Since the beginning of 2026, Google has rolled out several anti-spam updates in just a few months and fundamentally changed the way the Search Console reports performance data. The AI Overviews, already present in the American market, are gradually being deployed in Europe and are changing the very nature of what a “search result” is. This year’s SEO is no longer like that of 2024, and the usual benchmarks for web teams are shifting rapidly.
SEO Measurement in Search Console: What the New Filters Change Practically
Competitors are talking a lot about AI Overviews and AI-generated content. Few focus on a more immediate issue: the reliability of the SEO dashboards that most teams use on a daily basis.
Google announced a major evolution of the Search Console in September 2026. Impressions and clicks related to traditional results, AI Overviews, AI Mode, and multimodal searches (launched from a smartphone camera or a screenshot) are now distinguished. Before this update, all these interactions were mixed in the same report.
In practical terms, a site that was tracking its “average position” and “click-through rate” based on a single report was working with aggregated data, part of which came from formats where the click does not have the same meaning. The SEO updates published on www.virank.fr regularly detail this type of technical change in Google’s tools.
For SEOs, the practical consequence is clear: the old SEO dashboards have become incomplete. Reports need to be rebuilt by segmenting by result type, or else metrics that no longer measure the same thing will be compared.
AI Overviews and Click-Through Rate: Available Numbers Temper Enthusiasm
The prevailing discourse presents AI Overviews as an opportunity for additional visibility. The accessible data tells a more nuanced story.
According to a tracking report published by Loved by AI (review dated September 7, 2026, updated October 2, 2026), the proportion of AI Overviews containing external links has risen from almost zero to over 25% in just a few months. Google is therefore adding more citations in its generative responses.
On the other hand, an experimental study cited in the same report estimates that only 7.1% of clicks on a page featuring an AI Overview come from the citations embedded in the summary. The rest of the clicks are distributed between the traditional organic results located below the generative block and other elements on the page.
Generative Visibility and Real Traffic: Two Distinct Measures
Appearing in an AI Overview does not produce the same effect as occupying a traditional organic position. The generative block often answers the question posed without the user needing to click. Visibility exists, but traffic transfer remains low.
Field feedback varies on this point across sectors. Short informational queries seem to be the most affected by the lack of clicks, while transactional or highly specialized queries maintain a more stable organic click-through rate. The available data does not yet allow for conclusions about the exact extent of traffic redistribution across the French market.
Google’s Anti-Spam Updates in 2026: An Unprecedented Pace
Google launched its fourth anti-spam update of the year (September 2026 Spam Update) in September 2026, just over a month after the previous one. This pace has no recent precedent.
The accelerated cadence is partly explained by the proliferation of automatically generated content. Sites that massively publish AI content without editorial added value are the primary targets. The signals targeted by these updates include:
- Mass-produced content without proofreading or verifiable human expertise, often identifiable by its repetitive structure and lack of original data.
- Artificial link networks set up to simulate authority on recent domains, a technique that anti-spam filters are increasingly detecting.
- Satellite pages or “SEO parasite” sites that exploit the authority of a third-party domain to position low-quality content.
Four anti-spam updates in less than nine months means that the window between two algorithmic corrections is shrinking. A site affected by a filter has less time to recover before the next one hits.
Multimodal Search in Google: A New Type of Query to Watch
Since September 2026, searches launched from a smartphone camera or a screenshot appear separately in the Search Console. This is not an insignificant technical detail.
These multimodal queries do not rely on typed keywords. The user points to an object, a place, a product, and Google interprets the image to formulate a response. Traditional SEO, built around the match between a textual query and written content, does not cover this scenario.
Which Types of Sites Are Affected
Sectors where visual identification has a direct use are the first to be impacted: e-commerce (product recognition), tourism (identification of monuments or places), botany, decoration, fashion. For these fields, optimizing alt attributes and structured data for images becomes a concrete priority, not just a generic piece of advice.
Data on the actual volume of these multimodal searches remains limited. Google has not provided specific figures on their share of overall traffic. The separation in Search Console is a first step towards reliable measurement, but SEO teams will need to wait several months for segmented data before they can assess the real impact on their traffic.
SEO in 2026 is characterized less by a major break than by an accumulation of simultaneous changes: fragmented measurement, redistributed organic traffic, closely spaced anti-spam filters, non-textual queries. The challenge is not to understand each evolution in isolation, but to integrate them all into the same dashboard without losing data readability.
