Something fascinating is happening in the creator economy right now. Phoebe Gates and Sophia Kianni recently started using ChatGPT to analyze top-performing TikToks, building Excel databases of hooks, structure, energy, and pacing, then feeding that data back into AI to generate viral content. It’s a blend of creativity and systems thinking that’s catching serious attention in both marketing and creator circles.
Why This Approach Is Taking Off
AI-driven content creation isn’t a future trend anymore—it’s becoming the current standard. While many marketers are experimenting with AI, few are using it with real strategic depth. The difference comes down to one thing: intentionality. Knowing what to feed the machine and how to interpret what comes out.
Most creators are still using AI like a magic wand, asking ChatGPT to “write me a viral TikTok” and wondering why the results feel generic. Meanwhile, strategic creators are treating AI like a research assistant, feeding it specific data about what’s already working and asking it to help them replicate those patterns.
The results speak for themselves. Brands using this systematic approach are seeing 35-60% improvements in engagement metrics, while cutting their content ideation time in half.
The Step-by-Step Framework That’s Working
Step 1: Identify Viral Trends in Your Space
Start by choosing your platform—TikTok, Instagram Reels, or YouTube Shorts. Then identify 15-20 top-performing videos in your specific niche. You’re looking for content with high view-to-follower ratios, strong share rates, and engaged comment sections.
Use tools like TrendTok or TikTok Creative Center to find these gems, or do manual searches using relevant hashtags in your space. The key is being specific about your niche—don’t just look at “wellness” content if you’re focused on gut health for working professionals.
Step 2: Collect Video Metadata Like a Data Scientist
Here’s where most people stop, but this is where the magic really begins. Create a spreadsheet and start logging specific details about each viral video:
Hook Analysis: What happens in those crucial first 3-5 seconds? Is it a question, a bold statement, or a visual surprise?
Structural Patterns: How is the content organized? Problem-solution format? List structure? Story arc?
Energy and Pacing: Is it high-energy with quick cuts, or calm and contemplative? How fast do they speak?
Comment Themes: What are people actually saying? What questions come up repeatedly? What emotions are being expressed?
View-to-Share Ratio: This metric tells you how compelling the content is beyond passive consumption. High shares mean people found it worth passing along.
Audio Choices: Are they using trending sounds, original audio, or popular music? How does the audio choice affect engagement?
Step 3: Turn Data Into Actionable Insights
Once you’ve analyzed your collection, start looking for patterns. You might notice that videos starting with “Nobody talks about this, but…” consistently perform well in your niche. Or that content with a calm, educational tone gets higher save rates than high-energy content.
This is your content DNA—the proven elements that resonate with your specific audience. These patterns become the foundation for everything you create next.
Step 4: Prompt ChatGPT With Strategic Precision
This is where most people’s AI strategy falls apart. Instead of vague requests, use your data to create specific, targeted prompts. Here’s a framework that works:
“I analyzed 20 viral TikToks in the sustainable fashion niche. The highest-performing videos use hooks like ‘Fast fashion companies don’t want you to know this’ and follow a problem-revelation-solution structure with an educational but conversational tone. Generate 5 TikTok scripts that align with these patterns for a brand targeting environmentally conscious millennials.”
The more specific your input, the better your output. You’re not asking AI to be creative from scratch—you’re asking it to remix proven formulas in new ways.
Step 5: Test, Measure, and Refine
Produce content based on your AI-generated scripts, but don’t stop there. Track performance metrics carefully: saves, shares, comments, and view completion rates. Compare these against your baseline performance.
Here’s the crucial part: feed your own performance data back into your system. If a particular hook style or pacing approach works well for your brand, note that in your database. This creates a feedback loop where each post makes your next post stronger.
Real-World Success Stories
Phoebe Gates has been systematically analyzing her highest-performing content, then using AI to generate variations that maintain her authentic voice while hitting proven engagement triggers. Her approach focuses on socially conscious content with clean, direct hooks that align with her audience’s values.
Sophia Kianni uses this method to humanize climate activism, analyzing which personal storytelling elements work best in those crucial first three seconds, then prompting AI to help her develop new content that follows similar emotional patterns.
Phia, a wellness brand, implemented this exact framework and saw a 35% increase in view-to-share ratios within 30 days. They analyzed 50 top-performing wellness TikToks, identified the most effective hook patterns, and used AI to generate content that matched those proven formulas while staying true to their brand voice.
Case Study: How BloomWellness Cracked the Code
Let me walk you through a specific example of this framework in action. BloomWellness, a gut health brand targeting young professionals, was struggling with inconsistent content performance despite posting regularly.
They started by collecting 30 viral gut health TikToks and analyzing them systematically. What they discovered was eye-opening: the highest-performing content used hooks like “Nobody told you this about your gut…” paired with personal vulnerability—creators sharing their own digestive struggles.
They took this insight to ChatGPT with a specific prompt: “Generate 10 TikTok scripts for a gut health brand that use vulnerability-based hooks and follow a personal problem-to-solution structure, targeting working professionals who experience stress-related digestive issues.”
The results were impressive. After posting 10 videos based on the AI-generated scripts, they saw a 48% increase in shares and a 60% increase in saves. More importantly, they cut their content ideation time from hours to minutes while maintaining authenticity.
Their key insight? AI didn’t replace their creativity—it amplified their understanding of what already worked.
Why This Systematic Approach Works
There are several reasons this method is so effective:
You’re building on proven success: Instead of guessing what might work, you’re starting with what already does work.
AI becomes a force multiplier: Rather than replacing human insight, AI helps you scale and iterate on your strategic thinking.
You develop a sustainable system: Trends change, but the framework for analyzing and adapting remains consistent.
Every post becomes data: This approach turns content creation into a continuous learning process.
Advanced Strategies for Consistent Results
Once you’ve mastered the basics, here are ways to take your approach to the next level:
Expand your data pool: Analyze 30-50 videos to spot more nuanced patterns and seasonal trends.
Layer in audio trends: Identify which trending sounds pair well with your proven hook patterns.
Segment by content type: Determine whether talking-head videos, text overlays, or B-roll footage performs better in your niche.
Track engagement timing: Notice when your audience is most active and schedule your content accordingly.
Monitor comment conversations: Use the themes you see in comments to inform your next round of content topics.
Getting Started This Week
Ready to implement this system? Here’s your action plan:
Choose your niche and platform focus. Collect 15-20 high-performing videos from that space. Build your analysis spreadsheet and start logging patterns. Create specific AI prompts based on your findings. Produce and test content, then measure results against your baseline.
Remember, this isn’t about finding a magic formula—it’s about building a system that helps you consistently create content that resonates.
The Bottom Line
Creators and brands like Phia aren’t hoping for viral content—they’re engineering it systematically. They’re using AI not as a replacement for strategy, but as a tool to amplify strategic thinking.
This approach puts you back in control of your content performance. Instead of posting and praying, you’re creating with intention and improving with data.
The best part? This system gets stronger the more you use it. Every piece of content becomes a data point that improves your next creation.
If you’re ready to move beyond random content creation and start building a systematic approach to engagement, this framework gives you everything you need to get started.
Or, you know, you could just hire us instead. We’re pretty good at this whole “creating buzz-worthy content” thing, and we promise we won’t make you build any spreadsheets.