Derek Sivers' Advice on Writing: No Wasted Words

Derek Sivers founded CD Baby, which became the largest online seller of independent music. He sold it and gave the proceeds to charity. Then he wrote Anything You Want, Hell Yeah or No, How to Live, and Useful Not True. His books are deliberately short. How to Live started as a 1,300-page rough draft. He spent thousands of hours editing it down to 112 pages. Not a single unnecessary word survived.

His writing advice comes from David Perell's How I Write podcast, his essays at sive.rs, and the Writers Ink podcast.

The Seven-Step Process

Sivers published his writing process on his site. It is seven steps, and the last one is the point:

1. Write all of your thoughts on a subject.
2. Argue against those ideas.
3. Explore different angles until you are sick of it.
4. Leave it for a few days or years, then repeat those steps.
5. Hate how messy these thoughts have become.
6. Reduce them to a tiny outline of the key points.
7. Post the outline. Trash the rest.

Step seven is where most writers flinch. You have spent days, weeks, or years exploring a topic. You have pages of arguments, counterarguments, tangents, and discoveries. Sivers posts the outline and throws everything else away.

This is not minimalism as aesthetic preference. It is a conviction that the reader's time is more valuable than the writer's effort. The exploration was necessary for the writer to reach clarity. The reader does not need to see the journey. They need the destination.

One Sentence Per Line

Sivers has used this technique for twenty years. He says it improved his writing more than anything else.

The method: write each sentence on its own line. Not for publishing. For drafting. Your eyes only.

Four things happen. First, weak sentences become obvious. "Standing on their own, we notice. Delete any sentence not worthy of its own line." A bad sentence can hide in a paragraph. It cannot hide alone.

Second, you see sentence length variation. Short sentences and long sentences stacked vertically reveal rhythm patterns that paragraphs obscure. If every sentence is the same length, the writing drones. The vertical format makes that visible.

Third, rearranging becomes easy. Sentences in a column can be moved like items in a list. You find the strongest sentence and put it where it belongs - often first or last.

Fourth, you notice beginnings and endings. The first word and last word of each sentence carry disproportionate weight. The vertical format exposes weak openers like "I think" or "Whether or not" and lets you replace them with strong subjects and verbs.

Compressing Knowledge into Directives

"Compressing wisdom into directives - 'Do this.' - is so valuable, but so rarely done."

Sivers noticed a pattern when sharing book insights with friends. They would listen to his explanation of a 300-page book and say: "Just tell me what to do." Not because they were lazy. Because directives are useful. Explanations are interesting. Directives change behavior.

He points to Michael Pollan's Food Rules as the ideal. Pollan wrote two dense books about food and nutrition, then compressed the essential wisdom into 64 sentences readable in 30 minutes. That slim book probably changed more diets than the long ones.

Sivers acknowledges that telling people what to do seems arrogant. His reframe: "It's useful to people, so do it." Nobody is obligated to follow a directive. But a directive is easier to act on than a nuanced essay. The nuanced essay is what the writer needs. The directive is what the reader needs.

1,300 Pages to 112

How to Live is 27 short chapters. Each chapter argues for a different way to live. Each disagrees with the next. The structure is the point: there is no single answer, only contradictory answers that are each partially true.

The first draft was over 1,300 pages. Sivers wrote everything he had learned from 400-plus books and 50-plus years of living. Then he cut. And cut. And cut. Thousands of hours of editing. The final book reads like poetry because every word that was not earning its place got removed.

This is the extreme version of step seven in his process. Write everything. Then publish only the essential fraction. The ratio for How to Live was roughly 12 to 1. Twelve pages of thinking for every page of output.

Test Ideas in Conversation First

Sivers describes his best ideas as emerging from two-way banter. A friend says something that sparks a reaction. He responds. They push back. The idea sharpens through friction.

His most famous concept, "Hell Yeah or No," came from a phone call with musician Amber Rubarth. He was rambling about not wanting to attend a conference. She crystallized it: "So what you're trying to say is it's not a decision between yes and no, it's a decision between fuck yeah or no." He laughed, wrote the article that night, and the concept became a book title.

He connects this to John O'Donohue's definition of a great conversation: "When you overheard yourself saying things you never knew you knew." The comedian's process - testing material live, keeping what lands, discarding what does not - applies directly to writing. Ideas that survive conversation are ideas worth publishing.

Write Daily, Publish When Ready

Sivers tried publishing daily. It failed. Quality suffered because he was "spending more time being shallow, to get something posted." Rushed articles violated his implicit agreement with readers that everything on his site deserves their time.

He returned to a middle path: write every day, but publish only when an idea merits it. "I only post something to the public when I feel it's really worth sharing." Writing productivity and publishing frequency are separate concerns. You can be prolific in private and selective in public. The daily practice builds skill. The selective publishing builds trust.

His single biggest advice for aspiring book writers: "Don't think of it as a book. Put one idea at a time out into the world." Shine a spotlight on each idea. Let the world give you feedback. Test them. Then assemble the book with confidence that the ideas are ready.

Plain Text, Ancient Tools

Sivers writes in plain text files on a machine running OpenBSD with a window manager called Rat Poison - chosen because "it's the only one I've ever found that removes every single pixel from the screen except what you're working on." No menu bar, no icons, no clock. His friend Amber looked over his shoulder once and said, "I have never seen somebody use a computer like that."

The simplicity of the tool matches the simplicity of the output. When the tool offers nothing but words, the words have to be good. Even his book covers reflect this: he tested thirty-six designs with twelve graphic designers, then stripped everything to just the title and author on a plain hardcover. "I love it when you find a hardcover book and you take off that glossy, shiny outer wrapping, and what you have underneath looks like a library book."

Too Many Words Hide the Message

This is the throughline of everything Sivers writes about writing. Excess words are not neutral. They actively obscure. A 300-word blog post that makes one point clearly is more valuable than a 3,000-word post that makes the same point while wandering through qualifications, anecdotes, and tangents the writer found interesting but the reader did not need.

Sivers does not argue against long writing. He argues against unnecessary length. Some ideas need 500 pages. Most do not. The writer's job is to find the natural length of the idea and refuse to pad it.

His books embody this. Anything You Want is 88 pages. Hell Yeah or No is 118 pages. Useful Not True is 136 pages. Each could have been longer. None needed to be.

Key Takeaways

• Write all your thoughts. Then post the outline and trash the rest.
• Write one sentence per line during drafting. It exposes weak sentences, reveals rhythm, and makes rearranging easy.
• Compress knowledge into directives. "Just tell me what to do" is what the reader actually wants.
• Write daily. Publish only when the idea is worth sharing. These are different activities.
• Too many words hide the message. Find the natural length of the idea and stop.
• Cut aggressively. Twelve pages of thinking per page of output is not excessive.

Sivers and Morgan Housel share an obsession with saying complex things simply. Housel takes financial concepts and strips them to stories. Sivers takes life philosophy and strips it to directives. Both believe the reader should never work harder than necessary. The writer does the work. The reader gets the reward.

This post draws from Sivers's essays at sive.rs and the Writers Ink podcast. Athens is an AI writing editor that shows every change as a tracked diff - so your compression from 1,300 pages to 112 happens without losing the sentences that earned their place.