Against Malaria Foundation, GiveDirectly, and Direct Relief outrank charities many times their size when a donor asks an AI engine where to give. The Nonprofit Citation Share Study 2026 found these evidence-vetted organizations over-index in the trust layer — and in the global-health cause, GiveWell-associated charities capture the entire top of the table. Meanwhile some of the most recognized charity names in the country place lower than their fame would predict. The reason is simple and it is the most important lesson in the study: documented effectiveness is retrievable, and name recognition alone is not.
For every nonprofit, this is the finding to act on.
The data
In the study's global-health and poverty cause, the top of the table is almost entirely evidence-vetted charities — Against Malaria Foundation at 26 percent of citations, GiveDirectly at 21, Helen Keller Intl at 16, Malaria Consortium at 14. Doctors Without Borders, a far larger and more recognized organization, places fifth.
In the overall charity table, Against Malaria Foundation places third and GiveDirectly seventh — both ahead of, or even with, organizations many times their budget. The American Red Cross, one of the most recognized charity brands in the United States, places fifth and is frequently cited alongside the efficiency questions that have followed it in public coverage.
The study measures answer-engine citation, not charity quality — a larger organization ranking lower is not being judged a worse charity. But the pattern is unmistakable.
Why effectiveness compounds and recognition does not
An answer engine retrieving the trust question does not retrieve fame. It retrieves evidence. When GiveWell publishes a cost-effectiveness analysis showing a charity's impact per dollar, that analysis becomes a fact the engine can retrieve, quote, and attribute. When a charity is simply well-known, there is no comparable artifact — recognition lives in the public's memory, not in the source layer the engines read.
So the evidence-vetted charities compound. Each evaluation, each documented outcome, each cost-per-result figure adds to a retrievable record. A household name without that record has awareness the engines cannot convert into a citation.
What every nonprofit should take from this
The lesson is not "become a GiveWell charity." It is narrower and universal: measure your impact, document it, and publish it in retrievable form.
Measure outcomes, not just activity. Not "we served 10,000 meals" alone, but what changed as a result, measured.
Document it where the engines can read it. Outcome data and independent evaluation on owned properties and in evaluator profiles — not buried in an annual PDF.
Get the evaluator standing. A strong, current Charity Navigator rating, a Candid seal, a BBB accreditation, and GiveWell consideration where the cause and evidence fit. The evaluators are the engines' primary source for the effectiveness question.
Do not rely on the name. Recognition built over decades is real and valuable — but in the answer-engine trust layer it does not, by itself, produce a citation. Evidence does.
A nonprofit with genuine impact and weak Citation Share does not have an impact problem. It has a documentation problem — and that is fixable before the next giving season.
Common questions
Why do small effective charities beat large famous ones in AI search?
Because the engines retrieve documented evidence of effectiveness, which the evidence-vetted charities have, and cannot retrieve name recognition, which famous charities rely on. Documented impact is a retrievable trust signal; fame is not.
Does this mean large charities are losing donors?
It means large charities can lose the trust-layer query if they rely on recognition rather than documented, retrievable effectiveness. The fix is documentation, not a change of mission or scale.
What should our nonprofit do first?
Measure and document your outcomes in retrievable form, and complete your evaluator profiles. Those are the inputs the engines retrieve when donors ask which charity is effective.
About this research
This article was produced by Everything-PR.
Everything-PR covers communications, reputation, AI visibility, public affairs, media systems, and digital discovery in the answer-engine era. Thirty verticals. Original reporting, research, and analysis.
For coverage of how nonprofits manage reputation and visibility, see the Nonprofit Communications pillar. For how brands across thirty verticals are cited inside answer engines, see AI & GEO.
Published: May 2026 · Series: Nonprofit Communications · GEO & AI Visibility
