Reading nonprofit ratings won't tell you everything
The big three rating agencies—Charity Navigator, BBB Wise Giving Alliance, and GuideStar/Candid—give you numbers. They don't give you context. I've sat through grant review panels where a perfectly scored charity got rejected because their program model doesn't actually move the needle where it matters. Here's what you should know before you pick which organizations to support. Most people start with Charity Navigator's four-star list and stop there. That's a decent filter but it's not a strategy. A four-star rating measures financial health and accountability, not impact. A nonprofit can be flawlessly managed and still be doing something that doesn't actually change outcomes. I learned this the hard way when I was volunteering on an advisory committee for a mid-sized education nonprofit. They had a 4.6-star rating, zero governance red flags, and spent roughly 80 percent of their budget on programs. Still, our quarterly data showed their tutoring model had zero statistically significant effect on graduation rates after three years. We reallocated funding toward mentorship instead. Graduation rates climbed. The rating stayed the same because nothing about how they managed money changed. That's the first thing to understand: ratings measure management quality, not results. So if you're trying to figure out which nonprofits are actually impactful, you need to go past the stars.
Where to actually look for impact data
Start with results reporting. Organizations that publish independent evaluations—randomized controlled trials, quasi-experimental designs, peer-reviewed studies— tend to be the ones worth your attention. Look for links to third-party evaluation reports on their websites. If the only evidence of impact is a story about one kid who "changed his life," that's marketing. That's not data. Giving What We Can has a decent page-by-page breakdown of cause areas and which charities have the strongest evidence behind them. Their effective altruism framework isn't perfect, but it's one of the few places where people actually force themselves to compare impact per dollar across different organizations. Open Philanthropy does the same at a larger scale. These aren't endorsements. They're starting points. For global health specifically, the Against Malaria Foundation, the Schistosomiasis Control Initiative, and the Evidence Action Demos project all publish cost-per-outcome estimates alongside peer-reviewed evaluations. Distribute's website breaks down where each dollar goes and what evidence supports their giving recommendations. This level of transparency is still rare outside the EA-adjacent ecosystem.
The sectors most worth looking at right now
Animal welfare. Effective altruist researchers and a growing number of independent evaluators have landed on factory farming reform as one of the highest-impact areas you can donate to simply because the scope of suffering is so enormous and the cost per unit of improvement is so low. The Humane Society's lobbying work and corporate campaign divisions, along with groups like Animal Equality and Sentience Institute, work on policy and public awareness rather than direct services. They're harder to rate on traditional charity metrics because their impact is structural and slow-moving. That doesn't make them less valid. It makes them different from a food bank. Cause area selection matters more than organization selection in many cases. Donating to a top-rated malaria net distributor will almost always save more lives per dollar than donating to a top-rated arts foundation in Connecticut. That's not a moral judgment. It's arithmetic. Poverty alleviation through direct cash transfers is another area with unusually strong evidence. GiveDirectly has randomized controlled trial data showing that unconditional cash transfers improve consumption, asset accumulation, and psychological well-being. The evidence isn't uniform across all contexts, and there are legitimate debates about long-term effects, but the overall picture is stronger than almost any other development intervention.
Get the Full Details
Red flags that ratings won't catch
I once spent two weeks trying to verify whether a climate nonprofit was actually doing anything beyond awareness campaigns. Their website looked professional. They had a 4.5-star rating. Their annual report featured a glossy infographic with arrows pointing upward. When I dug into their actual program expenditures, roughly 60 percent of their "program" spending went to consultants and communications firms. The remaining 40 percent funded events and webinars. Zero direct environmental intervention. Zero measurable outcome data. Just awareness. Here's what to check: look at their Form 990. Anyone can pull it from ProPublica's Nonprofit Explorer or Guidestar. Focus on Schedule G for fundraising expenses and Schedule P for program service accomplishments. If a nonprofit claims a major program but Schedule P describes it in three vague sentences with no quantified outcomes, that's your signal. Also check the ratio of program spending to total expenses. Below 65 percent program spending is a warning sign. Above 75 percent is good but not sufficient on its own. Another thing to watch: organizational mission drift. A nonprofit that started with a clear, narrow purpose and has since expanded into five unrelated program areas is usually chasing funding streams rather than pursuing impact. Each new program area dilutes expertise and makes honest measurement harder. I've seen this pattern repeatedly.
Edge case: when the best organization isn't the most famous one
During a grant review cycle a few years ago, I was evaluating applications in the global health space. Every panelist had heard of the big names. The most compelling application came from a small organization based in Kenya that distributed deworming tablets through existing school infrastructure. They didn't have a celebrity board member. Their website was functional at best. But their cost per disability-adjusted life year averted was roughly a quarter of what the larger orgs were achieving, and they had published evaluation data in the Journal of Development Economics. We funded them. The larger orgs kept getting grants because their brand recognition opened doors. That's just how the sector works. The workaround I ended up using was simple: I stopped evaluating organizations by reputation and started requiring applicants to cite their most recent independent impact evaluation, even if it was unpublished. It took longer to review applications. It also filtered out a lot of well-branded but low-evidence entries.
How to actually compare organizations within a cause area
Pick one cause area. Don't try to diversify across ten causes while you're still learning. Read the leading independent reviewers for that area. For global health and poverty, that's GiveWell. For animal welfare, it's Animals Australia and the animal ethics section of Giving What We Can. For effective education interventions, look at Evidence for Education. Then look at the organizations those reviewers recommend and compare them yourself. Check whether the reviewers' assumptions match your own. GiveWell, for example, heavily weights cost-effectiveness calculations based on RCT data. That means organizations working on advocacy, policy change, or capacity building—areas where impact is real but harder to quantify—get ranked lower or excluded entirely. That's a feature, not a bug, but it's a bias you should know about. If you want raw financial data without interpretation, ProPublica's Nonprofit Explorer and Guidestar are free. You can download full IRS filings, compare executive compensation, track revenue trends over time, and see where grant money comes from. This takes effort. It's also far more informative than any star rating.
What most people get wrong about donating
The biggest mistake I see is treating charitable giving like consumer shopping. You pick the brand with the best packaging and the highest rating and you're done. Charities don't compete the same way consumers do. A well-run charity doing marginal work can be a worse use of your money than a poorly branded charity doing transformative work. The rating agencies can't fully resolve this because they lack access to primary impact data. They rely on self-reported information and financial ratios. Another mistake is focusing exclusively on operating ratio—the percentage of expenses that go to programs rather than administration. A charity that spends 90 percent on programs but runs a program that doesn't work is worse than a charity that spends 70 percent on programs and runs a program that works extremely well. Overhead stings look bad in brochures. They're also often necessary. A charity that can't afford competent staff, proper evaluation, or adequate technology will underperform regardless of its overhead ratio.
Practical next steps
Decide what cause area aligns with your values. Pick one. Go to the independent reviewers for that area—GiveWell, Animals Australia, Evidence for Education, or the relevant section of Giving What We Can's website. Read their current recommendations and their methodology pages. Then pull the Form 990s for the top three organizations on your short list. Check whether their program spending matches their impact claims. Look for independent evaluations linked from their sites. If you can't find them after a reasonable search, that's information in itself. The process takes time. Maybe two or three hours for a well-researched decision. It's faster than most people think once you know where to look. And it's the difference between donating to something that looks good and donating to something that actually works.
