Dan Koe reveals his AI-powered content workflow, leveraging Large Language Models (LLMs) and strategic repurposing across platforms to generate viral ideas and maintain high output with minimal effort.
Takeways• Leverage LLMs to analyze and deconstruct successful content, not just for basic summaries or initial drafts.
• Repurpose content strategically across platforms, starting with a core piece and adapting it for different mediums to maximize reach.
• Develop a systematic content workflow, using validated ideas from a 'swipe file' and continuous experimentation to drive audience growth.
Dan Koe outlines a systematic, AI-augmented content creation process designed for maximum reach and virality. His approach prioritizes a weekly newsletter and daily social media posts, originating ideas from validated high-performing content or industry trends, then expanding and adapting them for various platforms. By understanding human psychology and utilizing LLMs for idea generation and content deconstruction, this method enables consistent, high-quality output and scalable audience growth.
Content System Overview
• 00:02:41 Dan Koe's content ecosystem is built on a weekly newsletter and two to three daily social media posts. He primarily writes for X (Twitter) first due to its character limit, then repurposes this validated content across platforms like Threads, Instagram, LinkedIn, and short-form video formats. This strategy ensures content consistency and efficiency, believing that algorithms favor human psychology and that a strong initial idea translates well across different mediums.
Idea Generation and Validation
• 00:04:40 Newsletter ideas are generated from top-performing tweets or popular YouTube videos within a niche. The process involves identifying successful content topics from others, then crafting original perspectives rather than copying. Dan Koe maintains a 'swipe file' of high-performing tweets from other accounts, studying their structure and ideas to refine his own writing style and brand, emphasizing synthesis over direct imitation for unique content.
AI-Powered Content Expansion
• 00:07:58 LLMs like Claude and ChatGPT are central to research and content expansion. Dan Koe uses them to summarize long videos or PDFs, extract key points, and ideate from existing content like newsletters. He employs specialized prompts to generate YouTube titles based on his best-performing ones and to deconstruct high-performing social posts, identifying psychological patterns and structural elements. This helps him gather 'building blocks' for new content ideas rather than having the AI write full posts.
Advanced Prompt Engineering for Growth
• 00:32:35 Dan Koe utilizes a two-step process to create sophisticated prompts for content creation. First, he asks an LLM to break down successful content (tweets, landing pages, book paragraphs) into its core structure, psychological patterns, and underlying principles. Second, he uses a 'prompt to create prompts' to transform these insights into interview-style questions for context gathering, followed by instructions for generating new content (e.g., tweet variations or offer blueprints) that align with the learned successful structures and his specific brand voice.