The best post so far is the only one which has the same title as the column itself (link). Key advice:
1) Just pick three things. Don't make a list of everything you plan to do each day - that way lies failure. Instead, choose the three most important things you'd like to get done, preferably including one that's meaningful-but-not-urgent (or you risk spending the whole day putting out fires, which is fulfilling only if you're a firefighter). On a good day, you'll do plenty more, but you get to count as a success any day that you do your three.
2) Do the least enjoyable task first. Otherwise known as the eat-the-frog principle: if you eat a live frog each morning, you have the satisfaction of knowing that nothing else that day can possibly be as unpleasant.
3) Think quantity, not quality. If your life is unstructured, or you often worry about whether you're doing things well enough, return to the rigidities of the factory production line. Decide how many hours you'll dedicate each day or each week to a project, and the reverse of Parkinson's Law often kicks in: the work gets done in the time available.
4) Do one thing every day that scares you - a heuristic from Eleanor Roosevelt with good sense behind it, so long as you don't apply it to how you cross the road or under what circumstances you eat puffer fish. Risk-taking is how everything significant gets achieved, but it's much more comfortable to act according to habit than to take risks - ergo, turn risk-taking into a habit.
Of course, this has been said many times before, by many others. Reading it doesn't make the application any less easy. I have one more thing to add: switching tasks, as I often do, seems rather inefficient. If you can focus for at least one hour, preferably two, on a single task, you'll increase efficiency.