Things I wish I knew about grant writing when I started
1. Submit the same idea (and same grant if possible) to as many agencies as you can to increase chance of funding.
2. If an idea gets rejected but gets decent reviews, don't change the whole idea just fix the things that the reviewers point out.
3. Don't write to what you think will get funded, write to what makes sense scientifically.
4. Do think about the goals of the agency though and try to match your tone and presentation to what is expected.
5. It's probably better to write many grants about few ideas than to write many grants about many ideas.
6. Do your homework. You'd be surprised how many times proposed research has been published by others before the grant even gets submitted.
7. Don't confine yourself to agencies/foundations that you are familiar with. Explore areas outside your "home department" but that are related to your work.
8. Keep it simple stupid. Propose the simplest part of your experiments that will demonstrate proof of concept rather than trying to build a complicated house of cards all dependent on success of aim 1 even if it is flashy and exciting.