← Insights
News Analysis21 September 20266 min read
CB
Costin BucuciCo-Founder & Commercial Lead

We Checked Whether Ahrefs, Semrush, Profound, and Otterly Actually Optimise for AI Citation. Most Don't.

If a company sells AI visibility monitoring, its own website is the first place you would expect to find every AI-citation best practice applied. We checked four of the best-known names, Ahrefs Brand Radar, Semrush's AI Visibility Toolkit, Profound, and Otterly.ai, against three concrete signals: whether they publish an llms.txt file, whether their product pages carry schema.org structured data, and how deliberately their own educational content is written to be quoted by an AI system rather than just read by a person. We built our own AI Visibility Score around exactly these signals, so we wanted to see whether the tools measuring everyone else's AI visibility were actually applying the same discipline to themselves.

llms.txt: none of the four have one

An llms.txt file is a plain text file at a site's root that tells AI crawlers what the site is and where its most useful content lives, the same job robots.txt does for search crawlers, purpose-built for language models. We checked all four directly. ahrefs.com/llms.txt, semrush.com/llms.txt, tryprofound.com/llms.txt, and otterly.ai/llms.txt each returned a 404. Kelriva AI's own llms.txt has been live since before this AI Visibility Score tool launched. It is a small file and an easy one to publish, which makes its absence on four companies whose entire product is AI visibility monitoring the most surprising finding in this check.

Structured data: thin or unconfirmed

Schema.org JSON-LD is what lets an AI system parse a page's facts directly instead of guessing from prose. Ahrefs Brand Radar's product page carried no JSON-LD in the HTML we checked. Profound's homepage was the same, pure marketing copy with no structured data behind it. We could not conclusively confirm the presence or absence of JSON-LD on Semrush's or Otterly's pages in this check, so we are not claiming either way there, only reporting what we could and could not verify.

Citable content: Otterly is the exception

Ahrefs Brand Radar's product page has no glossary or definitional content of its own (Ahrefs runs a general SEO glossary elsewhere on the site, unrelated to Brand Radar specifically). Profound's homepage is close to pure feature marketing, with no explanations of AEO concepts or defined terminology. Semrush has a broader knowledge base and academy, but nothing GEO-term-specific on the AI visibility product page itself. Otterly.ai is the clear exception: a FAQ-style section, a free GEO guide, and even a named experiment testing whether schema markup affects AI search results. That is the closest any of the four come to Kelriva's own approach, structured, citable glossary entries under DefinedTermSet schema, each written to be a self-contained, quotable answer rather than marketing copy.

What this actually means

None of this changes what these four platforms are good at. Ahrefs and Semrush have scale and data most small tools cannot match, and Profound and Otterly both do real monitoring work for enterprise clients. What it shows is narrower: on the specific, checkable signals that determine whether an AI system can read and cite a page accurately, the companies selling AI visibility monitoring are not consistently applying it to their own sites. Kelriva AI's free AI Visibility Score and glossary were built around these same signals from the start, not retrofitted after the fact. You can see exactly what an AI system says about your own business, verbatim, for free, in under two minutes.

GEOAEOGenerative Engine OptimizationAI SearchAI Visibility Score
Discuss this with AI:
Want more like this in Google Search?
Thinking about your AI infrastructure?

Let's talk about how your business is built today — and how we'd make it AI-ready.

Fixed-fee. Delivered in weeks. No lock-in to any single provider.

Book a discovery call →
Related reading
EducationWhat Is llms.txt, and Does Your Site Need One?robots.txt tells search engines what they can crawl. llms.txt does something different: it tells AI systems what actually matters. Here is what the standard is, what a real one looks like, and whether your site needs one yet.News AnalysisWe Asked AI Who's Best in Accountancy, Law, Fintech, and Coaching. Here's Who Showed Up.We ran 15 real UK accountancy, legal, fintech, and coaching firms through the same test our AI Visibility Score runs for clients: real, unprompted buyer questions, asked live to ChatGPT, Claude, Gemini, and Perplexity. In every single vertical, one firm dominated the answers. Most of the others never came up at all.EducationWhat AI Visibility Monitoring Actually Costs in 2026We compared the publicly listed pricing of every AI visibility monitoring platform we could verify, from $29/month tools up to $495/month enterprise platforms, against a free report and £6.99/month monitoring.
← Back to all insights