Article handout · Dan Koe · X article · November 3, 2026
Dan Koe argues that capturing attention is the skill that decides whether any business makes money. AI does not replace that skill. It turns one person into a whole creative team. You can be that team for your own brand, or sell it to founders and creators. The article gives the three jobs on that team, a four-step method for building a brand, and a way to land the first client. This handout sums it up, applies it to building apps, and marks the sales pitch. Tap any ¶ chip to read the exact passage in the source panel.
If you read nothing else, read this. Each point links to the passages it comes from.
Attention decides who gets paid. A great product earns nothing if nobody knows about it. Attention is scarce and getting scarcer, and it now lives on social media, not TV or radio.
The money is moving to creators. Brand spend on creators tripled from 2021 to 2026, and founders are now expected to market their startups through their own posts.
The business is a one-person creative team. AI drafts and researches. A person still directs and rewrites. Do it for your own brand, or sell it to people who have no time for content.
Brand beats virality. Most AI content is a writer with no strategy. Add the brand strategist and the content strategist, and build a body of work that people explore over 6 to 12 months.
Method: interview, collect, map, then pitch with free work. Capture the person's story and beliefs, gather their best posts, have AI build a research-backed content map, and land clients by sending ten finished posts unasked.
Dan splits the work into three jobs. One person with AI can do all three, but most people only do the last one. That is why most AI content goes nowhere.
Owns positioning and perception. Decides what you stand for, and protects trust over reach.
Studies what performs: formats, hooks, titles, thumbnails, retention. Proposes the next posts. Without a brand, the growth does not stick.
Drafts the posts, scripts and newsletters, plans them and posts them, in line with the other two jobs. AI does most of the first draft now.
One real person: their bad day, their complaints in their exact words, what they already tried.
What you lived through that they live through now, and the moment things changed.
The people and works you never skip, and the ideas you keep coming back to.
What you could talk about for hours, the 2 or 3 that matter most, and the angle only you have.
What you want to help people do, and what you believe that others in your space do not.
The way of living that hurts your audience. A habit or idea, not a person.
What just changed that most people have not noticed.
Their life five years later, what they would call themselves, and your role: guide, documenter, or someone who came back with a map.
5 to 12 rules your reader lives by, and names for your own ideas and methods.
Dan writes for people who want to become creators or sell content services. You build consumer apps. These are the parts that carry over. The "For your setup" lines are our suggestions, not Dan's.
Dan offers two businesses: your own brand, or a content service for others. He also says a creator rarely builds software without a team, because the brand alone is a full-time job.
Do not add a content agency next to the apps. Use option one in service of what you already build: one brand, run as a one-person creative team, that brings users to your apps. The service business would split your focus in two.
Dan says good work earns nothing without attention. The Commanding Heights handout reached the same place from the other side: the durable asset is an audience you own and can reach directly.
Treat followers, email subscribers and push-enabled users as the asset, and each app as one way to grow them. Paid ads rent attention. A brand keeps it.
Growth without a brand is forgettable: if the account stops, nobody asks where it went. Most AI content is only the writer job.
Your AI character accounts and remakes cover the content strategist job well: find what performs, then recreate it. Run Dan's nine-part interview for each persona and for the app's brand. Then every remake also builds one recognisable point of view.
Steps 2 and 3 are a research job: pull the top posts across platforms, then have AI turn them and a brand document into a content map.
Your social-search skill already scrapes top TikTok, Instagram and Threads posts by account or keyword. Pair it with Dan's two prompts in Claude, and you have the content map without a new subscription.
Dan's first question asks you to describe one real person and quote their complaints in their exact words. Research on forums and app reviews gives you guesses, not those words.
Before you run the content map for an app, talk to three to five real users of that app. Use the same nine-part interview, and keep the recordings as the brand document.
Dan says the brand comes across over 6 to 12 months, and that most people fail because they want quick attention. He also warns against treating this as a way to get rich quick.
Set the success test for content before you start, and set it on a quarter or longer, not on a few weeks of views. Track leading signals weekly, such as saves and returning viewers, but do not judge the brand on them.
