research! A solid foundation makes a huge difference in creating a high-converting product description.

    Taking It Further: Beyond the Description

    Once you’ve evaluated and iterated on the product description until it’s perfect, you can keep going—all within the same context. This is where things get exciting.

    For example, you could ask the AI to come up with ideas for info-graphics or other forms of visual communication. Honestly, I’ve been surprised more than once by the creativity of its suggestions. Even if it doesn’t give you a final design, it’s great for getting inspiration and sparking your own ideas.

    Here are a few other ways you could expand on what you’ve already created:

    1. Banner or Landing Page Ideas:
      Use the AI to generate taglines, hero section ideas, or even full-page layouts. Want to brainstorm catchy phrases or unique selling points to highlight in a landing page? The AI can help you refine those in no time.
    2. Blog Post Creation:
      You’ve already provided the AI with a wealth of information in the context window. Why not ask it to reformat your product description into a blog post? Start with an outline and let the AI adapt the content, expanding on the details in a format tailored for blog readers.
    3. Social Media Campaigns:
      Turn your product description and bullet points into engaging posts for social media. You can also ask the AI for ideas to create short-form content like reels, tweets, or TikTok captions. This could be anything from product highlights to how-to-use scenarios.

    The Power of Repurposing

    The beauty of keeping all your information in the same context window is that it’s easy to reformat and re-purpose. Whether it’s for info-graphics, banners, blogs, or a full marketing campaign, just tell the AI what you want to create and let it guide you.

    The key here is iteration and inspiration—you’re not just relying on the AI to do all the work but using it as a creative partner to get started and refine your ideas.

    The Secret Sauce: Iteration and Guidance

    From what I’ve noticed, guiding the AI with a Chain-of-Thought process—starting small, building on previous outputs, and iterating—yields some of the best results. Iteration is where the magic happens.

    Think of yourself as the pilot of a ship: you’re steering the AI, making course corrections, and providing the input it needs to deliver exceptional outputs. When something doesn’t fit, rework it. Add your own ideas. Tweak the prompt and try again. This process not only ensures the results align with your vision but also makes them uniquely yours.

    These days, everyone seems to be searching for shortcuts. Just look at YouTube—videos about AI that get the most clicks are often the ones promising to “automate everything completely.” But here’s the thing: if you remove the human element—the evaluation, iteration, and repurposing—your outputs will likely be mediocre at best.

    When you actively guide the AI, step by step, improving and iterating along the way, your results will stand out. Why? Because you’re constantly adapting and refining the output to better suit your needs. This way, your final product isn’t just “okay”—it’s exceptional.

    Conclusion: Is AI the Promised Land That Will Automate All Our Jobs?

    No, not at the time of writing this article. In my opinion, we’re not as close to full automation as marketing might try to sell us.

    That said, AI is incredibly powerful and works best when guided by a human with a clear vision. The more I use AI, the clearer it becomes: its true power lies in collaboration—not in blindly generating everything without a second thought.

    If you embrace this technology and use it as your sidekick—your editor, coach, or brainstorming partner—while iterating, evaluating, and building on the context you provide, AI really shines. It can handle complex problems, save you time, and help you focus on the creative, meaningful parts of your work.

    AI is here to stay. So let’s embrace this exciting new technology and incorporate it into our lives to automate the boring stuff and make space for the things that truly matter.

    Thanks for reading, and stay tuned for more!