A first attempt should be judged by what it reveals, not only by what it produces.

A first attempt produces two things: the object it makes and the knowledge of what the problem actually is. Judged only by the object, almost every beginning looks like a failure worth abandoning. The information it surfaces, which assumptions were wrong and where the difficulty really sits, is usually the more valuable half. Early versions are questions asked out loud.

After your next rough draft, write down three things you now know that you did not before starting. Decide whether to continue from that list rather than from how the thing looks. Keep the list where the next attempt can use it.