Or, viewed another way, "showrooming" your local bookseller's stock, then ordering the book more cheaply from Amazon, works well for you- until the bookseller goes out of business.

Applying utilitarian philosophy to charitable giving can produce grindingly "fair" results, say two authors:

...Determining which charities should receive donations is at the heart of effective altruism. Singer and MacAskill begin by noting the thoughtlessness with which most people give – motivated by sentimental media appeals, associations with personal experience (funding research on a disease that killed one’s child), local connections (the local museum or hospital or community chest), or salient current disasters (earthquakes, tsunamis), without any attempt to find out how much good a donation would do by comparison with other destinations for that amount of money. This is partly a matter of motivation: a great deal of charitable giving is just a form of self-expression or a mark of solidarity rather than an attempt to have an effect. Or if it aims to do good, it does not aim to do so impartially, by the utilitarian metric, but deliberately favours some types of good and some people over others. Still, there are plenty of donors who want to benefit the worst-off people in the world, but who still make no effort to determine on the basis of evidence how to do so most efficiently. Or they rely on irrelevant evidence, such as the percentage of the charity’s budget that is spent on fund-raising and administration. As MacAskill says, one would never decide to buy a computer based on the size of the manufacturer’s executive salaries or its advertising budget. What one wants to know is how much value one is getting for the price, compared with the alternatives, and that is what should determine charitable giving as well. 

This kind of empirical calculation, if it accepts utilitarian impartiality, gives stark results. Westerners should give to no beneficiaries in their own countries, only to charities that benefit the poorest people in the world in the cheapest possible way – usually by preventing or treating illnesses that hardly exist in our countries because they are so easy to eliminate. To give one of Singer’s and MacAskill’s favourite examples, you might think it worthwhile to contribute for the provision of guide dogs for the blind, which certainly improve the quality of life for blind people. But it costs $50,000 to train one such dog which will work for nine years, and for that amount of money it is possible to pay for surgery that will save 500 people in the developing world from blindness caused by trachoma...