Wikipedia can answer a question without the person reaching its website. That distance also separates use of the encyclopedia from an opportunity to explain who supports it and invite a contribution. For the Wikimedia Foundation, extending the reach of knowledge and acquiring the next donor are related but different outcomes.
The growth challenge is to turn recognized usefulness into a voluntary relationship that lasts. More citations in automated answers can spread the content; funding also requires a clear reason to contribute, a straightforward way to do so and trust after the first payment.
Fewer new donors can coexist with higher revenue
The 2024–2025 fundraising report records a 10.2% decline in new donors, from 2.6 million to 2.3 million. These are worldwide, rounded counts for July 2024–June 2025; the reported change is not recalculated from those rounded totals.
The audited annual report provides a counterweight: $208.6 million in revenue and $190.9 million in expenses in 2024–2025. It also records$296.6 million in year-end net assets. Net assets are not available cash, and these accounts do not justify describing an imminent shutdown emergency.
The distinction matters for marketing. Total funds raised can hold up while fewer people contribute for the first time. A dashboard that rewards only the annual amount would detect a replenishment problem late. Tracking people, gifts and money separately reveals whether the base is renewing or increasingly depends on existing supporters.
The comparison requires separate accounting bases. The campaign report measures gifts received using its own methodology; audited accounts include other items and recognition rules. A percentage of fundraising does not become a percentage of total revenue simply because both cover the same fiscal year.
An answer citation does not replace a reader relationship
Wikimedia’s 2026 global planning assessment identifies changes in traffic, awareness and participation. It is the organization’s own diagnosis, not a measurement assigning every loss to a single platform. The analysis must distinguish use of the knowledge from visits returning to the site.
In Pew’s panel, pages with AI summaries received fewer traditional-result clicks
Google, United States, March 2025. Panel of 900 adults and 68,879 searches. Share of visits with a click on a traditional result.
Tap or hover for details. All values are also available in the data table.
Collapse notes
Observed association without experimental assignment. Search results reconstructed April 7–17. Does not measure donations, losses caused to Wikipedia or behavior in Spain.
Explore the data table
| Category | Visits with a click |
|---|---|
| Without AI summary | 15% |
| With AI summary | 8% |
The U.S. Pew panel provides search context in which AI summaries coincided with fewer clicks. It cannot calculate how many Spanish donors Wikimedia loses. It does challenge a commercial equivalence: appearing in an answer does not ensure an opportunity to explain funding.
A visible reference to the source can help people recognize the human work behind an answer, but its value for acquiring support depends on what follows. Brand recognition, arrival at an owned destination, understanding the funding model and making a first contribution should be separated. Improvement at the first stage does not prove improvement at the last.
For a free encyclopedia, that separation also protects the mission. Recovering visits at any cost could reward headlines or friction that make knowledge harder to consult. The opportunity is to offer a direct relationship to those who want it, while keeping reading independent of payment.
The first gift should begin a relationship with permission
Recurring gifts and email already supported two funding routes
Wikimedia Foundation, July 2023–June 2024. Two selected channels from the historical breakdown published by Statista.
USD million
Tap or hover for details. All values are also available in the data table.
Collapse notes
Selected channels, not a complete ranking. Excludes the Endowment. Global nominal amounts, not Spanish revenue. Do not add to audited accounting figures: the bases differ. Does not represent FY 2024–2025. Dollar amounts verified directly in the original report chart.
Explore the data table
| Category | Amount received |
|---|---|
| Recurring donations | 36.2 USD million |
| 31.6 USD million |
Statista’s historical breakdown compares two ways of supporting the project: permission for recurring support and a relationship that can continue through email. The snapshot covers 2023–2024. It helps explain the mechanism without presenting it as the current channel mix.
The subsequent fundraising report describes annual recurring gifts introduced in December 2024 and a milestone of 1 million monthly donors in 2024–2025. Renewal is already part of the model. The editorial question is how to expand that base without confusing clear authorization with a payment the person did not understand.
A first gift and recurring authorization address different decisions. The latter requires trust about future charges: amount, frequency, confirmation and the ability to change or cancel. The journey can explain those conditions before requesting a commitment and retain an equally understandable one-time option.
Continuity should be assessed through support that remains, after failed payments, refunds and payment-processing and service costs. A larger first payment does not guarantee a better relationship if it ends sooner. Nor should every cancellation be read as rejection of Wikipedia; the person’s capacity to give may have changed.
Enterprise access pays for a different service
In January 2026 , Wikimedia Enterprise announced new partners including Amazon, Meta, Microsoft, Mistral AI and Perplexity. Its commercial explanation distinguishes packaged enterprise access and support from the content, which remains freely available, including for commercial reuse.
This distinction avoids selling incompatible promises twice. Readers are invited to support a common resource; businesses are offered a service for consuming data at scale under specific conditions. Enterprise acquisition can complement small gifts, but it does not automatically turn every Wikipedia use into an invoice or guarantee a replacement for new donors.
The two relationships need their own economics. API revenue should be compared with delivery, maintenance and customer-support costs. Individual giving requires attention to acquisition, collection and continuity. Adding contracts and donations without distinguishing their costs would obscure which route actually funds more free-knowledge work.
Trust must survive the fundraising appeal
A sound invitation connects everyday usefulness with concrete uses of money and understandable financial information. Reserves can be explained alongside operating needs, without calling all net assets cash or reducing the organization’s work to server hosting. Transparency is part of the reason to support it.
A useful acquisition decision would compare messages and timing through a complete measure: new people who maintain their support and net funding raised, alongside the effect on reading. An immediate fundraising increase would not compensate for an experience that weakened trust or made content harder to access.
The distinguishing criterion is reader autonomy. Wikimedia can spread knowledge beyond its website while building a reason to return, participate or donate. Sustainable growth occurs when people understand who makes that usefulness possible and choose to support it while access remains free.
In this article
Marian Vasile
Marian Vasile is a MarTech and growth professional. His work spans product marketing, SEO, paid advertising and deep online tracking, including behavioral analytics and attribution.
