Google Search Console adds AI impression data

Search Engine Journal's walkthrough of Google's new Search Console AI Mode report lands at an instructive moment: for the first time, brands have a first-party signal telling them which of their pages Google is actually pulling into AI-generated answers, and how that compares to where those pages sit in conventional rankings.

The gap between those two columns is where the real information lives.

The divergence that should unsettle your SEO assumptions

Conventional search rewards a predictable hierarchy: domain authority, backlinks, on-page signals, structured data. AI Mode answers do not have to follow that order. A page sitting at position 11 in organic results can appear in an AI answer; a page at position 2 may never surface in one. Search Console's new report makes this visible at scale, and for most large sites the picture will be uncomfortable.

The mechanism is selection by relevance to a conversational query, not by ranking position. Google's AI answer layer is retrieving content it judges most useful to a specific intent, which means optimising for ranked position alone no longer guarantees AI visibility. The two metrics measure different things, and treating them as interchangeable was always a mistake. Now there is data to prove it.

For a multilateral institution like UNDRR or a standard-setting body like ISO, the practical stakes are specific. These organisations publish large volumes of technical guidance, frameworks, and data that practitioners routinely query through conversational search. A framework document that ranks at position 8 but appears in AI answers for high-intent queries is doing more reputational work than a homepage that dominates traditional results but never gets cited in generated responses. Without this report, the distinction was invisible. With it, editorial and SEO teams can prioritise accordingly.

What the data actually requires you to do

Three comparisons matter most inside the new interface. First, pages with high AI impressions and low traditional rank: these are the content assets currently punching above their SEO weight in AI answers, which suggests they match conversational intent well and should receive more structured support, cleaner entity signals, and explicit authorship attribution to consolidate that advantage. Second, pages with high traditional rank and zero AI impressions: these are the most revealing. Google is choosing not to use them in generative answers despite their authority, which usually signals either that the content is too navigational, too thin, or structured in a way that resists extraction. Third, pages accumulating AI impressions but generating no clicks: a growing category as AI answers resolve queries without sending traffic, and one that demands a different measurement frame entirely.

The click question is pointed for financial services firms and large industrial groups. If a Holcim sustainability report or an Adecco workforce data publication is being cited in AI Mode answers but producing no referral traffic, the brand is providing source material without receiving the attribution signal that previously would have converted into a session. The asset is doing work; the analytics dashboard is not recording it. Impression data without corresponding click volume is not a reporting failure. It is evidence that AI visibility and traditional traffic are decoupling, and that share of citation is becoming a separate metric from share of click.

The practical adjustment is straightforward. Pages surfacing in AI impressions should be examined for how cleanly they state their claims. Google's AI layer favours content with clear declarative sentences, named entities, and attributable facts. Long discursive introductions, buried conclusions, and dense jargon are extraction barriers. Organisations whose authority rests on technical depth, which describes most multilaterals, policy institutions, and industrial standards bodies, face a structural editing task: making existing rigorous content more parseable without making it less rigorous.

Google adding first-party AI impression data to Search Console is less a tooling upgrade than a reclassification event. It is the moment at which AI visibility shifts from inference to measurement. Brands that begin auditing the gap between their AI impression footprint and their conventional rank now will define their citation strategy before their competitors think to ask the question.