I second this question. There are good nonprofits which will find it difficult to answer precisely, e.g. one which focuses on basic science research (which has, over the course of human history, contributed rather a lot to human welfare). But it is certainly a question that nonprofit founders should be thinking about right from the start, and producing vague estimates accordingly, just to get into a quantitative frame of mind; otherwise it is a very bad sign.

I don't like this question. It's akin to asking a startup, "What is your revenue model, and how will you be cash-flow positive?" Eventually every startup will need to answer that question, but asking it too early is more likely to stunt important and creative ideas that don't come with cookie-cutter business models than it is to reveal a winner. Much better to ask open-ended questions that get at the same thing: What's the potential impact of your idea? Why are you going to succeed where others failed? And if the answer is "We're smart and we have these advantages and everyone seems to like what we do" then it may be OK not to have the cost-effectiveness worked out in detail yet.

There's a nuance to asking about cost-effectiveness for nonprofits that doesn't exist for for-profit corporations -- and that is that a nonprofit can be wildly successful by every metric except helping people.

There are charities who take in millions of dollars, keep some, and spend the rest to send bibles to Africa. By every measure, they're "killing it": they have a well-paid staff, a happy donation base, and they send out a heckuva lot of bibles. It's an easy business model to copy, and an almost-guaranteed smash-hit in terms of getting returns on investment. But, here's the rub--do you think there's any evidence that sending people bibles increases their quality-of-life?

Basically, I'm hoping here that PG is getting into the nonprofit sector because he wants to help people--not just to become (more) rich by looking like he's helping people.