If there's one universal feature that unites anyone running a research lab, it's that there are never enough hours in the day for them to get done everything that needs doing. From manuscripts to meeting, reviews to reports, there is always something due yesterday. The general hierarchy of How Fast Shit Gets Done, runs something like this:
Will it bring money into the lab?
Is it for a lab publication?
Is someone seriously pissed about the lateness of it?
Was it due yesterday?
Is it due today?
Are there meaningful consequences for missing the deadline?
Is it for my trainees?
Note that things from "everything else" can work their way up the ladder, but only if someone starts making some noise. I used to feel like an ass when I would get those reminders from journals that my review is due today. Now, depending on the time of year, that reminder is what gets me to download the damn thing.
Why is this important? Because outside of a few incredibly efficient PIs, most of us need reminders. In fact, for some tasks, if I don't get a reminder it will fall off the deck completely. This often comes up in the bloggosphere and IRL as an issue that a lot of trainees have, and the simple answer is that you need to stay on top of people if you want to keep the thing you need from them on the forefront. Always be pleasant and polite, but sending reminders is something you'll need to get used to. Remember that letting someone else's work sit on the back burner is an easy way to manage one's limited time, and unless there's some sense of urgency from that other person, it's gonna sit until I have nothing more pressing to do.