In my late 20s, I picked up a metaphor that I wish I'd come across earlier in my career. It was a metaphor for going about the journalistic process of reporting and writing, but it applies just as much to any writing practice. Here it is:
Writing is sculpting, first by addition and then by subtraction.
In other words, first you sculpt your piece by gathering all the raw materials you can think of. You jot down your thoughts, arguments, and assumptions. You gather your citations if you're writing about research. You go talk to some people if you're including outside perspectives and quotes. You pile all that up.
Then, after you've done that, you figure out how to combine those elements so that you select only the ones that truly serve the piece you're writing. You can't include everything, so you need to get rid of some stuff—an interesting-but-not-so-useful piece of data, a shaky quote, an errant thought. You won't know going in how much to subtract, but just know there should be a lot of it. More than you think.
After you've subtracted all the unnecessary bits, leaving only the elements that are essential, in theory you should have a lean, mean piece of writing. First you sculpt by addition, then you sculpt by subtraction.
Metaphors only get us so far, though. To add some concreteness, let me walk through the step-by-step process I go through for probably 80% of writing-based client projects. It's also how I wrote as a full-time reporter.
General outline
Detailed outline
Rough draft
Second draft
Final polish
To start, the general outline is the loosest assemblage of ideas. At this point, I haven't done much, if any, research yet, so whatever outline I create is necessarily going to change. That's okay. It's always been important to me to put shape to the sequence of ideas as soon as possible. Structure keeps me sane, even if I have to rebuild as I go.
Between Steps 1 and 2, I'm collecting more information and refining the flow of the piece. Maybe something further down in my general outline actually is more important than I first thought, so I need to move it higher up and move something else further down, or delete it altogether. By the end of Step 2, a detailed outline looks something like this:
Working Title
I. Intro
Opening story, insight - Line of detail
Context, "who cares?" - Line of detail, data point(s)
Additional detail teasing next section, rest of piece
II. Section title
...
This usually continues for five or so sections, but it always depends on the piece I'm writing. Around this stage, I'm also asking myself structural questions, like, What does the reader need to know as early as possible? What is my thesis? Which details support that thesis? Which ones are less relevant, but still helpful to include? How long should each section be in order to make the strongest case possible?
I'm exhaustive in Step 2 because the transition from outlining to writing is a big one, and I hate having to reverse course from composition (the fun part) to address something structural (the boring stuff). You can save yourself this headache if you remember that choosing what to say is always more important than choosing how to say it.
That said, drafting has to start sometime, so how about right now? Steps 3 through 5—the writing stages—are a back-and-forth process.
If you have an editor, a friend, or simply a pet who's willing to give you their attention, these steps are where you do the refining—the subtraction, most often—to reveal the sculpted piece within. If it's literally just you, it's still a back-and-forth, just with the "you" who will read the piece tomorrow, or next week, or next month, with fresh eyes. I always catch something I didn't notice before.
And if I ever feel I'm too familiar with the text, even after those grace periods, I'll change the document's font to something ridiculous, like Comic Sans, to see it anew; or I'll read it aloud to hear how it sounds. Or I'll do both, just for good measure.
Once you're happy with the final product, the last step, as always, is to surrender.
You must be willing to acknowledge that your writing will never be perfect. Even if it feels perfect now, it will always live under the microscope of future judgments, once life has had time to change your and others' perspectives.
If you don't acknowledge this, you'll cling to your writing as a precious object and inevitably be let down when it begins to tarnish. It sounds strange, but to paraphrase Chuck Klosterman, you must always write holding the simultaneous belief that what you're writing is both the most important thing in the universe and also has no significance whatsoever.
I've been writing for a long time, so trust me—there's no escaping this feeling. It's all part of the process.