How to Check Whether AI Search Engines Can See Your Brand (Free, 20 Minutes)

Disclosure: I'm the maker of Quotalis, an AI search visibility monitor — one of the tools that does this kind of checking automatically. Below is the manual version anyone can run for free, and I'll be honest about what it can and can't do.

Last week I asked four AI engines about my own brand 168 times. They mentioned me zero times. (Here's the data.) That experiment was the starting point of a question every brand owner should be asking: when people ask ChatGPT, Claude, Gemini, or Perplexity about your space, does your name show up?

Here's how to answer that question yourself, with no tools and no account.

Step 1 — Write your question set (10 minutes)

Don't ask "tell me about {brand}". That's a vanity query. Ask what your customers actually ask:

• Recommendation queries (3–4): "Which tools monitor brand visibility in AI search?" / "best {your category} tools for small teams"
• How-to queries (3–4): "how to track brand mentions in ChatGPT" — anything your product claims to solve
• Comparison queries (2–3): "{your category} tools compared"

Mix languages if your audience is multilingual. I ran 7 English + 7 Chinese questions.

Step 2 — Run each question in each engine (5 minutes with free tiers)

Open ChatGPT, Claude, Gemini, and Perplexity (free tiers all work) and paste the same questions. One engine is not enough — engines disagree wildly about the same topic.

Step 3 — Score two things separately (5 minutes)

This is where most DIY attempts go wrong. Track mentions and citations as separate columns:

• Mention = your exact brand name (or handle) appears in the answer text. Watch out for partial matches: "Wu" is not "wu的智脑团". Be strict, or you'll inflate your numbers the way I almost did.
• Citation = a linked source points to your page. x.com/somebody-else doesn't count as citing you just because it's the same domain. Check the full path.

My score after 168 runs: 0 mentions, 0 citations across all four engines. A clean zero is actually useful information — it means you're starting from a true baseline, and any future movement is real signal.

Step 4 — Read what the engines cite instead

The most valuable column in my spreadsheet wasn't my own score — it was the list of domains the engines did cite for my topic: tech media, dev communities, LinkedIn posts, competitor product pages. That's your distribution map. If you want AI to cite you, you need to exist where those engines already look.

The honest limits of the manual method

Doing this by hand once gives you a snapshot. Three problems appear within a week:

1. Drift — engine answers change as indexes update; a one-time check goes stale fast.
2. Coverage — I ran 3 repetitions per question because single runs vary. Doing that manually every week gets old.
3. Scale — 14 questions × 4 engines × 3 runs = 168 queries. It's a lot of copy-pasting.

That's why I built Quotalis to run the same battery automatically on a weekly schedule — but you shouldn't pay for something before you know the manual version works for you.

The free 20-minute shortcut

If you want the site-side half of this without running 168 queries: Quotalis has a free AI-readiness audit — no login, no signup. It fetches your site's public files (llms.txt, robots.txt, sitemap.xml) the same way AI crawlers do and tells you what's blocking them. It's not a substitute for the mention check above — it's the other half: making sure your site is legible to the engines once they do come looking.

Run the manual battery, fix what the audit flags, then re-run the battery in a month. The difference between the two numbers is your GEO progress, measured honestly.

---

I write about AI search visibility experiments like this one. Follow for the next round of data.