We checked our own AI visibility: named in 0 of 24 answers. Here is the 30-day plan
If we tell customers to measure their AI visibility and fix what the answers show, we should do it to ourselves first and publish the result, good or bad. On 11 October 2026 we asked ChatGPT, Gemini and Perplexity 24 buyer questions about our category. Citeably was named in none of them. This page is the log: the baseline, what we change, and the re-check after 30 days.
By the Citeably team · · · 5 min read
The baseline (11 October 2026)
We ran Citeably's own check on citeably.app: 24 answers across ChatGPT, Gemini and Perplexity, to questions a buyer might type when looking for a way to track AI search visibility. The brand name was left out of every question.
Result: Citeably was named in 0 of 24 answers. Earlier smaller runs on 10 October (10 questions each, other engine mixes) also named it in none.
That is expected for a new product with little written about it elsewhere, but it is the honest starting point.
What this number does and does not tell you
It is one run. AI answers vary from run to run, so a single result is a snapshot, not a verdict. A change only counts if it is bigger than the normal swing between runs, and with 24 answers that swing is large.
We will not claim any cause for a change. If the number moves, we will show what we changed in the same period and say plainly that the two may be unrelated.
What we are changing
We are following our own fix list, in the order the report gave it:
- Be found: make sure every public page is crawlable, in the sitemap, and readable without JavaScript, and that AI crawlers are not blocked.
- Be clear: a plain statement near the top of the homepage of what Citeably is, who it is for, and how it differs from the tools that were named.
- Be mentioned: write guides that answer the exact questions in the check, and ask to be listed where those categories are already compared.
The 30-day re-check
On or after 10 November 2026 we will run the same questions again and publish the result here, whether it is better, the same or worse, with the date and the engines used.
Until then the result section below is empty on purpose.
Result
Pending. This section will be filled in after the re-check, with the number, the date and what changed in between.
A second case: Testimoni
We ran the same kind of check on Testimoni, a testimonial tool we also run. On 11 October 2026, across Claude and Perplexity with 12 buyer questions, it was named in 2 of 12 answers, while three rivals were each named in 4 of 12. We will re-check it on the same schedule and publish that too.
Frequently asked questions
Why publish a bad result?
Because a tool that only shows flattering numbers is not worth trusting. If the plan works we will show it; if it does not, you should know that too.
Can AI visibility really change in 30 days?
Sometimes pages are picked up within weeks and sometimes not. We do not promise a result, which is why we are measuring ours in public.