Generative Engine Optimization (GEO) for a nonprofit is not a new discipline. Google's own guidance says optimizing for artificial intelligence (AI) features is still search engine optimization (SEO): publish helpful, people-first content, keep it crawlable, and skip the llms.txt files and "AI schema" that vendors are quietly upselling this quarter.
The short version
- Google states directly that AI Overviews and AI Mode need no special markup, no llms.txt file, and no dedicated schema.
- Google calls optimizing for generative AI search "still SEO" in its official optimization guide.
- What earns a nonprofit a citation is a clear, first-hand point of view, not more pages.
- Search Console's Generative AI performance report, live since June 2026, lets you measure AI visibility instead of guessing.
- Most paid "GEO" tools charge for work Google has said is unnecessary.
GEO is still SEO, and Google put that in writing
This quarter, as boards ask what to do about ChatGPT and AI Overviews, I keep pointing people to the same place: Google's own documentation. In its guide to optimizing for generative AI features, Google is blunt. "From Google Search's perspective, optimizing for generative AI search is optimizing for the search experience, and thus still SEO." Answer Engine Optimization (AEO) and GEO are terms you will hear from agencies, but Google treats the work as ordinary SEO. That is not a dodge. It is the most useful thing a resource-constrained nonprofit can hear, because it means you do not need a separate budget line for a discipline that does not exist.
The llms.txt and "AI schema" pitches are selling you nothing
Here is where common advice is wrong, and it is worth naming the failure mode. Vendors have spent the last year telling nonprofits they need an llms.txt file, "AI-ready" markup, or a special schema so AI systems can read their site. Google's answer is explicit: "You don't need to create new machine readable files, AI text files, markup, or Markdown" such as llms.txt files, and "structured data isn't required for generative AI search." On its AI features documentation, Google repeats it: "There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary." If a tool is charging your organization a monthly fee to generate an llms.txt file, it is charging you for a file Google has said it does not use. Structured data still earns its keep for rich results and for making your entity unambiguous, but it is not a ticket into AI answers.
What actually earns a nonprofit a citation
The content that gets pulled into an AI answer is content that answers the question cleanly and comes from somewhere credible. When I rebuilt the resource library for a legal-services nonprofit, the pages that started surfacing in AI Overviews were not the ones stuffed with keywords. They were the ones where the first two sentences answered the exact question a person typed, the headings were phrased the way people actually ask ("Am I eligible for free representation?"), and the facts matched what the organization says everywhere else, from its About page to its Internal Revenue Service (IRS) Form 990. Google's guidance lands in the same place: create "valuable, non-commodity content" with "a unique point of view," support it with high-quality images and video, and make sure every page is "indexed and eligible to be shown in Google Search with a snippet." For a nonprofit, your unique point of view is the thing no content mill can fake: the population you serve, the outcomes you have measured, the local context you know. Put that first, in plain language, near the top of the page. If you are rebuilding your site anyway, bake this in from the start, because that is the cheapest time to do it.
Where AI assistants actually find nonprofit answers
AI answers are assembled from the open web and the sources a model trusts, which is why your own site is necessary but not sufficient. Google notes that AI Overviews and AI Mode "display a wider and more diverse set of helpful links" than a classic results page, so being one credible voice among several matters. In practice, that means the same off-site presence that has always helped nonprofits: an accurate, current profile on the databases people and models both reference, consistent mentions in local and sector news, and a website whose facts do not contradict those sources. If your mission statement reads three different ways across your homepage, your Candid profile, and your annual report, you are giving an AI system three chances to quote the wrong one. Consistency is an underrated ranking behavior, and it costs nothing but discipline.
Measure it, do not guess
You no longer have to argue about AI visibility from vibes. Since June 2026, Google Search Console (GSC) includes a Generative AI performance report that shows impressions and clicks from generative AI features, including AI Overviews and AI Mode. It is not perfect. It does not expose the queries behind that activity, and the data is not yet in the bulk export, so treat it as a directional signal rather than a full picture. Still, for a nonprofit that has never been able to prove AI search was happening, a monthly look at that report is enough to see which pages are surfacing and which are invisible, and to spend your limited content time on the gaps.
What this does not cover
This is about Google's AI features and durable practice that carries across engines. It is not a reverse-engineering of ChatGPT or Perplexity, which run their own retrieval systems and publish no ranking documentation, so anyone promising you a guaranteed placement in a specific assistant is guessing. I am also not walking through JavaScript Object Notation for Linked Data (JSON-LD) implementation here, or paid AI placements, both of which deserve their own treatment. The point is narrower and more useful: stop paying for GEO theater, and spend the same hours on content and consistency.
FAQ
Do nonprofits need an llms.txt file for AI search?
No. Google states you do not need to create new machine readable files, AI text files, or markup such as llms.txt to appear in AI Overviews or AI Mode. If a vendor is charging you to maintain one, that fee is buying a file Google has said it does not require.
Does structured data help my nonprofit appear in AI Overviews?
Structured data is not required for generative AI features, per Google. It still helps you win traditional rich results and makes your organization's key facts unambiguous, so it is worth doing well, just not as an "AI-only" tactic.
Can I see whether AI search cites my nonprofit's site?
Yes, partially. Google Search Console's Generative AI performance report, launched in June 2026, shows impressions and clicks from AI features. It does not reveal the underlying queries, so use it as a directional signal rather than a full accounting.
Is paying an agency for GEO worth it?
It can be, if the work is real SEO and content strategy. Be skeptical of anyone selling llms.txt files, "AI schema," or guaranteed citations in a named assistant, because Google has said the first two are unnecessary and no one can guarantee the third.
Book a read-out
If you want a read-out of which of your pages AI search is already surfacing, and the three content fixes most likely to earn your nonprofit a citation, book a 20-minute GEO read-out and I will walk your Search Console report with you.