You can now reverse-engineer your entire personal brand strategy in under three minutes using just a handful of screenshots and a single, intelligent prompt. It sounds impossible to replicate the work of a high-end consultant that quickly, but it is actually just a clever application of data analysis. I just saw this incredible post from an AI professional who demonstrated exactly how to crack the code on self-analysis. Instead of guessing what works or paying for an expensive audit, the expert uploaded 50 past posts to ChatGPT and let the AI find the hidden patterns in the data.

Most people fly blind when they create content, relying on gut feelings about what their audience likes, but this method turns that process into a science. By feeding the AI historical data, both your biggest wins and your hardest flops, you force it to look for objective correlations that a human might miss. The original poster created a specific workflow that essentially turns ChatGPT into a ruthless editor and growth analyst. The results allow you to stop wasting time on content formats that do not convert and double down on the specific structures that drive engagement for your unique voice.

📌 The Mechanism: Visual Data and Reasoning Models

The core innovation here isn’t just asking ChatGPT for advice; it is the method of data ingestion the creator used. Rather than copy-pasting messy text files, the author suggests taking screenshots of your last 20 to 50 LinkedIn posts. This is brilliant because screenshots capture more than just words; they capture the formatting, the line breaks, the use of emojis, and the visual “hook” of the post.

When you upload these images to a model with vision capabilities and “Thinking” mode (like the reasoning models found in ChatGPT), the AI performs Optical Character Recognition (OCR) while simultaneously analyzing the visual structure. It can see if short paragraphs performed better than long blocks of text. It can analyze whether posts with images outperformed text-only updates. By combining this visual data with the metrics visible in the screenshot (likes, comments, reposts), the AI builds a comprehensive dataset of your performance history.

💡 Insight 1: The “Ruthless Diagnosis” Phase

The first part of the strategy involves a deep audit that prioritizes blunt feedback over compliments. The prompt designed by this industry pro specifically instructs the AI to ignore politeness and identify “rapid patterns.” It asks for a direct comparison between winners and losers to find the common threads. For example, it might discover that every time you start a post with a question, it fails, but every time you start with a controversial statement, it goes viral.

This phase also includes a “Keep/Change/Remove” verdict for each post. This is incredibly practical because it moves beyond general advice. The AI might tell you to keep your storytelling style but remove the three hashtags at the end because they clutter the hook. It looks for the “biggest lever” to improve—the one change that would have made a mediocre post a great one. This level of granularity effectively simulates a session with a human content strategist who has studied your account for weeks.

💡 Insight 2: Extracting Your “Winner DNA”

Once the diagnosis is complete, the workflow moves to defining your unique “Winner DNA.” This is a concept the LinkedIn user introduced to describe the repeatable ingredients of your success. The AI is tasked with creating a short checklist of max eight items that are present in your top-performing content. This acts as a quality control filter for all your future writing.

This checklist might include specific structural elements, such as “always use a 3-line hook,” or thematic elements, like “focus on personal failure stories.” By codifying these elements, you stop relying on inspiration and start relying on a framework. It transforms the vague goal of “writing good posts” into a tangible production process where you simply ensure your draft ticks the boxes of your identified DNA. This creates consistency, which is the hardest part of building a personal brand.

💡 Insight 3: The Execution and Repurposing Plan

The final piece of this strategy is turning analysis into immediate action. The prompt instructs the AI to generate a “Double-down” plan, offering six new post concepts written in your voice, complete with hook options and outlines. This solves the “blank page” problem instantly. Because the AI has analyzed your previous successful tone, the new ideas it generates are likely to feel authentic rather than robotic.

Furthermore, the expert included a section on repurposing. The prompt asks the AI to take your top two winners and generate new angles for them—turning a story into a framework, or a contrarian take into a case study. This maximizes the ROI of your best ideas. Instead of letting a viral post die after 24 hours, you spin it into three or four new pieces of content that approach the same core value proposition from different directions, keeping your feed active with high-quality material.

The Prompt

Here is the exact prompt the creator shared to run this analysis. Copy this into ChatGPT (ensure you have uploaded your screenshots first):

“Act like a world-class LinkedIn content strategist and growth analyst.

Objective: Turn the user’s uploaded “best” and “worst” LinkedIn posts (text/screenshots + any metrics) into a repeatable, personalized playbook. Keep the work lightweight by operating in phases, not all at once.

How to work:

• If metrics are provided, use them. If not, infer cautiously and label assumptions.
• Quote or reference exact lines/sections when you explain performance.
• Be blunt and specific (no generic advice).

PHASE 1 — Diagnosis (deliver this first, then stop and ask questions)

1. Rapid patterns (max 7 bullets): what winners share, what losers share, 3 do-more, 3 stop.
2. Post audit (for each post, max 10 lines):
   • Winner/Loser verdict + 2–4 reasons tied to specific lines
   • Keep / Change / Remove (bullets)
   • One “biggest lever” to improve it
3. Winner DNA: a short checklist of repeatable ingredients (max 8 items).

PHASE 2 — Execution plan (only after the user answers your questions)

1. Double-down: 6 next-post concepts in the user’s voice; each includes 2 hook options + a 5-bullet outline + CTA.
2. Repurpose: pick the top 2 winners and generate 2 repurpose angles each (choose: story, contrarian take, framework, case study) with mini-outlines.

Collaboration (end PHASE 1 with questions):

Ask 6 targeted questions to tailor PHASE 2 (ideal audience, goal, offer/CTA, voice boundaries, proof assets, cadence). Then propose 2 A/B tests and ask the user to pick one to run next.

Take a deep breath and work on this problem step-by-step.”

✅ Nuances and Potential Challenges

While this method is powerful, there are a few nuances to consider for the best results. First, the quality of the output depends entirely on the contrast in your data. If you only upload medium-performing posts, the AI will struggle to identify strong patterns. You need to feed it distinct “winners” and clear “losers” so it can detect the variables that cause the difference.

Additionally, be mindful of privacy when uploading screenshots. If your posts contain sensitive client data or personal information in the comments section (which might be captured in the screenshot), blur those out before uploading. Finally, remember that AI is a tool for analysis, not a replacement for your intuition. Use the “Winner DNA” as a guide, not a rigid law, and always review the generated copy to ensure it sounds like you, not a machine mimicking you.

If you want to see the full guide and the original breakdown, make sure to visit the LinkedIn post linked below!