Start it badly

Here is the instruction. On the day a piece of work that matters arrives, spend fifteen minutes making a bad version of the whole thing. If it arrived a while ago, do it today.

Each part of that is doing something, so I'll take them in turn.

Bad, on purpose. I don't mean rough, with a secret hope that it turns out good. I mean bad. Clumsy sentences, guesses where facts should go, and SOMETHING ABOUT COSTS HERE in capitals. The reason is in chapter 5. What you've been avoiding is judgement, and a version that's meant to be bad can't be judged a failure. If it's bad, you've succeeded. For once the judge has nothing to say, and you didn't need one o'clock in the morning to arrange it.

The whole thing. Don't write a polished first paragraph. Write a bad version of all of it, start to finish, at whatever level of detail fifteen minutes allows. For a proposal, that might be every heading with one ugly sentence under each. For a talk, it's the six things you might say, in the order you might say them. For a spreadsheet model, it's the tabs and columns with invented numbers in them. For a portfolio, it's a list of eight pieces with a line on why each is there. Covering the whole thing shows you the size and shape of the job, which is the information you've been missing. It also stops you hiding in the one section you find easy.

The real thing. The bad start happens in the document itself, the one that will eventually be sent. Reading doesn't count. A plan of when you'll work doesn't count. Notes about what the proposal might cover don't count either. Those can come afterwards. The test is simple. When the fifteen minutes are up, is there more of the actual work in existence than there was?

Fifteen minutes, with a timer. It's short enough that you can't reasonably claim you haven't the time, and short enough that the swerve has little to push against. Dreading twelve days of work is easy. Dreading a quarter of an hour of deliberately poor typing is hard. When the timer goes, you can stop, and I mean that. If you want to carry on, carry on. But the deal you make with yourself has to be honest, or you won't believe it next time.

On the day it arrives. That's when the work is least frightening. No guilt has built up, nobody expects anything yet, and a bad version on day one looks like keenness. Every day of delay adds to what the first session has to overcome.

There's one more step. Before you close the document, write a line at the top saying what the next thing to do is, and make it specific. "Ask finance for last year's figures." "Write the second aim properly." Chapter 12 explains why that line is worth more than it looks.

A decorator's first coat looks dreadful. It's patchy and the old colour shows through. No decorator is troubled by this, because a first coat isn't for looking at. It's there so the second coat has something to go on.

Why would fifteen bad minutes change anything? This part is my reasoning, built on the earlier chapters.

It gives you on day one what the last night gives you. The judge is quiet because bad is the goal. The job is small because the timer made it small. The choice is gone because the rule says today. You get all of that without the fear.

It turns dread into a list. The gaps now have names, and a named gap is something you can send an email about.

It starts the clock on everything that takes time without taking effort. That covers the questions other people need days to answer, the incubation from chapter 6, and the chance to show someone a draft.

And it changes what the work is. Before, the task was to write the proposal, which meant making something from nothing in front of a judge. Now the task is to improve a bad proposal, which is editing. Most people find fixing a bad page far easier than filling a blank one. Your complaints about the bad version will be specific, and a specific complaint is halfway to an instruction.

I should be plain about the status of all this. The parts are supported by the research in the earlier chapters. The package, fifteen bad minutes on day one, hasn't been tested in a trial that I know of. It's my recommendation, and the evidence that counts will be your own.

Some objections come up reliably. The first is that a bad draft is wasted work, because you'll only throw it away. You'll throw some of it away. Count what you've been spending instead, though. Twelve days of dread and a night of lost sleep are a higher price than fifteen minutes of sentences you later delete. And less gets thrown away than you'd expect. The structure tends to survive even when no sentence does.

The second is embarrassment. Nobody sees it. If it still bothers you, type DRAFT ZERO at the top.

The third is quieter. What if the bad version shows that you've got nothing to say? Then you've learned it with eleven days left and not eleven hours, while there's still time to read, ask or rethink. That question is the fear from chapter 5 talking, and I'd bet on the opposite result. People who've been thinking about a subject for weeks usually find more on the page than they expected.

If you have anything waiting, try this once before you read on. Reading about it won't tell you much, and one go will.