AI Visibility for Schools and Tutoring Centers

Parents rarely ask an assistant for "tutoring." They ask for a subject, a grade, a place and a time. The prompts below are hypothetical, but they show the shape of the demand:

1. "Find a math tutor near Westfield for a 10th grader who also needs SAT prep."
2. "Which tutoring centers in Austin offer small-group reading help for second graders?"
3. "How much does a private school in Denver cost per year, and do any offer sibling discounts?"
4. "Is there a tutoring center open on Saturday mornings near me?"
5. "Best private schools in Charlotte with a strong STEM program and a bus service."
6. "Does Bright Path Learning still have spots for fall enrollment?"
7. "Compare two Montessori schools in Portland for kindergarten: fees, class size, application deadline."

Every prompt carries a hard fact: a grade band, a fee, a day, a neighborhood, a deadline. If an engine can't find that fact on a page it can read, it skips you or fills the gap from another source.

One caution on testing. OpenAI says ChatGPT search rewrites a prompt into one or more targeted queries before it searches. Your prompt list is a sample of parent demand, not the exact input the engine runs. Start by writing down the ten questions your front desk hears most. Those become your test prompts.

Parents use several assistants, and each one draws on a different source. As of October 2026:

• Google AI Overviews and AI Mode. Google says a page must be indexed and eligible to show with a snippet to appear as a supporting link, with no additional technical requirements. Ahrefs found that 37.9% of URLs cited in AI Overviews also appeared in the first 10 results (March 2026), so ranking helps but doesn't decide it.
• ChatGPT. Its search feature relies on OAI-SearchBot. Many answers are generated without search at all, from training data.
• Microsoft Copilot. It grounds on the Bing index, so Bing presence matters separately from Google.
• Perplexity, Claude, Gemini and Meta AI. Each has its own crawler or retrieval path. Meta is the only operator that states allowing its crawler, Meta-WebIndexer, helps it cite your content.

These engines can disagree about the same school. Read how AI search differs from Google search before you treat one engine's answer as the whole picture. Record results per engine, never as one number.

Schools and tutoring centers change their core facts on a calendar. Tuition goes up each year, session times change each term, and a program fills up and opens a waitlist. AI answers lag behind for four reasons.

Training data is frozen

When an assistant answers without searching, it uses what it learned before its cutoff. OpenAI lists a February 16, 2026 knowledge cutoff for its GPT-5.6 models and August 31, 2025 for the Instant path used in ChatGPT. A fee you raised in September exists in neither. The search on vs off test shows how far the two answers diverge for your school.

Third-party pages hold old facts

Directories and listing sites often keep last year's fees or a previous address. An engine that cites them repeats the error.

Your own site contradicts itself

A blog post with "2024-25 rates," a PDF brochure, and a pricing page can state three different numbers. The engine has no way to know which one is current.

Key facts load after JavaScript runs

Schedule widgets and booking embeds often load their content with JavaScript. Most AI crawlers never see it.

The facts most exposed to drift for this industry:

• Tuition, hourly rates, packages and registration fees
• Grade levels served (for example, a center that added middle school)
• Subjects and test prep offered
• Session times, term dates and application deadlines
• Enrollment status: open, waitlist or full
• Locations, especially a closed or moved branch
• In-person, online or both
• Confusion with a similarly named school nearby

The one on-page change with positive experimental support is simple. A specific fact, stated on a page an engine can fetch, gives it a sentence it can quote.

Say you run a tutoring center (hypothetical). Instead of "affordable packages for every family," write: "Weekly one-to-one math tutoring for grades 3 to 8 costs $65 per hour for the 2026-27 school year." That sentence answers prompts 1, 3 and 7 directly.

Apply it across the site:

• Put fees, grades, subjects, hours and addresses in page text, not only in images or PDFs. Keep the brochure if parents want it, but repeat the key numbers on the page.
• Date every fee and schedule ("for fall 2026"), so a reader and an engine can tell it's current.
• State enrollment status in one line on the program page, and update it when it changes.
• Give each location its own page with address, hours and the programs offered there.
• Delete or update old posts that quote past rates.
• Make sure the content is in the HTML before JavaScript runs.

Skip the shortcuts. Google says you don't need special markup, AI text files or a special writing style for its AI features. Ahrefs' matched test found AI Overviews citations fell 4.6% after pages added schema. Existing markup does no harm, but it isn't a lever for AI answers. The same plain-facts approach works for dental clinics and other local services.

