GiveWell started in the hedge fund world (as a collaboration between coworkers), and our staff and board is heavy on for-profit experience. When we talk about metrics and dashboards, a lot of people assume we’re applying business concepts, and that we want to see charities run more like businesses. That isn’t true.

Do businesses conduct academic-style studies employing randomization and statistical significance tests to figure out how they’re doing? If any do, I doubt they’re very good ones. If you’re running a chain of pizza joints, you don’t need to know anything about statistics, and you can get by without much of a dashboard too. That’s because the single most important thing you need to measure – profit – is hitting you in the face every day. And most of the other things you need to measure take care of themselves too. It’s easy to know how much you’re paying delivery boys and how that impacts your bottom line – and if they slack off, you’ll be hearing about it from your consumers. No followup surveys necessary.

The more complicated things get, and the more difficult-to-observe things you have that might affect you in long-term and difficult-to-notice (but still important) ways, the more you need to audit. That’s why companies do it. But there’s no business whose operations are as difficult to understand – and whose outcomes are as difficult to measure – as even the simplest charity trying to fight poverty.

If a charity doesn’t follow up with its clients, it will never know whether its efforts are very successful, moderately successful, or entirely worthless. It will have no way of figuring out when the plan that made sense in its director’s head is falling short in reality. It could keep doing things that make logical sense, but don’t work at all, for hundreds of years – and never find out. None of that is true of any business.

Reading fundraisers’ arguments for why their charities are good, I keep hearing things like “We’re very experienced,” “We’ve been around for a long time,” “We’re well recognized/well respected/well established.” These are all points that make perfect sense as praises for a for-profit business, and no sense as praises for a charity. Charities are mission-driven, not self-gain-driven. Continued existence is not evidence of success; fundraising excellence isn’t either. The fact is that without evaluation, a charity’s success is something no one can see.

Believe me, I know about the inherent limitations of measurement, and I know how expensive and annoying it is too. I don’t like to fill out a survey every time I blow my nose; if I were running a business, I wouldn’t want to be spending half our budget on figuring out whether our pizza really was the cause of improved customer happiness, or whether selection bias were involved. But if I were running a charity, I wouldn’t see a choice. It’s tough but true: unlike a business, a charity needs all that annoying stuff – or it’s working in the pitch dark.