AI SEO for Nonprofits: The Full Playbook

Right now, someone who cares about your cause is typing your mission into ChatGPT. They are not typing "best children's charity." They are typing four sentences about their kid, their diagnosis, and the bill they cannot pay. Or four sentences about the two hours a week they want to give and the cause that keeps them up at night.

And AI gives them three names. If your organization is not one of those three names, you do not get a second chance. There is no page 2 to scroll to.

I adapted this playbook, with permission, from Sabrina Ramonov's "AI SEO for Beginners" and rebuilt it for mission-driven organizations. In our AEO work with healthcare nonprofits, the pattern I keep seeing is that the organizations winning here are not better funded. They simply describe themselves the same way in more places. Every section below ends with a prompt you can paste straight into ChatGPT or Claude, so you are not just reading. You are doing.

1. What is SEO, and what just broke?

SEO stands for search engine optimization: making it likely a search engine shows your page to someone looking for what you do. For years the shape was simple. Someone types a question into Google, Google shows 10 blue links, you want to be near the top because the top gets the clicks. Every piece of SEO advice was built on one assumption: the person will click.

That assumption just broke. When someone asks ChatGPT or Claude or Perplexity a question now, they do not get 10 links. They get an answer, a short one, with maybe three organizations named inside it. And they usually stop there. If nobody clicks, ranking number one is worth nothing by itself. The prize is not the click anymore. The prize is being named in the answer.

Here is the difference, in nonprofit terms:

• What you win. Old SEO: a blue link on a results page. AI SEO: your name inside the written answer a donor reads.
• The goal. Old SEO: rank in the top 10 links. AI SEO: get cited as one of three recommended organizations.
• How you win it. Old SEO: keywords, backlinks, page speed. AI SEO: the same facts about your mission repeated consistently across many sites.
• What you measure. Old SEO: clicks and sessions. AI SEO: how often you get named, and what people do next.
• Who it favors. Old SEO: big national orgs with long histories. AI SEO: any organization described the same way in many places.

This is not a niche of tech people. A majority of US adults now use AI assistants, and for "who should I trust with this" questions they increasingly start in a chatbot before they open Google. A growing share of your donors, volunteers, and the families you serve are forming their first impression of you inside AI chat.

2. Why AI search is good news for nonprofits

Because AI traffic is tiny but far warmer than search traffic. Think about what the person just did. They described their exact situation to an AI. A parent typed out their child's condition. A donor typed out the cause they care about and the amount they can give. The AI gave them a shortlist. They are not browsing. They are deciding.

So a small local nonprofit named in 20 answers can outperform a giant organization that gets 100,000 random visitors who never convert. And here is the opening: big organizations move slowly. Comms gets approved, legal reviews the language, four people sign off before a page goes live. Most national orgs will never publish a page that plainly names peer organizations or answers a hard donor question in specifics. You can publish one tomorrow. Speed is the only advantage you need, and it starts with the cheapest thing on this list.

3. Beginner: pick the one thing you want to be known for

Start here because it takes five minutes and it is the one page about you that you fully control. AI models read LinkedIn constantly, both your organization's page and your executive director's personal profile. When an AI is deciding whether your nonprofit is credible enough to name, LinkedIn is one of the top things it trusts.

Most nonprofits write "501(c)(3) nonprofit organization dedicated to making a difference." That tells a machine nothing; every nonprofit could copy-paste it. Lead with the specific outcome, for a specific group, in a specific place instead:

"We close the pediatric wound-care gap for families across the Southeast."

That is the one thing. Everything after it exists to prove the claim: children served, clinicians trained, named hospital or school partnerships, a grant from a funder people recognize, years doing the work. The structure is: [the one thing you want to be known for] + [proof] + [proof] + [proof]. "501(c)(3) serving the community" tells a model where you filed. "The organization that gets wound-care supplies to uninsured kids in rural Georgia" tells a model what to recommend you for.

Paste your answer into your LinkedIn headline and your organization's About field before you close this page. Later means never, and this is the one thing everything else depends on.