When engines cite other sites instead of yours, being named correctly on those sites is the route into the answer. That's an association in the evidence, not a guaranteed effect, but these are the sources parents' answers tend to draw on:

• Map and local business listings for each location
• Local directories and your chamber of commerce listing
• Education directories and private school finders
• State or district lists of approved providers, where they apply to you
• Review sites parents use
• Tutoring marketplaces where your tutors hold profiles
• Local news and parent community pages that mention you

Say a reading center in Columbus moved across town last spring (hypothetical). Its new address is on its website, but three directories still show the old one. An engine citing those directories sends parents to an empty storefront.

To find which sites matter, run your test prompts and note every URL each engine cites. Fix those listings first. The competitor check also shows which rival schools AI recommends and where it found them.

Clear facts don't help if crawlers can't reach them. These checks remove blocks, and they're fully in your control.

• robots.txt, per bot. Rules apply to each crawler separately. GPTBot governs training only, and OAI-SearchBot controls ChatGPT search. Read whether to block GPTBot before you copy a blanket AI block.
• Firewall and CDN. OpenAI says eligibility needs both robots.txt permission and a host or CDN that allows its search bot's IP addresses. A firewall can return 403 to a bot that robots.txt allows. Cloudflare changed its default AI crawler settings for some sites on September 15, 2026, so check yours even if you edited nothing. 99Visibility's AI Crawler Checker requests a page as each AI crawler and shows which ones get turned away.
• JavaScript rendering. Vercel found ChatGPT and Claude crawlers fetch JavaScript files (11.50% and 23.84% of requests) but don't execute them. Googlebot does render.
• Snippet controls. noindex, nosnippet, max-snippet:0 and data-nosnippet keep content out of AI answers. Templates sometimes apply them by accident.
• Index status. Check Google Search Console and Bing Webmaster Tools separately. One being healthy says nothing about the other.

A minimal robots.txt that allows the main search bots looks like this:

User-agent: OAI-SearchBot
Allow: /

User-agent: PerplexityBot
Allow: /

User-agent: Claude-SearchBot
Allow: /

Days 1 to 7: baseline

Write your ten parent prompts. Run up to three of them through the free checker on this site, then run the rest by hand in each assistant. Record, per engine, whether you're named, which facts are wrong, and which URLs are cited. Note that API answers and consumer app answers can differ.

Days 8 to 14: access

Audit robots.txt per bot, test your firewall and CDN, compare raw HTML with the rendered page, and search templates for snippet controls. Confirm index status in Google and Bing.

Days 15 to 21: your own pages

Publish one dated, plain-text statement of fees, grades, subjects, hours and enrollment status per program and location. Remove contradictions in old posts and brochures.

Days 22 to 30: listings and recheck

Correct the directories and listings engines cited. Rerun the same prompts and compare with your baseline.

Set expectations honestly. Search-on answers can pick up fixes once pages are recrawled. Training-data answers won't change until the next model release, and no one can promise a citation.

Does ChatGPT know our current tuition? Only if it searches. Without search, ChatGPT answers from training data with a fixed cutoff, and OpenAI lists August 31, 2025 for the Instant path. A fee set after that date isn't in the model. With search on, it can find your current fee if OAI-SearchBot can reach your page and the number is stated in plain text. Test both states and fix what search-on answers get wrong.

Should we add schema markup so AI shows our programs? Not for that reason. Google says structured data isn't required for its generative AI features, and Ahrefs' matched test found AI Overviews citations fell 4.6% after pages added schema. Existing markup does no harm, so there's no need to remove it. Spend the time stating fees, grades and schedules in clear, dated sentences on pages crawlers can fetch.

Can Google Analytics show parents arriving from AI answers? Partly. ChatGPT adds utm_source=chatgpt.com to many referral links, but not all clicks arrive tagged. Google AI Overviews send no separate referral at all. Search Console's generative AI report shows impressions only, with no clicks or position. Treat analytics as a floor on AI traffic, and use prompt tests to see what engines actually say about your school.

Should a school block AI crawlers to protect privacy? Private student information shouldn't be on public pages at all, so fix that at the source. Blocking search bots such as OAI-SearchBot or PerplexityBot removes you from those engines' answers. If you want to limit training use, rules are per bot: you can disallow GPTBot and still allow OAI-SearchBot. For a specific section, data-nosnippet keeps it out of Google and Bing AI summaries.

How often should we recheck AI answers? Recheck after every change parents care about: a new fee schedule, new term dates, a program that fills, or a location that opens or closes. A prompt check is a snapshot, not continuous tracking, and answers vary between runs. Rerunning the same prompts at the start of each term gives you a consistent comparison.