
Is Your Brand Invisible to ChatGPT? How to Check
You can check whether ChatGPT, Perplexity, Gemini, and Google's AI Overviews mention your brand in about 30 minutes, for free. Here's the exact method.
By Team COACT
Here's a test worth running before you spend anything on "AI visibility": open ChatGPT, Perplexity, and Google, ask the questions your customers ask, and see whether your brand appears in the answer. It takes about half an hour, it costs nothing, and it tells you the one thing the dashboards charge for — whether the AI answer layer knows you exist.
Most brands have never actually looked. This post is the exact method: what to ask, what to record, how to read the result, and an honest tour of the paid tools that automate it (short version: they exist, they're useful, and none of them discloses a rigorous methodology — so start with the free check).
Key Takeaways
- You can audit your AI visibility manually in ~30 minutes across ChatGPT, Perplexity, Gemini, and Google's AI Overviews — no tool required.
- Record three things per query: are you mentioned, is the mention accurate/positive, and which page (if any) gets cited.
- Paid AI-visibility trackers exist (Profound, Peec, Semrush's AI tools), but none publishes a transparent methodology — treat their scores as directional.
- Being absent from AI answers is a content-and-crawlability problem, not bad luck. The fix is the same specific, sourced, citable content that wins AI citations generally.
The 30-Minute Manual Audit
You don't need software to answer "does AI know my brand." You need a list of real questions and a place to log the results. Here's the whole procedure.
Step 1 — Write 8–12 real customer questions. Not your brand name — the problems people search before they've heard of you. For a D2C skincare brand: "best affordable sunscreen for oily skin India," "is mineral sunscreen better than chemical." For a fintech: "cheapest way to send money Singapore to India." These are the prompts where you either appear or you don't.
Step 2 — Run each query in four places and read the answer, not just the links: ChatGPT (with search on), Perplexity, Google Gemini, and a plain Google search (to catch the AI Overview). Do this in a logged-out or incognito session so your own history doesn't skew it.
Step 3 — Record three things per query in a simple sheet:
| Column | What to log |
|---|---|
| Mentioned? | Does your brand appear in the answer at all — yes / no |
| Sentiment | If mentioned: is it accurate and positive, or wrong/negative |
| Cited page | Which of your URLs (if any) is linked as a source |
Step 4 — Repeat monthly. AI answers are volatile; a single snapshot is noise. The trend across months is the signal. Run the same query set each time so the comparison is clean.
That's the audit. The output is a brutally clear scorecard: on the questions that matter to your business, how often are you in the answer, is what's said about you right, and which of your pages the engines actually trust.
How to Read Your Results
Four patterns show up, and each points to a different fix.
| What you find | What it means | The fix |
|---|---|---|
| Not mentioned anywhere | The engines don't have you as a source for these topics | Publish specific, sourced content on exactly these questions |
| Mentioned, wrong details | The engines have stale or incorrect info about you | Correct it at the source — your site, and the third-party pages they cite |
| Competitors cited, not you | Their content is more citable on your topic | Study what they get quoted for; out-specific them |
| Cited, but a weak page | You're in the answer via the wrong URL | Strengthen the page that should be representing you |
The most common result for brands that have never optimised for this is the first row — simply absent. That's not a mysterious algorithmic snub. It usually means one of two concrete things: your content isn't specific and sourced enough to be worth citing, or the engines' crawlers can't access it. Both are fixable, and both are the same work that wins AI citations in general.
The Paid Tools: What Exists, and What to Trust
You can automate the audit above, and a small industry has appeared to do it. Being honest about these matters, because their marketing runs ahead of their transparency.
Tools like Profound, Peec AI, and the AI-visibility features now bundled into suites like Semrush will run prompt sets across multiple engines on a schedule and report how often you're mentioned, your "share of voice" versus competitors, and sentiment. That's genuinely useful — it's the manual audit, automated and tracked over time.
Here's the part their landing pages won't tell you: none of these tools discloses a rigorous, reproducible methodology. How many prompts, chosen how, run against which engine versions, on what schedule, scored by what definition of a "mention" — the details that would let you trust a number are not published. AI answers also vary run-to-run for the same prompt, so any single "visibility score" is a sample, not a measurement. That doesn't make the tools worthless; it makes their numbers directional. Use them to watch a trend line over months, not to trust a decimal. And know that the free 30-minute audit above is, in one respect, more transparent than any of them — because you can see exactly what was asked and what came back.
Why Brands End Up Invisible — and How the Engines Decide
If your audit comes back mostly blank, it's worth understanding the mechanism, because it tells you where to push.
The AI engines split their crawling into distinct jobs, and you can see this in how OpenAI documents its bots: OAI-SearchBot surfaces and links your site in ChatGPT's search answers, while GPTBot is the crawler used for training — and OpenAI states the two settings are independent, controlled separately in robots.txt. Google does the same: normal Googlebot handles Search, while Google-Extended is a separate control for whether your content helps train and ground its generative models. The practical implication: if you (or an over-zealous security setting) have blocked these crawlers, you can be perfectly visible in classic search and invisible to the AI answer layer — and never know it until you run the audit.
