TL;DR: What an AI visibility audit checks, how to run one free in 30 minutes, and what real results look like — from the Singapore AEO agency that audits its own site. ChatGPT, Perplexity, Google AI Overviews and Claude.
What Is an AI Visibility Audit?
An AI visibility audit is a structured test of whether AI engines — ChatGPT, Perplexity, Google AI Overviews and Claude — cite or recommend your business when buyers ask questions you should be the answer to. Instead of tracking keyword rankings, the audit runs real buyer queries ("best custom cake delivery in Tampines", "accounting firm for Singapore startups") across each engine, records whether you are cited, mentioned or invisible, and diagnoses why — so you know exactly which gaps to fix first. In Singapore, LOOP Marketing Group runs this audit free of charge: no email wall for the quick scan, no card, and a full report across the major AI engines if you want one.
Why Singapore Businesses Are Running These Audits Now
The buyer journey has split. Alongside Google's blue links, a growing share of research now happens as direct questions to AI assistants — and IMDA's data shows GenAI adoption among Singapore SMEs tripled from 4.2% to 14.5% in a single year. When someone asks ChatGPT for a recommendation, they get an answer, not a list of links. If your business is not in that answer, you are not in that conversation at all.
The uncomfortable part: most Singapore SMEs who run the test for the first time land on invisible for the majority of their money queries. We know because we faced the same result on our own site before we started fixing it. An audit does not solve the problem — but it is the only honest starting point, because it tells you where you actually stand rather than where you assume you stand.
What a Proper AI Visibility Audit Checks
A real audit goes beyond "did my name come up". It covers four layers:
| Layer | What gets checked | Why it matters |
|---|---|---|
| Citation results | Whether each engine cites, mentions or omits you for 5–10 real buyer queries | This is the score — the ground truth of your AI visibility today |
| Crawler access | Whether GPTBot, PerplexityBot, ClaudeBot and Google-Extended can read your site (robots.txt, site speed) | Blocked crawlers mean AI engines literally cannot read your content |
| Content structure | Direct-answer formatting, FAQ coverage, schema markup (Organization, FAQ, Article) | AI engines extract passages; unstructured content is hard to cite |
| Entity and authority signals | Google Business Profile completeness, directory consistency, third-party mentions, reviews | Engines triangulate — if only your own site mentions you, that is a weak signal |
How to Audit Your AI Visibility Yourself (Free, ~30 Minutes)
You can run a basic version manually — we published the full walkthrough in our step-by-step ChatGPT visibility guide. The short version:
- Write 5–10 buyer queries. Think like someone who does not know your brand: "best [service] in [area]", "how much does [service] cost in Singapore", "[service] vs [alternative]". Never test your own company name — of course the AI knows you when you name yourself.
- Run each query on four platforms: ChatGPT, Perplexity, Google (check the AI Overview) and Microsoft Copilot. Different training data means different results.
- Score every result as Cited (named and recommended), Mentioned (appears but not the pick) or Invisible.
- Check crawler access: open yourdomain.com/robots.txt and confirm GPTBot, PerplexityBot, ClaudeBot and Google-Extended are not blocked. Many sites block them without knowing.
If you are invisible on three or more queries across multiple platforms, you have an AI visibility problem. The good news is that it is fixable — and less competitive to fix right now than traditional SEO.
What a Professional Audit Adds
The manual test tells you whether you are invisible. A professional audit tells you why, and in what order to fix it. LOOP's audit adds the diagnostic layers a manual check misses: schema and structured-data validation, entity consistency across directories, competitor citation mapping (which queries your competitors win that you don't), and a prioritised action plan rather than a raw score. The quick scan on our site tops out at 8 out of 10 deliberately — the last two points require a live citation check across all four engines, which is what the full report covers.
The full audit is free, delivered as a report you keep, with no sales call attached. If the findings show we are not the right answer for you, we say so.
What Happens When You Act on the Audit
An audit is worthless if nothing ships afterwards. Here is what real, recent results from our own client work look like — anonymised, and honest about the shape of the curve:
- A Malaysian F&B POS provider went from a 0% to 60% AI citation rate in roughly three weeks, on the back of about five targeted guides written against the queries the audit surfaced.
- One comparison guide from that programme was cited by ChatGPT, Perplexity, Google AI Overviews and Claude within six days of publishing.
- The pattern behind both: specific buyer-intent queries get cited within days ("X vs Y", "how much does X cost", "X for [specific niche]"), while generic "best X" head terms take structural work — entity building, third-party mentions and authority signals that compound over months, not days.
That last point is the honest one most agencies skip. Fast wins are real, but they come from precision, not magic. The head terms everyone wants take sustained effort.
What an Audit — and AEO — Cannot Do
A few things we tell every client up front:
- No one can guarantee citations. AI engines change behaviour frequently; what they cite today can shift. An audit measures and a programme improves the odds — nobody controls the answer.
- You cannot buy your way in. ChatGPT and Perplexity do not factor in ad spend. Paid ads will not get you into AI answers.
- Measurement is still maturing. There is no "Google Analytics for AI citations" yet — tracking means re-testing real queries on a schedule, which is exactly what a repeatable audit process exists for.
- An audit is a snapshot, not a fix. The value is entirely in what you ship afterwards.
This is the discipline behind Answer Engine Optimization (AEO) — the audit is step zero, and everything after it is content, structure and authority work.
Where to Start
Run the free manual test this week — it costs you thirty minutes and tells you the truth. If you would rather have the full picture, including the live four-engine citation check and competitor gaps, contact LOOP for a free AI visibility audit or read about our AEO Engine. We run the same playbook on loop.com.sg that we run for clients — this article is part of it.
Frequently Asked Questions
How much does an AI visibility audit cost in Singapore?
The manual method is free and takes about 30 minutes. LOOP's professional audit is also free — a full report across ChatGPT, Perplexity, Google AI Overviews and more, with no card and no sales call. Where you invest money is in fixing what the audit finds, not in the audit itself.
How often should I re-run an AI visibility audit?
Monthly if you are actively publishing, quarterly at minimum. AI engines update frequently and competitors are starting to optimise, so a six-month-old audit is stale. Ongoing AEO programmes typically re-test 10–20 core queries weekly.
How fast can results improve after an audit?
It depends on the query type. Specific buyer-intent queries can earn citations within days of publishing targeted content — one of our comparison guides was cited by four AI engines within six days. Generic "best X in Singapore" head terms take structural work over months. One client went from 0% to 60% citation rate in about three weeks by targeting the specific-query end first.
Is an AI visibility audit the same as an SEO audit?
No. An SEO audit measures your Google rankings and organic traffic. An AI visibility audit measures whether AI engines cite you in generated answers, and checks AI-specific factors like crawler access for GPTBot and PerplexityBot, extractable content structure and schema markup. Related foundations, different channel — see our AEO vs SEO comparison.