The illusion of a unified metric
Search traffic is disappearing into generative answers while researchers struggle to agree on the exact scale of the void, creating a fragmented picture of click behavior where reported losses vary wildly depending on the source.
The differences are stark.
The core issue preventing a unified consensus stems from fundamentally incompatible samples and conflicting definitions of a successful user interaction, because some analytical teams measure the clickability of specific ranking positions while others evaluate broad user session behavior across entirely different search intents.
This methodological discrepancy means that aggregating these figures into a single unified metric creates an illusion of clarity that the underlying data completely fails to support.
Incomparable methodologies and outdated figures
The drastic variance in traffic loss estimates originates from researchers using completely incomparable methodologies to track human behavior.
Comparisons fail.
A direct comparison between a user visiting a session and the specific clickability of a given ranking position remains statistically invalid, and furthermore, the relevance of certain cited studies is compromised by their publication dates falling in 2025 rather than the current year.
Data from both Amsive and the Pew Research Center date back to 2025, which naturally reduces their applicability to the modern interface environment.
It is crucial to note that all discussed metrics apply strictly and exclusively to artificial intelligence overviews within Google search.
Any attempt to utilize this data to predict the future situation for alternative search engines lacks foundational evidence and remains strictly prohibited by the limits of the research.
| Research Source | Publication Date | Key Metric Observed | Reported Change |
|---|---|---|---|
| Ahrefs | February 4, 2026 | Position click loss | Significant positional drop |
| Pew Research Center | July 22, 2025 | Standard result clicks | Interaction decrease |
| Seer Interactive | 2026 | Summary clickability | Slight interaction growth |
| Amsive | 2025 | Average traffic drop | Overall visibility loss |
Our previous analysis regarding where the clicks go position by position demonstrated a classical distribution curve across standard ranking placements, yet this new evaluation exclusively captures how that established performance curve morphs and flattens upon the introduction of a generative answer.
Positional drops versus session interactions
An extensive analysis published by Ahrefs on February 4, 2026 provides the most severe assessment of the current search landscape.
According to their updated tracking metrics, the introduction of a generative answer suppresses traditional organic traffic across the entire first results page.
The data indicates a 58 percent reduction in clicks for the top organic position when a generated summary appears above it.
The second organic position suffers a similarly drastic 50.8 percent decline under the exact same experimental search conditions.
Even the very bottom of the first page feels the impact, with the tenth ranking position experiencing a documented 19.4 percent drop.
The earlier April estimate of 34.5 percent describes overall traffic loss rather than the first position alone, so the two figures measure different things and cannot be read as one trend getting worse.
Methodologies completely diverge.
This position tracking counts results, while the survey work counts people and their visits, and a single person can run many searches inside one visit.
The reluctance to click generated links
A distinct approach to measuring user interaction comes from a Pew Research Center study published on July 22, 2025.
Rather than tracking individual ranking positions, this prominent organization evaluated the overall propensity of users to select traditional hyperlinks during a session.
Their detailed findings demonstrate that only 8 percent of users click on a regular result when an artificial intelligence summary is present on the screen, although this baseline interaction rate nearly doubles to 15 percent when the search interface displays standard blue links without any generated text appearing at the top.
The study also reveals an extreme reluctance among participants to interact with the generated content itself, showing that a link inside the summary itself was clicked in 1 percent of those visits.
Context matters immensely.
While these statistics offer valuable insight into fundamental human behavior, the 2025 publication date inherently limits their relevance for assessing the current state of search.
The rapid evolution of generative interfaces means older session data might not accurately reflect how users interact with the continuously updated modern search environment.
The transparency problem with closed reports
The landscape of available data becomes even more convoluted when incorporating findings from Seer Interactive and Amsive.
Coverage published by Search Engine Land in 2026 detailed Seer Interactive data showing that the overall clickability of generative summaries grew from 1.3 percent to 2.4 percent.
Their proprietary data also highlighted a unique advantage for specific entities, revealing that cited brands receive 120 percent more clicks than their unmentioned competitors.
Meanwhile, a separate 2025 report published by Amsive calculated an average overall traffic drop of 15.49 percent across their measured datasets, breaking this overall decline down into an unbranded traffic decline of 19.98 percent alongside a contrasting branded traffic growth of 18.68 percent.
Both the Amsive report and the Seer Interactive data share a critical transparency flaw that significantly complicates independent external verification, as these valuable datasets are presented exclusively through closed reports and secondary publications without any direct public web address linking to the primary tables.
Raw data remains hidden.
The complete absence of accessible tables prevents independent analysts from fully understanding the precise mathematical approaches used to calculate these specific percentage changes.
The missing industry standard
The fundamental incompatibility of these four research approaches ensures that no single definitive number can accurately represent the total loss of search traffic.
Researchers attempting to measure the exact impact of generative summaries must choose between tracking specific positional drops and evaluating broader session behavior, guaranteeing that the resulting figures will describe entirely different aspects of the evolving search ecosystem.
Readers reviewing these statistics must carefully examine the specific methodology and publication date before accepting any percentage as a universal truth.
Since no single definitive number can accurately represent the total loss, comparing these conflicting reports requires a careful examination of the specific methodology used by each research group.
Sources and statuses
- 1 Ahrefs Verified
- 2 Ahrefs Verified
- 3 Pew Research Center Verified
- 4 Pew Research Center Verified