4. Say the same thing everywhere

Here is the part almost everyone gets wrong. A nonprofit writes a good description on their website. Their Candid (GuideStar) profile says something different. Their Charity Navigator listing says a third thing. A grant application from two years ago says a fourth. An AI reading all of that gets a fuzzy picture, and fuzzy means it will not risk recommending you.

A model is not looking up a fact about you in a database. It is noticing which words keep showing up next to your name across the whole web. If your name appears next to "wound care for uninsured children" on 30 sites, the model treats that pairing as true and repeats it. If it appears next to four different descriptions, it has nothing confident to repeat, so it names a peer organization instead. Consistency is not tidiness. It is the actual mechanism.

Before you paste your claim anywhere, check one thing: some website hosts block AI crawlers by default. Open your robots.txt and make sure the search crawlers are allowed: OAI-SearchBot for ChatGPT, Claude-SearchBot for Claude, PerplexityBot for Perplexity, and plain Googlebot for Google's AI answers. If they are blocked, everything on your own site is invisible.

Then paste a version of your claim into your website About page and homepage; your LinkedIn, Instagram, Facebook, YouTube, and X bios; your Candid, Charity Navigator, and GreatNonprofits profiles; your annual report's opening paragraph; the "about the organization" boilerplate in every grant application; your email signature; and your ED's speaker or podcast bio. While you are in those profiles, ask five people you have served or five long-time volunteers to leave a GreatNonprofits review. Real reviews with specific details get quoted by AI far more than a star rating does. Follow your consent and privacy practices for anyone you serve.

5. Which pages do AI answers actually quote?

Guides that compare options, "best of" lists, and "how to help" pages get cited far more than a blog post about your gala, because they answer the exact question the person asked. Three formats do the work:

And yes, you can write a "best organizations for [your cause]" guide that includes your own organization. The rules if you do: be honest (if a peer is better at something, say so); include the organizations that genuinely belong, even strong national ones; give real specifics, not "a great organization"; and put a comparison table near the top. Nonprofits are not competitors the way software companies are, and a generous, accurate guide earns more trust, not less.

6. Intermediate: find the questions your people actually ask

Pull them from your own donor inquiries, volunteer applications, intake and help-line questions, and grant reviewer questions. Those are the exact words your people use. Do not guess. The words are already sitting in places you own.

You are looking for three buckets. Decision terms: "is [org] a good charity," "best nonprofit for [cause]," "where does my donation actually go at [org]." Need terms: the specific problem you solve in the person's own words, not "addressing food insecurity" but "my kids and I ran out of food and rent is due Friday." Fit terms: "nonprofits helping [cause] in [state]," "volunteer opportunities for [skill]," "grants for families dealing with [condition]." Skip anything purely educational; "what is a 501(c)(3)" brings in students, and AI answers it without naming an organization.

Here is what differs from old keyword research: when someone asks an AI a long question, it does not run one search. It quietly breaks the question into several smaller ones and searches each. Someone asking "what's a trustworthy organization that helps uninsured kids get specialty medical supplies in Georgia" triggers hidden searches for pediatric medical-supply charities, charity trust ratings, and uninsured medical help in Georgia. Every one of those is a door into your page.

7. Intermediate: format a page so AI can quote it

An AI grabs one passage from your page and drops it into an answer with none of the surrounding context, so every chunk has to stand by itself. Five writing changes, easiest to biggest difference:

• Make headings into real questions. Not "Our Programs" but "How do I get wound-care supplies if I'm uninsured?" That matches how people ask. Question-style headings in a clean H2 to H3 order roughly triple your citation odds.
• Answer in the first two sentences. A found chunk has about 40 to 60 words to be useful. Say the answer, then explain.
• Trade every vague word for a specific one. "87 cents of every dollar goes directly to programs" beats "most of your donation goes to our mission." "We served 1,240 children last year" gets quoted; "we help many families" never does.
• Say where your info came from. One line like "figures from our 2025 annual report and IRS Form 990" does real work, because models weigh visible sourcing.
• Write table rows that stand alone. A cell saying "Free wound-care kits for uninsured children 0 to 18 in Georgia" can be quoted straight into an answer. One good table can get you named a dozen times.