When the engines can read you, what gets you cited is what gets everyone cited: specific, attributed, sourced content. The single peer-reviewed study on generative visibility found that adding statistics and quotations were the biggest levers, and that keyword stuffing actively hurt. So "how do I stop being invisible" has a boring, evidence-based answer — publish the citable thing, and let the crawlers reach it.
What to Do With a Bad Result
A concrete sequence once your audit shows you're absent or misrepresented:
- Confirm the crawlers can reach you. Check your
robots.txtisn't blocking OAI-SearchBot, GPTBot, PerplexityBot, or Google-Extended if you want to appear in AI answers. Being uncitable-by-accident is the most fixable failure of all. - Publish specific, sourced content on the exact questions from your audit. Not keyword pages — answers, with real numbers and named sources, on the problems your customers actually ask.
- Correct wrong information at its source, including the third-party pages the engines cite about you.
- Re-run the audit monthly and watch the trend, not any single snapshot.
That's also, not coincidentally, the discipline behind everything we publish — our breakdown of what a good ROAS actually is leads with a specific, sourced number for exactly this reason.
Frequently Asked Questions
How do I check if ChatGPT mentions my brand?
Open ChatGPT with search enabled and ask 8–12 real customer questions — the problems people search before they know you, not your brand name. Record whether your brand appears in the answer, whether the mention is accurate, and which of your pages (if any) gets cited. Repeat across Perplexity, Gemini, and Google's AI Overview, ideally in an incognito session, and re-run monthly since results vary over time.
Is there a free way to audit AI visibility?
Yes — the manual method costs nothing and takes about 30 minutes: run your customer questions across ChatGPT, Perplexity, Gemini, and Google, and log mention, sentiment, and cited page for each. Paid tools automate this over time, but the free audit is more transparent because you can see exactly what was asked and returned.
Why isn't my brand showing up in AI answers?
Usually one of two concrete reasons: your content isn't specific and sourced enough to be worth citing, or the AI crawlers can't reach it. Some brands block crawlers like GPTBot or Google-Extended without realising it, which leaves them visible in classic search but invisible in the AI answer layer. Both causes are fixable.
Are AI-visibility tools like Profound and Peec accurate?
They're useful for tracking a trend, but treat their scores as directional, not precise. None publishes a rigorous, reproducible methodology — the prompt sets, engine versions, and scoring definitions that would let you trust an exact number aren't disclosed, and AI answers vary run to run. Watch the trend line over months rather than trusting a single visibility score.
How often should I check my AI search visibility?
Monthly. A single snapshot is noise because AI answers change frequently for the same prompt. Running the same fixed set of customer questions each month turns that noise into a trend you can actually act on.
How do I get my brand cited by ChatGPT and Perplexity?
Make your content the obvious thing to cite: specific statistics, attributed quotes, and named sources on the exact questions your customers ask — the tactics the one peer-reviewed study on generative visibility found most effective. Then make sure the AI crawlers can reach those pages. There's no trick beyond publishing genuinely citable content and letting it be crawled.
Conclusion
The uncomfortable thing about AI visibility is how easy it is to check and how rarely anyone does. Half an hour, four engines, a dozen real questions, three columns in a sheet — and you'll know whether the answer layer that's quietly intercepting your traffic even knows you exist. If it doesn't, the fix isn't a tool subscription; it's citable content the crawlers can reach. Run the audit first. You can't optimise for a problem you haven't measured.
How this post was compiled. The manual audit method is a procedure, not a claim, and needs no external data. The crawler mechanics — OAI-SearchBot vs GPTBot, and Google-Extended — are verified against OpenAI's and Google's own bot documentation. The named tools (Profound, Peec, Semrush) were confirmed to exist; we state plainly that none publishes a rigorous methodology rather than repeating their marketing figures. The generative-visibility tactics reference the one peer-reviewed study in the field (Aggarwal et al., arXiv:2311.09735). Coact is a performance marketing agency working with ecommerce and app businesses across Singapore, India, and Indonesia.
Continue Learning
- What is AEO? Answer Engine Optimization Explained (publishing soon) — the plain-language primer
- AEO & GEO in 2026: A Guide for Southeast Asian Brands (publishing soon) — the full playbook
- SEA & India Growth Benchmarks: What's Real — the same audit-then-act discipline, applied to regional data
- Talk to our team about auditing and fixing your AI search visibility.
Want results like these for your brand?
COACT runs performance-based growth across SEA & India. Start with a free 15-day pilot.
Book a discovery callKeep reading
Attribution & Measurement: What Actually Tells You What's Working
Across 2,226 Meta experiments, last-click attribution explained just 19% of real incrementality. Here's how to measure marketing that actually holds up.
How to Optimise Your Site for AI Search in 2026
Google says there are no special requirements to appear in AI Overviews. So what actually works? Three layers — content, structure, and crawler access — sourced.