Are any companies completely invisible to AI search?

A natural fear for any business is that it simply does not exist as far as AI assistants are concerned, that asking ChatGPT, Gemini, or Claude about its category would never surface it at all. This study tested that directly. Across 100,966 companies, 14,951 distinct prompts, and three engines, the count of companies that never appeared in any answer on any engine was 0, out of 0 being 0% of the cohort. In plain terms: zero companies in this study were completely invisible.

This is worth sitting with because it changes the question a company should be asking. If total invisibility were common, the first priority would be getting mentioned at all. Since it is not, the real variation in this study lives somewhere else, in how often a company is mentioned and how prominently, not in whether it is mentioned even once.

The companion figure makes this concrete. While 0% of companies were never mentioned, 95.7% of the cohort, 96,644 companies, were visible on at least one engine but not on all three. That is the overwhelming majority of the cohort. Very few companies sit at either extreme of total absence or universal presence. Most sit in between: known to one or two engines, absent from the third, or named occasionally rather than reliably.

It is important to be honest about what this result can and cannot support. A finding of zero total-invisibility companies is a strong result given the cohort, but it describes this specific cohort of 100,966 companies across the categories this study covered, built from a list of 100,966 ranked companies overall. It says that within the set of companies this study chose to test, drawn from established category lists, none were completely absent. It does not mean every company in existence shows up somewhere in AI search. Companies without any established web presence or category listing were less likely to be included in a study cohort like this one in the first place, so the method is better at measuring unevenness among companies that already have some footprint than at measuring the odds of appearing for a company that currently has none.

The practical implication for a company worried about AI visibility is to reframe the worry. The question is less likely to be "will we ever be mentioned" and more likely to be "how often, how prominently, and on how many of the three engines." A company mentioned by one engine in a handful of prompts is visible by this study's zero-invisibility standard, but it is a very different competitive position from a company named consistently across all three engines and placed early in the answer. The next two findings in this study, on mention position and on cross-engine agreement, are where that distinction actually lives, and that is where a company's real work should be aimed once it has confirmed, as nearly every company in this cohort could, that it is not starting from zero.