The average brand fails to appear in 84% of the AI search responses where it would want to be present, according to a new report distributed through GlobeNewswire. The finding puts a hard number on what many marketers have suspected since generative engines began absorbing query volume that once flowed to the ten blue links: visibility in AI search is not a default outcome of traditional SEO effort, and most brands are effectively absent from the majority of conversations that matter to them.

What the study measures

The report evaluates how often brands appear in AI-generated answers to queries relevant to their business — what the industry increasingly calls "AI visibility" or share of voice in generative search. The headline figure, 84% invisibility across target responses, means a typical brand surfaces in only about one in six answers where a mention, citation, or recommendation would be strategically valuable.

This is a tool-and-vendor measurement, not a statement from Google, OpenAI, or any search engine. Google has made no official claim about how often brands appear in AI Overviews, and OpenAI publishes no comparable metric for ChatGPT Search. The 84% figure comes from a third-party report, so readers should treat it as an independent measurement of AI answer content rather than a platform-confirmed benchmark.

That distinction matters. Confirmed statements from the search engines — such as Google's documentation on how AI Overviews select and cite sources — describe the systems generally, without publishing inclusion rates by brand or vertical. What this report adds is an outside measurement of the outcome: when you sample AI answers at scale and check whether a given brand shows up, the average brand is missing most of the time.

Why the gap exists

The mechanics of generative answers differ from ranked lists. An AI response synthesizes a small number of sources into prose, so the number of visible "slots" collapses from ten blue links to a handful of citations — or, in many chat interfaces, to no explicit links at all. A brand can rank in position five organically and still be paraphrased out of the answer entirely. Being crawled and indexed is no longer sufficient; being selected as raw material for the model's synthesis is the new threshold.

For brands in verticals with heavy review and recommendation intent — retail, software, travel, financial products — the invisibility problem compounds. AI answers tend to favor a compact set of names per category, which suggests early winners in each niche are capturing a disproportionate share of mentions while the long tail drops out of responses entirely.

What marketers should take from it

The practical implication is that AI visibility needs to be measured, not assumed. Teams that track only traditional rankings have no signal telling them whether models cite them, paraphrase them, or skip them. The 84% figure — whatever the exact methodology behind it — indicates that for the average brand, the default state in AI search is absence.

There is a caution here as well. Vendor-published visibility reports often accompany a commercial product, in this case distributed via a press-release wire rather than a peer-reviewed or independently audited study. The headline number is concrete and directionally consistent with what other AI-visibility tools have reported over the past year, but the sample composition, engine mix, and query-set design behind the 84% figure were not detailed in the announcement. Buyers of AI-visibility tooling should ask for exactly those parameters before benchmarking themselves against the average.

What to watch next

Watch for follow-up data from the same research covering per-engine breakdowns — AI Overviews versus ChatGPT versus Perplexity — and trend lines over time, since a single-point snapshot cannot show whether visibility is improving or deteriorating as the engines refine their citation behavior.