Write an outline, then stick it away where you can’t see it. The only helpful thing about an outline is the act of writing it, because it gets you to set down—and set aside—all of your bad ideas. The good ones are mostly subconscious at this stage.

Once you’ve ceremonially buried that outline, go with your gut. This is not mushy advice, but a technical practice. Storytelling is a complex process that your subconscious understands far better than your conscious brain. If you aim to make your entire writing process explicit rather than implicit, you will dumb yourself down and write boring, one-dimensional pablum. The act of writing bears almost no opportunity cost, so you have no reason to avoid taking risks and pursuing hunches. Try things out, rather than censoring yourself or sticking to a plan.

A book is different from an essay. An essay has a hook, perhaps a chorus. A book is a symphony, not a song: it includes movements, variations, strains, themes that repeat and build and sometimes conflict. It is more complex an entity than an author or reader can grasp from any single perspective or moment in time.

This is the reason most books made out of blog posts don’t sell: a collection of blog posts isn’t a book in this very important sense of being symphonic, the sum greater than its parts. A collection of essays can be a book, and a very good one, when the author has gone back, broken all those pretty, perfectly smooth eggs and made an omelette of them—something new and more complex, with a gestalt of its own. That is what Tobias Mayer did when turning his collected writings into The People’s Scrum—he came at it all afresh, reconsidering the forest, the trees, and the sounds they made.

You can read a book in a linear fashion, but you cannot write one that way. A book is written in layers, at different speeds and from different angles. You might charge through a first draft at break-neck speed, only to discover at the end that you’ve been driving backwards, or in a circle. Your next pass through the material might look more like a blower throwing up snow, or a flurry of licks of a cat’s sandy tongue—or a writer dying of a thousand paper cuts. It is okay to die; the world doesn’t need you to write a book.

Great books, like great works of music, aren’t written, but re-written—in iterations long and short, ranging from inspired all-night flights of hypomania to rock-breaking sessions in the quarry.

The best writers I know can all tell you that they often “don’t remember” writing their best work. That is because the trance that descends upon a writer in a state of advanced composition is unconscious, a fugue state. This sounds scary and exotic to non-writers, but it is basically the same thing that happens when you drive a car—when you are doing something so complex and so important with so many moving parts that the whole cannot exist within your consciousness—in fact your consciousness might leave the building entirely. Remember the last time you were so lost in thought you drove home rather than to the office? It’s the same thing.

Write for a reader—and I mean a specific reader. The novelist Edmund White has said that his ideal reader is a “sixty-year-old straight, European or European educated woman.” Or you can be even more specific: When I write fiction, my ideal reader is my friend Tom Christie. When I write or edit non-fiction, I usually aim for my mother, Ricki Grady—she is an author herself, but also a natural polymath and a great critical thinker, so if she can get excited about a book on an obscure technical topic she knows nothing about, then it’s a really, really good book.

If you are writing a book to satisfy a goal of your own, rather than a goal of your reader, then you are squandering your time and theirs. When you ask a total stranger to read a book you wrote, you are asking them to give you seven or eight hours of their thoughtful attention. That is a generous gift on their part, and if you don’t honor that as you write by putting that reader’s needs and desires first at all times, then you aren’t earning their readership.

Listen to your editor. As the creator, you are too close to the work to context-switch as rapidly between the big picture and the details as one must to achieve the level of harmonic refinement all great work shares. Nothing worth reading is unedited. Even professional editors need editing when they write.

But also: Don’t listen to anyone. That is, do not take any one critic’s words to heart—positive or negative. All mediocre work is praised. If everyone likes what you wrote, all this means is that you have written something mediocre. All great work is threatening to mediocre minds, and even to some great ones. If your work disturbs someone you admire, you’re onto something.

Non-fiction is no different whatsoever than fiction in these ways. The field of non-fiction suffers from writers and editors deeming it a programmatic endeavor, and this results in a huge number of utterly soul-less non-fiction titles on the market—many of them written by “subject matter experts” who want to have “a book out” to further a consulting career. Authors of this kind of book usually think the work needs to be stuffy and “professional” the way their website needs to be blue and white and talk about “business solutions.”

I have edited and published several books by experts that are by those standards extremely unprofessional—full of narrative, whimsy, irreverence, conviction and controversy. People tell me they enjoy reading those books. What writers of cookie-cutter non-fiction books do not realize is that writing with heart and personality is also easier than writing stiffly, to a plan. If you doubt me, try it.

Hillary Louise Johnson is the author of the novel, Physical Culture, and co-author with Chris Sims of the technical bestsellers The Elements of Scrum and Scrum: A Brief and Agile Introduction. She is the founder of Dymaxicon, which recently published Tobias Mayer’s The People’s Scrum and Gabrielle Chavela’s novel, Part of Your World.