• AI visibility

Guide · AI search

Can AI search bring an EdTech company more schools, teachers and families?

Partly, and the path differs for each buyer. Teachers, students and families already ask AI tools for help, so being named in those answers can start adoption. Districts and universities still buy through committees, evidence reviews, privacy checks and, more and more, state approval lists, so an AI answer only helps if it can find that proof and repeat it accurately.

By Underneath Team Guide Updated

The short version

  1. Districts are drowning in tools: they have access to an average of 3,001 digital tools, yet students and teachers typically use about four learning apps inside their learning platform, according to Instructure data reported by EdWeek Market Brief (opens in a new tab).
  2. The people who use EdTech already use AI: six in 10 US public school teachers used an AI tool for their work in 2024-25 (Gallup and the Walton Family Foundation (opens in a new tab)), and 42% of college students use generative AI weekly (Tyton Partners (opens in a new tab)).
  3. State lists now steer curriculum buying: at least 18 states publish approved materials lists, and 76% of school and district leaders say those lists influence elementary reading purchases “a lot” (EdWeek Market Brief (opens in a new tab)).
  4. Guardrails are tightening: 79% of districts now have AI guidelines, up from 57% in 2025 (CoSN (opens in a new tab)), and New York City banned student-facing generative AI through eighth grade for one year.
  5. AI search can also take traffic away: the homework help company Chegg sued Google in 2025, claiming AI Overviews hurt its traffic and revenue (The Verge (opens in a new tab)).

Who buys education software, and how?

Four different buyers: districts, teachers, colleges and families, each with its own path to a purchase.

K-12 districts. The US has 98,577 public schools, according to federal data (opens in a new tab), grouped into districts that buy through budgets, requests for proposals and board approval. Influence is spread out. In an EdWeek Research Center survey on who influences purchases (opens in a new tab), superintendents were named by 63% of district and school leaders and school boards by 46%; 31% pointed to finance leaders and 29% to technology leaders. That survey was about security purchases, but the chain it describes, from vetting to the business office to the board, is typical of large district buys.

Teachers. Teachers often find a tool first and bring it to the district. Their influence shows in usage: across 12.6 million K-12 users of the Canvas platform, Instructure counted 618,000 educators launching learning apps in 2025-26.

Colleges and universities. Instructors and administrators choose course tools and platforms. Tyton’s 2025 study drew on more than 3,300 students, instructors and administrators across more than 900 colleges.

Families and students. Parents and learners buy subscriptions directly: tutoring, test prep, language learning, homework help. This is the closest EdTech gets to consumer marketing, and the most exposed to AI answers.

What a customer is worth varies widely, and public benchmarks are thin. What the data do show is pressure on budgets. The public school population fell 2% over the past decade, and federal projections suggest a possible 5% drop by 2031, as EdWeek Market Brief reports (opens in a new tab). In its survey, 55% of district and school leaders said enrollment had dropped in the past five years. Fewer students means fewer seats to sell and harder questions at renewal.

Where do teachers, students and district buyers already use AI?

With the end users first: teachers, students and families use AI daily; district buyers’ use for vendor research is unmeasured.

  • Teachers. In the Gallup and Walton survey of 2,232 public school teachers, 60% used AI tools for their work, and 32% did so at least weekly.
  • Teens. Pew Research Center (opens in a new tab) found 26% of US teens had used ChatGPT for schoolwork, up from 13% in 2023, and 54% said it is acceptable to use ChatGPT to research new topics.
  • Colleges. Tyton found weekly generative AI use among 42% of students and 30% of instructors.
  • District technology leaders. In CoSN’s 2026 survey (opens in a new tab), 96% of education technology leaders said AI could positively affect education, and 64% of districts were using AI in operations, up from 37% the year before.

None of these surveys asks whether district buyers use ChatGPT, Gemini or Google’s AI features to find vendors. For vendor research, the nearest figure comes from outside education: among software buyers of every kind in G2’s 2026 survey (opens in a new tab), 51% started research with an AI chatbot more often than with Google. G2 sells review visibility and did not survey schools.

We infer that AI answers matter most at two points: when a teacher, parent or student looks for a tool to solve a problem today, and when a district team does its first pass over a crowded category before writing a request for proposals.

Which questions do educators, parents and students ask?

Questions about fit, evidence, privacy, price and alternatives. We wrote the classroom and household prompts below ourselves to show the pattern; none was captured from a real teacher, parent or student.

