The 15-Minute Competitor AI Visibility Audit You Can Run on Any Rival

TL;DR:

A competitor AI visibility audit shows whether your rivals are built to be cited by ChatGPT, Perplexity, and Google’s AI Overviews, or whether they are invisible to those engines. You can run one on any website in 15 minutes using four public signals. The signal most executives ignore, the robots.txt bot policy, is where a competitor’s entire AI strategy leaks.

Open any competitor’s robots.txt right now. Type their domain and add /robots.txt. In ten seconds, you know something their CMO probably does not: whether they let ChatGPT read their site or shut it out. Google’s AI Overviews already cut clicks to the top organic result by 58%, according to Ahrefs analysis of 300,000 keywords. The companies winning that traffic are not the ones with the largest content budgets. They are the ones machines can read, parse, and quote. A competitor AI visibility audit turns four public files into a map of who is positioned for AI search and who is exposed. Here is the version you can finish before your coffee gets cold.

What a competitor AI visibility audit actually reveals

A competitor AI visibility audit is a structured inspection of the public signals that decide whether AI search engines can access, understand, and cite a website. Every signal is public. None requires a paid tool or a login. The audit answers one question your own analytics never will: Is a rival defended, exposed, or absent in AI search? It also reveals intent. A site configured deliberately for AI search visibility looks different from one that stumbled into it by accident, and that difference tells you exactly where they are beatable.

Signal 1: The bot policy hidden in robots.txt

A website’s robots.txt file is the clearest public statement of its AI strategy. It lists which crawlers the site allows and which it blocks. The names that matter are GPTBot and OAI-SearchBot from OpenAI, Google-Extended, ClaudeBot, and PerplexityBot. Roughly 29% of all sites now block GPTBot, and 25% of the top 1,000 sites do. Here is the nuance almost no one checks: blocking GPTBot stops training, but blocking OAI-SearchBot is what removes a site from ChatGPT’s live citations.

Only 10.7% of sites use the Google-Extended distinction that blocks AI training while keeping search intact, a split Google documents in its crawler, and robots.txt guidance. A rival that blocks OAI-SearchBot has voluntarily deleted itself from ChatGPT answers. That is a gift. Confirm your own robots.txt configuration before you celebrate theirs.

Signal 2: Structured data, the signal everyone overstates

Structured data, also called schema markup, tells AI engines what a page means rather than what it says. The evidence on its impact is contested, and pretending otherwise is how agencies oversell it. A Search/Atlas study found no correlation between schema coverage and citation rates. A BrightEdge study found a 44% increase in AI citations after adding structured data and FAQ blocks. Treat schema as necessary, not sufficient.

To check a competitor, view page source and search for “application/ld+json,” or run the page through Google’s Rich Results Test. The schema markup for WooCommerce types they deploy, Product, FAQPage, and Organization, reveals which AI surfaces they target. The vocabulary itself is public at Schema.org. Implementing it well across a large catalog is structural work; teams like WPRiders handle schema and headless WooCommerce builds precisely because hand-editing markup at scale breaks the moment a template changes.

Signal 3: llms.txt, mostly theater

The llms.txt file is the most overhyped signal in AI search, and its presence on a competitor’s site tells you more about their marketing team than their traffic. Adoption sits near 10% across 300,000 domains. Google’s Gary Illyes confirmed Google does not read the file and has no plans to. Major crawlers skip it and parse HTML directly.

So when a rival publishes an llms.txt file, read it as a signal of attention, not advantage. It proves they are watching AI search. It does not prove they are winning it. Check the date on the file too, because a stale llms.txt untouched in a year confirms a rival ran an AI checklist once and moved on. The one exception is documentation-heavy sites whose pages feed AI coding assistants, where the file earns its place.

Signal 4: The citation footprint

A competitor’s citation footprint is the number of times AI engines name or link them in answers to category questions. Citation footprint is the only signal in this audit that measures outcome rather than setup. Pick the five buying questions your customers actually ask, in the shape of “best tool for small teams,” “top alternatives to the market leader,” “is this category worth the money,” “what does implementation cost,” and “product A versus product B.” Put each into ChatGPT, Perplexity, and Google’s AI Overviews. Record who gets cited and how often.

A rival cited in four of five answers owns the category in AI search regardless of their Google ranking, and that gap is invisible in every traditional rank tracker you already pay for.

The Four-Signal Teardown: a Monday-morning scorecard

The Four-Signal Teardown converts those checks into one competitive read. Score each rival from zero to two on each signal: access, structure, llms.txt intent, and citation footprint. Then plot them on two axes: AI Access (open or closed crawlers) and Machine Readability (strong or weak structure). Four types emerge. The Magnet is open and well-structured, cited often, and hard to displace. The Sitting Duck is open but unstructured, accessible yet messy, and the easiest rival to outrank in AI answers. The Leaky Fortress is closed but well-built, invisible in ChatGPT despite good engineering, and beatable there immediately. The Ghost is closed and weak, absent from AI search entirely.

Your fastest wins come from out-structuring Sitting Ducks and occupying the space Leaky Fortresses abandoned when they blocked the crawlers. Rank your opportunities by the size of the citation gap, not the polish of the website, because a beautiful site that machines cannot read is a beautiful site that does not get quoted.

Reading the gap between setup and citations

The most valuable finding in any teardown is the gap between a competitor’s technical setup and their actual citation footprint. A rival with a clean schema and open crawlers that still gets cited rarely has a content and entity authority problem, not a technical one.

A rival cited constantly, despite a weak setup, is coasting on brand authority that you can erode. Closing these gaps touches server configuration, theme templates, and structured data at once, which is why the work rarely survives as a side project inside a busy marketing team. In our work with WordPress and WooCommerce clients, the teardown is step one because it tells you whether the fix is technical, editorial, or both before anyone spends a budget guessing.

Key Takeaways

• A competitor AI visibility audit uses four public signals: bot policy, structured data, llms.txt, and citation footprint.
• A website’s robots.txt file is the clearest public statement of its AI strategy.
• Blocking OAI-SearchBot removes a site from ChatGPT citations, while blocking GPTBot only stops AI training.
• Schema markup is necessary but not sufficient for AI visibility, and its measured impact is contested across studies.
• The llms.txt file signals a competitor’s attention to AI search, not their success at it.
• Citation footprint is the only AI visibility signal that measures outcome rather than setup.
• The biggest competitive opening is the gap between a rival’s technical setup and their actual AI citations.

Conclusion

The teardown is a snapshot, and snapshots age fast. Google’s AI Overviews appeared in roughly 13% of queries and keep expanding into commercial searches where revenue lives. The rivals who treated AI access as an afterthought are about to learn what 58% fewer clicks feels like. The ones who structured their sites for machines will compound their lead quietly, because AI engines reward consistency over time. Run this audit on your three closest competitors this quarter, then run it on yourself. The companies that win the next phase of search will pair WordPress fluency with AI search architecture, and they will treat the gap between setup and citation as the real scoreboard.

FAQs