Tenva's AI visibility baseline: 0 of 40

Why would a company publish its own zero?

Because the zero is the point. Tenva sells AI visibility measurement and the work that moves it, so the honest way to prove the method is to run it on ourselves first and publish the worst possible starting number. Every number on this page comes from a probe run Tenva can re-execute on demand.

This baseline is the BEFORE in a public before-and-after experiment. The 26 answer pages on this site went live on June 6, 2026. The same ten questions get re-run monthly, and this page updates with each run. If the method works, the citation count climbs and you can watch it happen. If it does not, that will be visible too.

How does the measurement work?

Each question is asked once per engine through the engine's own API with live web search enabled, exactly as a buyer would ask it. Every answer is recorded in full. A citation means the engine linked tenvaai.com as a source in its answer. A mention means the answer text named Tenva at all, even without a link. Both are counted separately for every one of the 40 answers.

The four engines are OpenAI (ChatGPT), Anthropic (Claude), Google (Gemini), and Perplexity. One sample per engine per question keeps the run small and repeatable; the trade-off is that single samples vary between runs, which is why the monthly re-run uses identical questions and why movement only counts when it repeats.

What were the ten questions?

The questions are the ones Tenva's own buyers ask, written in plain language. Six are commercial, for example "best AEO agency for private medical practices", "AI visibility services for plastic surgery practices", "answer engine optimization for healthcare practices", "who can help my practice show up when patients ask AI for recommendations?", and "companies that help orthodontists show up in AI search".

Four are informational, for example "how do private practices get recommended by ChatGPT?", "how to get my med spa recommended by AI assistants", "how to get more patients from ChatGPT", "AEO vs SEO for medical practices", and "how do I find out what ChatGPT says about my practice?". The exact wording is frozen so every monthly re-run measures the same thing.

What were the results?

Tenva was cited in 0 of 40 AI answers to its own buyer questions in June 2026. Tenva was also mentioned in 0 of 40. All four engines returned usable answers for all ten questions, so the zero is not a data gap: the engines answered every question and recommended other companies every time.

That is the floor this experiment starts from, and it is the same floor most private practices sit on today. The answers themselves named a long tail of marketing vendors and blogs, which is exactly why the recommendation slots are winnable.

A second probe a day later widened the picture: across both June 2026 runs combined, Tenva appeared in 0 of 95 AI answers, because Tenva had no pages targeting the additional gap questions when they were probed. Tenva re-runs its own baseline probe across two question sets to keep the measurement honest.

Does the same method work for a practice?

Yes, and it has been run. In May 2026 Tenva ran the identical method for one Philadelphia cosmetic dentistry practice using ten geo-specific patient questions instead of buyer questions. The practice was cited in 5 of 40 answers. The other 35 answers recommended someone else, which is the gap the work targets.

The practice is kept anonymous here because the audit was a private measurement, not a public case study. The method, the counts, and the dates are exactly what is reported in the evidence panels across this site.