BuyerIllustrative prompt
Teacher“What is a free tool to give feedback on student writing that works in Google Classroom?”
Curriculum director“Which K-5 reading programs are on the Texas approved list and align with the science of reading?”
District technology leader“Which math platforms have ESSA evidence and a signed student data privacy agreement in our state?”
University administrator“What are alternatives to our current proctoring software that faculty and students accept?”
Parent“Is this online tutoring service worth it for a 7th grader struggling with algebra?”
Student“What’s the best app to practice Spanish for the AP exam?”

These questions mix discovery with trust. A parent wants to know whether a tutoring service is safe and worth the money; a district wants proof of impact and a privacy agreement before anything else.

A parent’s single question about a tutoring service can set off several searches behind the answer. For AI Overviews and AI Mode, Google describes a “query fan-out” technique (opens in a new tab) that may issue multiple related searches; on OpenAI’s side, the help center says ChatGPT search typically rewrites a question (opens in a new tab) into one or more targeted queries. Those hidden searches lean on outside judgment, which is where an EdTech vendor’s reviews and awards come in: in our hidden-searches study, 46.2% of ChatGPT’s answers involved a search for reviews or ratings, and 43.8% targeted a named publication, ranking or award.

How does being named by AI become subscriptions, classroom adoptions and district contracts?

By different routes for each buyer: subscriptions for families, teacher adoption for schools, and shortlists for district contracts.

  1. Families and learners. An AI answer names a tutoring service or app; the parent or student tries it; a subscription follows. This is the shortest path, and the one most like consumer software.
  2. Teacher-led adoption. A teacher asks for a tool, tries the free version, and uses it with a class. If enough teachers do the same, the district buys a license. Because students and teachers use only about four apps in their learning platform, the prize is becoming one of those few.
  3. District and state contracts. A curriculum or technology team narrows the field, checks state approval lists, evidence and privacy agreements, then issues a request for proposals or buys from an approved list. An AI answer can put a vendor on the first list; the evidence decides the rest.
  4. Higher education. Faculty and administrators compare options; the institution signs a campus license.

The more formal the purchase, the less the AI answer decides on its own. State approval matters most in core subjects: 62% of leaders said state lists influence elementary math purchases a lot, against 59% for secondary reading, while just 17% said the lists don’t influence secondary reading decisions at all.

Why would an assistant recommend one learning tool over another?

No platform documents how it chooses products; studies point to independent sources, and education buyers reward evidence and privacy.

What Google and OpenAI say. Both companies describe AI answers that search the web and cite sources; neither explains how a tutoring app or classroom tool gets picked.

What research has measured. Earned sites such as reviews and independent publications supplied 72.7% of AI search sources for US software questions, Chen and colleagues (opens in a new tab) found. Repeat runs change the list, too: across five runs of the same question in our consistency study, only 25.2% of the brands ChatGPT named appeared every time. Neither study covered schools, colleges or family learning products.

What education buyers check. These trust factors are specific to the sector and are public, which matters because an assistant can read them too:

  • Evidence of impact. More than half of the tools in Instructure’s 2025-26 top 40 had documented proof of impact, which Instructure says is 13 points higher than the broader EdTech market. EdWeek Market Brief adds that districts are “demanding stronger evidence behind product claims.”
  • Privacy credentials. Nearly half of the top 40 tools hold third-party privacy certifications from groups such as 1EdTech or iKeepSafe. The Federal Trade Commission’s (opens in a new tab) children’s privacy rule requires verifiable parental consent before collecting personal information from children under 13, and was updated in 2025.
  • Security track record. On December 28, 2024, PowerSchool, a major student information system vendor, learned that personal data had been taken from some customers’ systems through a support portal accessed with a single compromised credential (opens in a new tab). CoSN found 65% of districts name insufficient cybersecurity staffing and the lack of a dedicated budget as top barriers. A breach becomes part of what any search, human or AI, finds about a vendor.
  • Policy fit. Seven states passed laws or directives on devices and digital tools in classrooms this year, EdWeek Market Brief reports (opens in a new tab), and New York City’s ban on student-facing generative AI affects nearly 600,000 students.

Our inference. The proof districts verify (evidence ratings, privacy agreements, state approvals, accessibility statements) is published by third parties and is easy to cite. A reasonable expectation is that vendors whose proof is public, consistent and current give AI answers more to work with. No published study has tested it for school or family learning products.

What does it cost an EdTech company to be missing, or to be summarized?

