A board member asked me last month how people were finding her organization. She meant Google. I told her the more useful question now is what an AI says when somebody asks it where to give, or where to work, or who does this kind of work in Cleveland. She had not thought about it that way, and most nonprofits have not.
Here is the change in one sentence. People used to search, get ten links, and click one. Now a growing share of them ask a question and read an answer. If your organization is not in the answer, there is no second chance, because there is no list of ten anymore.
Three different disciplines decide whether you show up, and they are not the same work. Most people know the first one, half have heard of the second, and almost nobody in our sector is doing the third.
SEO: can a machine read your site at all
This is the old discipline and it still comes first. Nothing else matters if a crawler cannot reach your pages and understand them.
The short list: your important pages are crawlable and indexable, your site has a clear structure with real internal links, each page matches a real thing somebody is looking for, the writing is original and comes from people who actually do the work, you have depth across related subjects rather than one page per topic, your titles and metadata are accurate, and the site is fast and usable on a phone.
That last one is not a technicality. I have watched good organizations lose candidates because the application form did not work on a phone. Structured data helps too, where it genuinely applies, and a job posting is one of the places it applies most.
AEO: can a machine extract an answer from your page
Answer engine optimization is about being quotable. A model reading your page is looking for a clean answer to a specific question, and it is going to take whatever is easiest to lift.
So answer the main question immediately, in the first paragraph, before the history of the organization. Use headings that are the questions people actually ask. Keep the important facts in short extractable sentences rather than buried in a long paragraph. Break hard ideas into scannable blocks. Use comparisons, steps, tables and a real FAQ. Define your terms, because a model cannot infer what your organization means by “wraparound services” if nobody on the page says it. Cover the obvious follow-up question in the same place. Keep every claim current and specific enough to check.
Most nonprofit About pages fail every one of these. They open with a mission statement written by committee in 2011 and never say plainly what the organization does, for whom, where.
GEO: will a generative engine cite you
This is the one almost nobody in our sector is working on, and it is the one with the most room in it.
Generative engines cite sources. What gets cited is original research and firsthand knowledge, evidence and examples, and a point of view that exists nowhere else. Comparison and decision-stage content gets pulled constantly, because that is what people ask about. Mentions from authoritative third parties matter, as does showing up in the communities and platforms where the subject gets discussed. Keep the valuable material fresh and accurate, and make your evidence easy to verify.
Nonprofits are sitting on exactly this and publishing almost none of it. You have program data nobody else has. You know what the actual cost per outcome is in your city. You know what the waitlist looks like. That is original research. A blog post about Giving Tuesday is not.
What this has to do with hiring
This is where it gets concrete for anyone running a search, and it is the part I care most about after 29 years of this.
Your job postings are content. A candidate increasingly asks an AI what a development director at a mid-size arts organization does, what it pays, and who is hiring. Your posting is either legible to that machine or it is invisible.
Most nonprofit postings are invisible, and it is not a technology problem. It is that they are written in a code that only people already inside the sector can read. “Dynamic leader sought to steward our mission” tells a machine nothing and tells a career changer less. A posting that says what the person will own, what the budget is, who reports to them, where the office is, and what it pays is legible to a human, a crawler and a model, all at once.
The same goes for titles. We see the same job posted as Director of Development, Chief Advancement Officer, and Philanthropy Lead. Pick the one people search for and put the alternatives in the body.
What to measure
The old scoreboard does not cover this. Alongside your rankings and traffic, start watching whether AI tools mention you at all, whether they cite you, which of your pages get cited, how often your brand comes up unprompted, how much of the conversation in your subject area you hold, what prompts you show up for, what prompts you are missing, how much of your subject you actually cover, and whether what gets said about you is accurate.
That last one is worth a line on its own. Ask an AI about your own organization. Several nonprofits I know have found out that the answer was two years out of date, or confidently wrong about their programs, and they only found out because somebody finally asked.
Where to start if you do one thing
Rewrite your About page so the first paragraph says what you do, for whom, where, and at what scale. No mission statement, no history, no founder story above the fold. Then do the same to your three most important program pages, and to every job you post.
It is unglamorous and it takes an afternoon. It will do more than anything else on this list, because all three disciplines reward the same thing: saying clearly what is true.
