Now that we have one funded proposal, I'm not going to act as though I have this stuff all figured out. I don't. There are lots of bloggers with better funding records and more insight into the process than I. But there are some things I learned along the way that might be helpful to those in the process of chasing that first proposal.
1) Have a couple irons in the fire. The proposal we ended up getting funded first was not the first or even second proposal I started sending out. We have several things going on in the lab and this just happened to be the first thing that got picked up. Sometimes I wonder if we riding the line between "project diversity" and "spread too thin" a little too hard, but in this case it paid off.
2) Get to know your agency. This might seem obvious, but if you are targeting NSF, get on a panel. Contact the relevant POs and put your name in the hat. They always need qualified people, so you will get called. NIH is less likely to call on unfunded peeps, but it's worth a phone call.
3) You're going to get a lot of rejections, so get used to it. CPP made this point recently, but it bears repeating: you need to keep swinging if you want to get a hit. Funding rates are down and there is a certain level of stochasticity to the process, where even a very good proposal is going to get beat up once in a while. Fix it up and get it back in.
4) Program officers are there to help. Even if they can't tel you things outright, more often than not you can learn a lot in a conversation with them.
5) Data. Especially right now, show you can do what you are claiming you want to do. If you have a new method, demonstrate it can work. Use a small, preliminary or simulated dataset, but assuage reviewer concern about the method, because that is the easiest way for a reviewer to tank your proposal.
6) Get the message across. Again, should be obvious, but a well written grant makes a HUGE difference. Use subheadings to guide the reviewers along and allow for easy back referencing. Make the important points clear. Provide figures to break up the text and make your point. The reviewers that are most critical to your case are going to be the ones who have to read 10-15 (or more) proposals in preparation for the panel. Make their lives easier and you're ahead of the game.
7) Have a solid Broader Impacts section. Obviously this is NSF-specific, but don't ignore the BI section and give reviewers any easy negative thing to say. It's really not hard to come up with something if you take advantage of existing programs. This section also tends to be the last thing a reviewer reads, so don't leave them on a down note.
I'm sure there are important points I am leaving out, but that's what I can think of off the top of my head. Above all, keep making improvements and resubmitting. Don't wait for the response to continue working on the project and making progress with the data, because you may need that critical piece for the next submission.