Here’s some great tips for writing scientific papers from Steve Ellner (link to original pdf). Ellner presents them as tips for theoretical papers, but I think they’re good practice for any paper:

1. Don’t maintain suspense.

• Present the topic clearly at the very beginning.
• Explain the relevance of the paper at the very beginning.
• Quickly telegraph where the entire paper will be going. Give away all your punchlines in the abstract, and do it again in the Introduction.

2. Make the paper easy to skim.

• Make sure that the “meat”—the core that everyone should read—is well labeled and easy to find.
• Explain your main results using graphs.
• Remove from the main text any technical details that aren’t needed for the flow of ideas. Readers shouldn’t have to stop and think about whether or not they have to think about an equation.
• Use signposting to help people “peel the onion”—get as deep into the paper as they want, but no deeper. Technical sections should be prefaced by an explanation of what and who it’s for, so it’s easy for a reader to tell if they should read it, skim it, or skip it for now.