For roughly two years, every B2B SaaS marketing team running organic as a channel had the same blind spot: AI-driven impressions were lumped into overall Search Console performance data with no way to isolate them. Google confirmed AI Mode and AI Overviews counted toward totals, but John Mueller clarified that all links within an AI Overview shared a single position in the reports. Figuring out what AI search actually did for your site meant guessing.
That measurement gap just got smaller. Not closed. Smaller.
What Google actually shipped
Google is testing two new Search Console features with a subset of UK websites before a global rollout. First, a dedicated generative-AI performance report showing impressions for URLs that appear in AI Overviews, AI Mode, and AI Overviews in Discover. The report breaks down by page, country, device, date, and (notably) hourly granularity. Second, a toggle letting site owners opt out of appearing in generative AI features entirely, without affecting traditional search rankings.
The toggle builds on existing controls like snippet management and Google-Extended, which governs whether content trains Google's AI models. This new switch is specifically about live AI search surfaces. Opt out, and you won't receive impressions or traffic from those features. Google says the choice won't be used as a ranking signal elsewhere.
The reports omit click data and query-level metrics. Google says it's "continuing to work with website owners to understand what insights will be most helpful" and plans to add more metrics over time. No specifics, no timeline.
Why this is useful but incomplete
Impressions by page, country, and device are a real step forward for Marketing Ops teams that need to segment AI-surface visibility. You can now build a monitoring cadence around which pages show up in AI features, track geographic variation, and spot device-level patterns. That's directional signal you didn't have before.
But impressions alone don't tell you whether AI search sends traffic that converts. The whole concern since AI Overviews launched in the U.S. in 2024 has been that Google can answer queries directly, reducing clicks to publishers even when a page contributes to the answer. A report that measures appearances without measuring click-throughs leaves the most important question unanswered: does AI visibility produce pipeline, or just vanity metrics?
Meanwhile, Bing moved faster. Bing Webmaster Tools launched an AI Performance dashboard in February 2026, added grounding query-to-page mapping in March, and previewed Citation Share at SEO Week in May. Microsoft isn't a major B2B SaaS traffic source for most teams, but the contrast in measurement maturity is hard to ignore.
The operational move for B2B SaaS teams
Don't wait for Google to ship click data. Pair the new AI report with your own downstream analytics to close the gap yourself.
Setup: Once the AI report rolls out to your property, export the page-level impressions data weekly. Cross-reference against GA4 landing page sessions, filtering for Google organic. Pages with rising AI impressions but flat or declining sessions are candidates for the "AI is answering the query without sending traffic" hypothesis.
The hypothesis (make it falsifiable): If pages with high AI-surface impressions show lower click-through rates than pages with comparable traditional search impressions, then AI Overviews are absorbing clicks for those queries because Google is resolving intent directly.
What to measure: Primary metric is the ratio of AI impressions to organic sessions per page. Secondary metrics are lead form submissions and trial starts from those landing pages. Guardrail: don't kill content that has low AI click-through but still drives branded search lift or assists in multi-touch attribution.
What not to over-interpret: Impressions in the AI report are not equivalent to traditional search impressions. A page appearing in an AI Overview supporting link may have been visible to the user for a fraction of a second, or buried behind a "show more" interaction. Treat the data as directional, not definitive.
Content strategy implication
Expert consensus points in one direction: generic SEO content is easier for AI systems to synthesize without sending traffic. Original data, proprietary benchmarks, and expert-led analysis are harder to replicate and more likely to earn both visibility and clicks. Google's own documentation notes that AI features surface supporting links from indexed pages eligible for snippets, with no extra technical requirements. "AI optimization" isn't a new technical discipline. It's a content quality and trust signal problem.
That's uncomfortable if your organic strategy has been built on volume plays and keyword coverage. It's an advantage if you've already invested in original research, named expert contributors, and content that can't be commoditized.
The toggle is interesting as an option but probably wrong as a default. Opting out of AI features means forfeiting whatever traffic those surfaces do send, and you lose the impression data needed to understand the channel. Unless you have strong evidence that AI appearances are cannibalizing your traditional search traffic with no offsetting value, keep the toggle on and measure.
Two years of blended data, then a report that shows half the picture. Google's AI visibility reporting has gone from nothing to something. The gap between "something" and "useful for pipeline decisions" is where your own analytics stack has to do the work.
