For me, writing and design have often come down to a specific set of rules: define your thesis, then prove it.
(I was one of those jerks that loved writing literature papers by the way. My undergrad’s in English for a reason.)
The biggest problem, of course, is that you start by defining your thesis, go research it, and discover that hey, maybe your thesis is wrong. You refine your thesis, start writing, try to bring all that research into something that makes sense, and realize that nope, thesis isn’t quite right yet. That super-critical point that you believe with your heart to be true? Totally false. And you can try to fake it, but any good reader is going to see that you’re not actually backing your thesis adequately. And that’s going to erode your position, making the entire piece less effective. So you have to loop around again.
(I feel the same way about design. It’s great that the “thesis” of your webpage is “buy this shirt” but when you advertise seventeen other shirts, fifteen features of other products, and a lawn mower, as well as invite your reader to visit three other sites of yours and like you on Facebook, you’re eroding your thesis that “this is the shirt you want to buy”.)
My column on The Pastry Box shifted from the 16th of the month to the 3rd with the beginning of the new year, and I found that although I’d started a long and rambling piece on how different people come with different cognitive tools for dealing with the world, it has no functional thesis. Well, it has five function theses, which is a problem.
So instead, I described how to edit any piece down to one thesis.
Maybe it will help you. Heck, maybe it will help me.