Holden's analysis seems sensible. To it I would add only two points:

1) Ratchet up your giving each time you get a higher income - at the time the new money is coming in, and before you start thinking about how to spend it. After you get a raise, especially an unexpected raise, is the best time to donate - rather than waiting to think about it at the end of the year.

2) Holden's analysis implies that if you have no income, e.g. as a student, you should wait to donate until later. This strikes me as basically correct but it is really really surprisingly important to give something on a regular basis - and put cognitive effort into efficient altruism / optimal philanthropy / rational charity, to get into that habit as well, even for small amounts.

At the end of every year, for example, you might donate $100 - or even $10, if $100 is too much - after looking over the latest list of efficient charities and doing some thought about where the $100 ($10) will do the most good, for purchasing utilons rather than fuzzies, just as if you were about to give $10,000. If the end of the year is far enough away and you don't have a trustworthy reminder system already set up, you might do that part now, then again a year later or at the end of the year, etc.

If you don't do this part, I would evaluate a surprisingly low chance that you would remember to start giving, and giving efficiently, later in life when you have income.