Lost trials and shortlist places, and for content-heavy businesses, lost traffic.

  • The first list in a crowded market. With districts able to reach about 3,001 tools, a product that is not named early may never be compared.
  • Teacher discovery. If teachers ask an assistant for a tool and a competitor is named, the bottom-up route to a district license narrows, we infer.
  • Content businesses can lose the click. Chegg sued Google in February 2025, claiming AI Overviews used its content and hurt its traffic and revenue; Google said AI Overviews send traffic to a greater diversity of sites. The case shows the risk for EdTech companies whose growth depends on free answers to homework questions. We look at this risk in which content businesses AI search threatens and whether AI Overviews reduce clicks.
  • Wrong descriptions. An outdated privacy or pricing description in an AI answer can stop a district review early, we infer. Our guide to fixing wrong brand information in AI answers covers how to correct a stale description.

What does GEO involve for a learning-product company?

It makes your evidence, privacy and fit easy for AI assistants to find and repeat, for each buyer. It cannot promise that a teacher, parent or district will see you named.

  1. One clear identity per audience. Say what the product does, for which grades or courses, and which platforms it works in, the same way on your site, app marketplace listings and review profiles.
  2. Public proof pages. Publish ESSA evidence and research summaries, privacy certifications, signed state data privacy agreements, accessibility statements and state approvals on plain web pages, not only in sales packets.
  3. Independent coverage. Earn mentions in education trade press, research clearinghouses and educator communities. The outside-coverage work is the same as in other sectors, described in how brands build authority for AI search.
  4. Answer real questions honestly. Write comparison, alternatives and “is it worth it” pages for parents, teachers and administrators, with real prices and limits. An independent “best tutoring apps” list tends to count for more than your own site, as our guide to which pages to target shows.
  5. Reviews from real users. Encourage teachers and families to review the product where they already look. Never plant reviews; see legitimate GEO versus manipulation.
  6. Measure each audience separately. Teacher, parent and district questions draw different answers. Track ChatGPT, Google AI Overviews and AI Mode, Gemini, Perplexity, Copilot and Claude; see how many prompts to track.

What don’t we know yet about AI search in education buying?

We cannot yet say how often districts pick vendors through AI, or whether it lifts EdTech sales.

  • No survey of buyers’ AI search use in education. The surveys above measure AI use for teaching and learning, not for choosing products.
  • Some sources are vendors. Instructure sells a learning platform; G2 sells review visibility.
  • The Chegg case is unresolved. A lawsuit is a claim, not a finding.
  • No link to seats or subscriptions has been shown. Whether being named turns into EdTech revenue is still open; see does AI visibility drive business results.

What should an EdTech company check before the next adoption and renewal cycle?

Check how each assistant describes you to teachers, parents and districts, then publish the proof each buyer needs.

Draft the questions a teacher, parent, student, district leader and university buyer would ask, and put each one to ChatGPT, Gemini, Google’s AI features and the other main assistants more than once. Note whether your product is named, how your evidence, privacy terms and price are described, and which sources are cited. Where answers fall short, the cause is usually proof that lives only in sales packets, or a page that has gone stale.

We can run that check with you: ask us to review how AI answers present you to schools and families. We will show where AI answers put you in front of teachers, families and district teams, and which gaps in your evidence and privacy proof are most likely costing you trials, adoptions and contracts. For the ongoing side, our generative engine optimization service page describes how proof pages, educator reviews and separate checks for teacher, parent and district questions are kept current.

Frequently asked questions

Do teachers use ChatGPT to find classroom tools?

Many teachers use AI for their work: 60% did in 2024-25. No survey measures how often they use it to find products.

Does being on a state approved list help AI visibility?

Untested for AI, but it matters to buyers: 76% of leaders say state lists influence elementary reading purchases a lot.

Should we publish our privacy agreements and certifications?

Yes, at least the facts districts check before they sign. Nearly half of the top 40 tools in Instructure’s data hold third-party privacy certifications.

Will AI search hurt our free content traffic?

It can. Chegg claims AI Overviews hurt its traffic; plan for answers that summarize your content without a click.

Is GEO different for consumer and school products?

Yes. Families decide quickly from answers and reviews; districts need evidence, privacy agreements and approvals before a contract.

Sources

Free strategy call

Some questions are easier to answer about your own business.

Bring the one that matters most. On a free 30-minute call we’ll take a first look at it and send you a short written read afterward.

Ask about my businessHow we work