Our read, not Dan’s
The method is sound and specific. But the article is also a sales letter, and four claims are weaker than they sound.
Dan founded Eden, and almost every step sends you there: the research, the brand board, the creator lists, the MCP, two workshop ads. He says nothing else can do the job. The method itself needs only a scraper, a document and a chat model.
A growing creator economy does not mean a newcomer earns $10k to $75k a month. The proof is a few winners: JK Molina, and one reel that got a reply from Grant Cardone. Many people who send ten DMs a day never get an answer.
Dan says AI cannot copy a voice. His own agency quotes say AI already drafts posts that are almost there, and writers work 4 hours instead of 12. Agencies are "not firing people yet". If the drafts keep improving, a content service gets cheaper to replace. A brand you own does not.
The Permissionless Apprentice works because few people do it. Ten finished posts a day for strangers is a lot of unpaid work, and the more people copy the tactic, the less it stands out.
The names Dan uses for his ideas. Each chip opens the passage that explains it.
The figures Dan leans on, as he states them. The market numbers come from Goldman Sachs and the IAB; the income numbers are his estimates.
Creator economy, 2023 to 2027 (Goldman Sachs forecast)
Creator ad revenue in 2025, the first year it passed traditional media
US brand spend on creators, 2021 to 2026 (IAB)
An agency writer's day once AI drafts the posts
What Dan thinks a skilled one-person team can earn from 1 to 3 clients
A typical content client landed through the free-work DM
Dan's opening claim is that capturing attention decides whether any work earns money. Good software, a great book or hard work do not matter if nobody knows about them. As job security falls, many people will start a business, and most will do it through a personal brand. He says he will teach how to capture attention on social media, for your own business or as a paid service for others.
Dan argues that the quality of a product or book does not matter if nobody knows it exists. Capturing attention and getting customers is the main thing that decides success.
“If nobody knows who you are or what you offer, how in the world are they going to pay you?”
As job security falls, many people will choose to start their own business. Dan expects most of them to build it on a personal brand.
“the way they will build that business is by starting a personal brand.”
People can earn more money, but they cannot get back attention they spent. So whoever holds attention holds the scarce input.
“Because attention is a finite resource, and it’s only becoming more scarce.”
Part I uses market numbers to argue that attention has moved to creators. Brands and companies now pay creators large sums, and founders are told to market their startups through their own content. So the most valuable skill is making good content. But one viral post is not a brand: you need a body of work that people want to explore.
Dan cites IAB data on US brand spend on creators. He also cites Goldman Sachs: the creator economy grows from $250B in 2023 to $480B by 2027.
“US brands spent $13.9 billion on creators in 2021, and the IAB expects that to reach $43.9 billion in 2026, which is about 3x in 5 years.”
Dan pays creators himself. He says venture money pushes prices up, and most paid placements do not bring results.
“paying thousands of dollars for a single reel or 60-second YouTube time slot (that don’t even bring results 90% of the time).”
Startups grow through the founder's own personal brand. Dan says most founders are bad at it, which is the gap he points to.
““Founder-led marketing” is now a buzzword in the startup space”
A million views on one post means little if the rest get a thousand. Good content gets views. A brand makes people come back.
“That signals that your content is decent, but your brand is non-existent. We need both.”
Chasing virality on every post is why most people fail. Dan's alternative is a connected body of work that pulls people from one post to the next.
“or you can build a world of content that people can’t help but explore.”
Part II names the business: a one-person creative team, run with AI. You either do it for your own brand or sell it to founders and creators who have no time for content. Agencies say AI raises output, but a person still has to direct and rewrite. The team has three jobs: brand strategist, content strategist and writer. Most people only do the third.
Founders know AI can write. They hire someone because they have no time to think about content, and because direction still takes creativity.
“they still know they need a strategist or creative director who leverages AI better than they would.”
Agencies that use Eden say AI researches what performs and drafts posts that are almost there. A person still reviews and rewrites, but in a third of the time.
“My entire job nowadays is reviewing and rewriting drafts. It takes me 4 hours a day instead of 12 now.”
Option one: be your own creative team and grow your brand, list and products. Option two: sell that service to founders and creators. Dan says the first is a full-time job even with AI.
