ps: Don't use ChatGPT for layout and optimization, it's really hard to use...
Many people overestimate the importance of information but underestimate the importance of structure.
In today's era, there is never a shortage of information. What is missing is: Whether you can organize scattered information into a complete industry cognitive system.
When many people enter a new industry, it looks like this:
- Google search
- Baidu search
- Check Zhihu
- Read WeChat Official Accounts
- Read a few industry reports
After three days of looking, they have 200 more links in their bookmarks, but their brains are still a mess. They don't know who the top players are, who is making money, what the supply chain looks like, or where the opportunities lie.
This used to take a researcher several weeks. But now, you can completely let Codex become your own industry researcher. When I enter a new industry now, I basically do the following actions first.
Prepare two software programs in advance:
1. Codex
2. Obsidian
# Step 1: Let Codex build an industry database
The way most people use AI is: Ask a question, get an answer, and then it's over. For example: \"How is the US weight loss supplement industry?\" or \"Help me search for the market situation of US weight loss supplements.\" AI gives you an analysis of several thousand words. You feel great after reading it, but three days later, you've basically forgotten it all. Because what you got was just information, and we only used our brains to memorize it instead of storing it.
I now prefer to let Codex help me build an industry database. For example: I want to study the US weight loss supplement industry. The first step is not to study products, but to build an industry structure. I will tell Codex directly: Help me build a US weight loss supplement industry database. Output a complete directory structure and create Markdown files for each directory. All content should be suitable for importing into Obsidian.
Example:
Weight-Loss-Supplement-Industry
│
├── Brands
├── Products
├── Keywords
├── Communities
├── Influencers
├── Competitors
├── Business-Models
├── Supply-Chain
├── Regulations
├── Trends
└── Opportunities
At this point, you are creating a large information retrieval directory for the industry database.
## Brand Database
Let Codex compile: Organize the Top 100 brands in the US weight loss supplement industry. Output: Brand name, Official website, Main products, Price range, Sales channels, Estimated scale, Core selling points, Founder background, Social media accounts. Eventually, you get a Brands folder with files like Goli.md, HUM.md, etc. Whenever you encounter a new brand, just put it in. The database grows larger, providing the data support for your research.
## Product Database
Next, let Codex compile the main product types in the US weight loss supplement industry, sorted by market size. You can add a timeframe like Q1 2026. Then break it down further: Fat Burners, Appetite Suppressants, GLP-1 Support, etc. For each, analyze ingredients, user reviews, pros/cons, viral brands, and market size. This is when the AI starts to understand the industry and user pain points.
## User Pain Point Database
This part is often the most valuable because the money is hidden in the pain points. Let Codex organize Reddit discussions about weight loss supplements, focusing on high-frequency complaints, needs, questions, and goals. This forms a Pain Points database (e.g., rebound, no effect, side effects, high price). Many product opportunities are right here.
## Content Database + Traffic Acquisition Channels
Most people only look at products, but content often determines traffic. Let Codex compile top channels on YouTube, TikTok, Instagram, X, and Newsletters. Analyze the content with the highest views, likes, comments, and shares. You will find that certain topics and viewpoints go viral repeatedly. This tells us what kind of content is validated by the market.
## Keyword Database
This is often overlooked. Let Codex organize keywords from Google, Amazon, Reddit, YouTube, and TikTok, categorized by intent (Commercial, Informational, Comparison, Review, Buying Intent). This can be directly used for SEO, ads, and content creation.
## Final Result
Instead of dozens of random articles, you get a complete Industry Operating System in Obsidian. You see a complete cognitive system rather than scattered info.
# Step 2: Let Codex reverse-engineer how the industry makes money
The fastest way to learn is to deconstruct your peers. They have already paid the tuition for you. You just need to study why they make money.
## Build a Competitor Database
If you are doing Shopify, identify the top brands and give the URLs to Codex. Ask it to analyze the site structure: navigation, product categories, collections, tags, footer, blog, SEO, and landing pages. Organize this into a Markdown report.
## Deconstruct the Navigation Bar
The navigation bar is the boss's brain. It determines the user's journey. It tells you what the profit products, traffic products, and conversion products are.
## Deconstruct Collections
Collections are the actual transaction paths. By analyzing multiple competitors, you'll find their structures are very similar because they are all doing what has already been validated.
## Deconstruct Product Tags
Tags tell you how users search for products, how Google understands them, and how the recommendation system works. Analyzing 1,000 products gives you the industry's entire tag system.
## Deconstruct SEO and Blog Content
Don't write articles blindly. Study what the top players are writing. Identify the high-traffic keywords and the content they use for customer acquisition. Use Codex to find the most viewed and frequently updated blogs.
## Deconstruct Social Media
Analyze which content types (exposure, growth, save, conversion, persona) perform best on social platforms. You'll move from seeing screenshots to understanding the entire industry's monetization logic.
# Step 3: Research the Industry Content Ecosystem
In today's era, content is a means of production because traffic determines orders. Don't just follow a few influencers; let Codex build a database of 100 accounts across all platforms. Identify the \"traffic harvesters\" by analyzing the top-performing content over the last 90 days. You'll find that viral hits aren't random; they follow patterns. Categorize content into Exposure (controversial), Growth (resource-based), Save (SOPs/templates), Conversion (results/cases), and Persona (stories). This turns content from entertainment into a growth system.
# Step 4: Build an Industry Knowledge Map
Information overload is the enemy. A knowledge map solves this by giving every piece of info a place. Let Codex create a multi-level map of the industry (e.g., AI Industry -> Models, Applications, Infra -> LLM, Multimodal, etc.). Create knowledge cards for each node. This transforms your knowledge from a tree structure into a network, allowing you to see opportunity maps where competition is low or growth is high.
# Step 5: Make it Your Private Knowledge
Industries change. To stay relevant, switch from a search mode to a subscription mode. Build an information source system using RSS to monitor top blogs, competitors, and content trends. Let Codex generate weekly intelligence reports and automatically update your industry map. You are no longer just a consumer; you are an industry researcher with a continuously growing cognitive asset. Codex is not just a coding tool; it's a researcher that never gets off work.