Then three jobs make the page findable: link your own pages to each other (five or more internal links per important page); add Organization schema on your homepage and FAQPage schema on any page with questions; and refresh what already ranks every three months. About 83% of AI citations for evaluation questions come from pages updated in the last 12 months, and pages not refreshed quarterly are roughly 3x more likely to lose their citations. This mirrors the citation-drop shift we tracked across our own client work, and it maps to the moves in our AEO 2026 guide.

8. Get AI to mention your nonprofit

Get other websites to name you. Roughly 85% of the brand mentions AI makes come from sites that are not yours, so everything on your own site has a ceiling until outside pages talk about you. The free version takes an afternoon: ask ChatGPT, Claude, and Perplexity your main cause question ("what are the best organizations helping with [your cause] in [your area]?"), write down every article each one cites, open each one, and see whether you are on it. If not, email the author.

Hi [name], I read your piece on the best [cause] organizations. I run [your org], and we [specific thing the others on your list don't do]. Happy to send our latest program numbers, outcome data, and a family story we have consent to share, if it's useful for your next update. No pressure either way.

Every author who adds you becomes a source an AI can quote. While you are at it, claim the free listings AI models lean on for nonprofits: Candid (GuideStar), Charity Navigator, and GreatNonprofits, each filled out completely with the same claim from section 3.

9. Advanced: press releases, podcasts, and other people's platforms

News sites get crawled constantly and carry a lot of weight, so one wire release can put your name on outlets you would never reach by pitching. Use a service like EIN Presswire (they have nonprofit rates). Nonprofits have real news constantly: a grant awarded, a new program, a milestone (the 10,000th family served), a hospital or school partnership, a new board chair, an impact report. Front-load the news in the first paragraph (who, what, why anyone should care, under 50 words), write it in third person with one substantive quote from your ED, and point your two or three link slots at your donate, program, and "how to get help" pages, not your homepage.

Then get onto platforms that keep working. A podcast show-notes page sits there with your ED's name on it forever and gets re-crawled every month; twelve guest spots build twelve permanent pages. Search your cause on listennotes.com and email the small and mid-size shows, which need guests constantly. YouTube is the one you do not have to wait to be invited to: make short videos answering the exact donor and help-seeker questions from section 6. Sector publications (Nonprofit Quarterly, the Chronicle of Philanthropy, your community foundation's blog) and conference talks all create trusted pages with your name and mission on them.

10. Advanced: track whether AI is recommending you

Ask the AI models your donors' and families' questions every month and log whether you got named. That count, not traffic, is your scoreboard now. Re-run the crawler check from section 4 first, because a blocked crawler is the one thing that makes this scoreboard read zero for reasons that have nothing to do with your value. Then measure three things, in order: ask the AIs your 10 buyer questions monthly and log who got named; watch branded-name searches in Google Search Console (getting named leads people to look you up); and count what matters, tagging where donations, volunteer signups, and help requests come from.

One honest expectation: this takes time and it compounds. Every page you publish, every profile you fix, every mention you earn stacks on the last one, which is also why it keeps paying long after you stop. Anyone promising results next week is selling something.

11. Bonus: the board conversation

Do not ask your board to fund "AI SEO." It is abstract. Sell them the result: "In 90 days, when a donor asks ChatGPT which organizations to trust for [our cause] in [our region], our name comes up. Right now it doesn't. Here is the proof." Then show them the proof. Take the 10 questions from section 10, run them on your own organization across ChatGPT, Claude, and Perplexity, screenshot every answer, and show the board a peer organization's name highlighted and yours missing. You are not asking for a big budget. You are showing them a gap a competitor for the same donors is already filling.

A model does not rank your website. It reads your site, your profiles, other people's lists, a news story, a podcast page, and it repeats the version of your organization that shows up the most. The winner is not whoever writes the cleverest page. It is whoever is described the same way in the most trusted places.

Key takeaways

• Beginner: pick your one claim, say it identically everywhere, and write cause guides, "how to help," and "how to get help" pages.
• Intermediate: mine the real questions your people ask, and format pages so any single paragraph stands alone.
• Advanced: run press releases, guest on podcasts and YouTube, and track how often you get named across the AI models.
• Measure the named-in-answers scoreboard monthly, not website traffic.
• The nonprofits winning at this are not better funded. They just started earlier, when it felt too early to matter.