“In other words, you become the creative director of your own brand or someone else’s.”
Dan says a skilled operator can charge a fraction of what a full team costs. He also warns that this is not a way to get rich quick, and it needs practice.
“$10-$75k a month for 1-3 clients isn’t that far out of reach”
Job 1: Brand strategist
The brand strategist owns positioning and perception. They push back on following every trend just to go viral.
“Growth and engagement are important, but not at the expense of trust and authority.”
Job 2: Content strategist
The content strategist studies what performs: formats, hooks, thumbnails, retention. They grow accounts, but growth alone builds no loyalty.
“if the creator were to stop posting, nobody would ask where they went.”
Job 3: Writer and social manager
The writer drafts, plans and posts. Most people only do this job, so they turn random ideas into posts and hope. Dan says that makes them easy to beat.
“They think that if they just have Claude pump out scripts without direction or data they’re going to magically build a connection with their audience.”
Part III is the method, written as if Dan were onboarding you as a client. First, a long interview builds a brand document. Then you collect the person's best posts, or the best ideas of the people who shaped them. Last, AI combines both with research on what performs to make a content map: about six months of post ideas.
The method in four steps
Ask the nine groups of questions. Save the transcript as a brand document.
Have AI pull the top posts and note the ideas, voice and worldview behind them.
List the 5 to 10 people who shaped you most, and pull their best ideas instead.
AI merges both documents with what performed in the past 6 to 12 months: pillars, topics, ideas, angles.
Step 1: Onboarding call
The onboarding call collects the context that makes the brand unique. Save the transcript so AI can use it later.
“Record the call so you’re able to store the transcript somewhere and reference it with AI.”
The interview answers become topics you post about for months. You cannot fit them into one post. Most people quit because they want quick attention.
“You don’t. This is a long term game, and is why most people fail miserably.”
Who you help, your story, who shaped your mind, content pillars, mission and beliefs, what you are against, why now, where you take people, and your rules and words. See the full list in the reader.
“Who do you most want to help? Describe one real person.”
Step 2: Scrape their best content
Dan agrees AI cannot copy a personal voice. The aim is context, so AI drafts and ideas start from what is unique about the person.
“you’re building a knowledge base of everything that is unique or noteworthy about that person.”
Have AI collect the best posts on each platform. Then note the core ideas, the voice and the worldview behind them, in one document.
“10-20 posts from each platform they are on.”
Step 2.5: If you have not started posting
Your mind is a mix of the people who shaped you. List the 5 to 10 who affected you most, and pull their best ideas into a document.
“Your best ideas - the ones that will lead to your growth - are how you interpret the ideas you’ve adopted from others.”
Step 3: Research-backed content map
Many creatives dislike studying viral posts. Dan reframes it as an artist's reference board: take the patterns, not the copy.
“You’re collecting references.”
Feed the brand and content documents to AI. Ask for 3 to 4 pillars, 5 to 7 topics each, 5 to 10 ideas per topic and 4 to 5 angles per idea, checked against what performed over the past 6 to 12 months.
“3-4 unique and broad content pillars that branch into 5-7 topics”
AI can then draft each idea using the saved context. You check each post's structure against what performs.
“Now you have about 6 months of content ideas that you can work with.”
Part IV is how to get the first client without cold outreach. Do the work for free first: write ten posts for a busy founder or creator and send them, ten DMs a day. If DMs fail, pitch in public. Dan calls this the Permissionless Apprentice approach.
JK Molina, who started the X ghostwriting trend, wrote ten posts for Dan without being asked. Dan saw it as a better way to get clients than cold outreach.
“I wrote 10 posts for you anyway. Try posting them, if you like them, then let me write more, or just give me a testimonial.”
Run the whole process for a target client, draft ten posts, polish them and send them. Dan expects at least one client a month at $2,500 to $5,000.
“have it draft 10 posts, polish them up, and send 10 DMs a day.”
If DMs fail, pitch in public. A creator posted a reel called “Grant Cardone needs a new creative director”, and Cardone replied “Hit me I’m stuck”.
“You do free work for a dream client, then get it in front of them however you can.”