This article first published in MoneyWise magazine.

MacKenzie Bezos, who recently divorced the world’s richest person, Amazon founder Jeff Bezos, has pledged to give away at least half of her $37billion fortune. To whom should she give it?

The basic trick is to notice that some causes are much better than others. And some charities are much better than others: they achieve better results with the same resources. So it pays to be proactive and do your research.

The best opportunities may not be from the ones who stop you in the street, clipboard in hand. Nobody sensible makes financial investments reactively, solely in companies that put flyers through their letterbox, and it is the same with charitable donations. Smart donors go and seek out the best opportunities.

I’m not talking here about deciding whether to sponsor your friend running a marathon for charity: that’s more about supporting your friend. Rather, I’m talking about your main donations. And the same principles apply whether you are giving your own money, or if you’re involved in a foundation or your company’s charitable giving programme.

There are two levels of choice. The first is what cause to give to. This choice may be quite obvious to you if, say, you’ve lost a family member or friend to a particular disease. But either way, it’s worth knowing the numbers.

The charity 80,000 Hours, which gives advice on maximising the impact of your career (the average career involves 80,000 hours of work), recommends choosing causes that have three characteristics.

First, that they are important: the problem is widespread and has serious consequences. On that basis, it is better to support the poorest of the poor, where a little additional resource can improve lives dramatically, than to support people who are doing rather better, where that same additional resource won’t achieve as much.

Many Brits like to support animals. The UK has about 8 million cats. But check out how many factory-farmed animals there are! In the US alone, over 8 billion chickens are killed annually for human food, and around 95% of the laying hens there each live in a space smaller than a piece of A4 paper. On that basis, factory farming would be a priority cause.

Secondly, the cause should be neglected. In other words, go where other donors aren’t, where there are opportunities that are short of resources. Take health research. The UK spends £9 each year on research into mental illness for each person who is affected. By comparison, we spend £228 per person affected by cancer – which is 25 times more.

Third, the problem should be ‘tractable’: there should be some solution that has a reasonable chance of working. For instance, you may think that the alleged interference in Western democracy of foreign powers is an alarming problem, but if there is no organisation working on that to which you could donate and which any evidence suggests will succeed, then the problem is not tractable for you.

Let’s turn now to the other level of choice, which is choosing a particular charity to support. People often seem to give to the first charity that comes to mind within their chosen sector. In donor interviews and research, I’ve often heard people say things like: “I support Shelter because I wouldn’t want to be homeless.” That’s fine, but it’s worth understanding what the organisation actually does, and seeing if that is the area that you think is most important and neglected and tractable.

For instance, Shelter provides advice, support and legal services to people who have bad housing or who are homeless, and it does research and campaigning. It does not actually provide housing. Or a charity such as Cancer Research UK often comes to mind for people who’ve been affected by cancer. I have nothing against Cancer Research UK, but it’s worth remembering that (as its name suggests) it does research: it does not provide support for people with cancer or their families, in the way, say, Maggie’s Centres do at hospitals or hospices.

Some charities achieve more than others. The Home Office has analysed the effects of various programmes aimed at reducing crime levels. It found that a programme by Key4Life reduced offending by 18 percentage points, whereas the programme by Langley House Trust reduced it by just eight percentage points.

And even though that is only half as good, that is still a decent achievement: several programmes made the problem worse, and actually increased offending – one programme by as much as 12 percentage points. You can’t be aware of those results without looking at the data.

Giving to charity is a wonderful thing to do, and motivated by our hearts. But to give really well, we need to be proactive, engage our heads, and look at the evidence of what is needed and what really works.