By Madeleine Alegria, Senior Researcher, Head of Research
Local nonprofits across the country are closest to the issues but hardest to find. When donors sit down to choose where to give, search engines bury their neighborhood nonprofits beneath well-branded and well-funded institutions. This disparity comes at a time when community need is surging to all-time highs, with 77 percent of nonprofits reporting increased demand for services last year, and 36 percent ended 2024 in an operating deficit.
Americans are meeting the increased need with greater generosity and empathy, including a 3.3% increase in philanthropic giving in 2025, adjusted for inflation. However, barriers to giving, such as information overload, decision fatigue, and lack of trust, are prohibiting this generosity from reaching nonprofits on the margins. As a result, fewer than half of U.S. households donate to nonprofits, a 23 percent drop in participation over two decades. Individual giving participation is declining alongside public trust in institutions. Donors see nonprofits as well-intended but less competent than their for-profit counterparts, and 2023 saw the largest single-year drop in public confidence in nonprofits.
The current search infrastructure is quietly functioning as a funding mechanism, redistributing charitable dollars in ways that have almost nothing to do with impact, community need, or donor values. Nonprofit stories and impact data are hidden in dense reports, embedded across tedious 990 filings, annual reports, and website links that search engines ignore and donors rarely see. The sector doesn’t have an impact problem; it has a visibility problem.
What Gets Found Gets Funded
In 2024, we conducted a study and found that 69 percent of donors preferred supporting local nonprofits, yet more than half made their most recent donation to a nationally serving organization.
The nonprofit sector is far more diverse than search engines that are trained on website traffic, domain authority, and SEO investment suggest. Organizations with annual revenues over $5 million represent fewer than 3 percent of all U.S. nonprofits, yet consistently dominate philanthropic and government grant awards. On the other hand, 92 percent of nonprofits operate with budgets under $1 million. These are the food banks, legal aid clinics, youth programs, and advocacy organizations embedded in the places where people actually live. Caught in a starvation cycle, these organizations are increasingly being asked to respond to greater community needs with fewer resources.
This has material consequences for rural communities, which account for over 24 percent of the U.S. population and represent 91 of the 100 most disadvantaged communities in the U.S., yet receive only 7 percent of philanthropic funding. Without the proper overhead to invest in program offerings, let alone optimize for search and marketing, these nonprofits lose ground to larger organizations that can afford to be found.
Experiment in Donor Search
In fall 2024, we ran a study to test whether a different search environment could produce different giving behavior. We recruited 291 American donors and randomly assigned them to two conditions: one group used our AI-powered nonprofit discovery platform, and the other used the general internet, including Google and any other tools they typically relied on. Both groups received $50 to donate to a nonprofit they had never given to before.
Donors using our platform were 79 percent more likely to give to a nonprofit in their own city. They chose organizations with significantly smaller net assets, roughly five times smaller based on logged asset size scores. They were more than twice as likely to give to a vetted, often lesser-known nonprofit. And by the end of the two-week study period, they reported meaningfully higher levels of trust in the nonprofit sector overall. These differences did not come from telling donors what to do; they came from changing what donors could see.
The platform uses natural language search and draws on verified data from millions of IRS filings and expert-reviewed content. It also incorporates behavioral nudges, short prompts embedded at the bottom of search results. When donors asked questions like “which nonprofits help people experiencing homelessness and direct the most to the cause?”, they were prompted to explore local organizations or consider giving criteria beyond brand familiarity.
Importantly, the two groups donated at almost identical rates and amounts. The intervention did not make donors more or less generous in absolute terms. It changed the direction of their generosity, and that direction shift matters enormously for the organizations on the receiving end.
The Information Gap is a Structural Problem
The philanthropic sector has invested heavily in evaluation, transparency tools, and capacity-building to evaluate grantees. We have been slower to reckon with the upstream problem: funders cannot support organizations they cannot find, and the infrastructure that donors and funders rely on is not designed with equity in mind.
This is not simply a technology problem; it is a design and values problem. Funders should consider how recommendation systems perpetuate biases toward large institutions. Algorithmic design determines which data to surface, how to rank results, which criteria to treat as relevant, and, ultimately, which nonprofits are funded. The result is that well-funded and well-staffed institutions continue to dominate recommendation algorithms and funding decisions.
When foundations or intermediaries build or fund donor-facing platforms, they make choices about what data to surface, how to rank results, and what criteria to treat as relevant. Those choices collectively determine which organizations are discoverable and funded.
Several findings from our study highlight how natural language and conversational search might break the cycle. Among the nudges that were clicked in our study, 35 percent of donors clicked prompts that narrowed their search to nonprofits near them. Nearly 15 percent were prompted toward vetted nonprofits that emphasize transparency, systems change, and values alignment. Donors engaged with an average of 2.5 platform features, compared to 1.9 for the general internet group, suggesting that richer information environments support more deliberate decision-making.
Donors and funders alike need infrastructure that reflects the full breadth of the nonprofit sector, not just its most visible slice.
Lessons for the Field
As the way we search shifts, funders have a clear opportunity to reinvest in infrastructure with nonprofit equity in mind. Funders should invest in search and discovery tools designed to surface community-based organizations, and audit existing tools for size and brand bias that may inadvertently concentrate resources among certain organizations. Additionally, disclosing data sources, vetting criteria, and recommendation logic will help build trust among donors and nonprofits alike.
This paradigm shift also means supporting nonprofit capacity-building in a new domain. Smaller organizations are increasingly disadvantaged not just in fundraising, but in discoverability. As donor search shifts towards natural language queries and recommendation platforms, nonprofit data access and transparency become increasingly important. Funders should consider how they support digital presence and data infrastructure for community-based organizations.
These results also have implications for nonprofits seeking to improve discoverability. While funders should do more to diversify their sources beyond IRS Form 990 filings, nonprofits should expect this data to become increasingly visible. Ensuring accurate, up-to-date, and machine-readable data is essential not just for financial data but also for program descriptions, leadership bios, and impact data. The more context a nonprofit makes available, the easier it is for the right donor to find them.
And for the field broadly, there is a question worth sitting with: If the tools donors use function as de facto funding mechanisms, quietly shaping which organizations receive support and which remain invisible, who is accountable for how those tools are designed and what metrics we prioritize?