Zero-Click Measurement and Impression Value

Clicks and sessions are the metrics most SEO reports are built around. They measure something real, but they measure only the fraction of organic exposure that results in a site visit. The majority of organic exposure (brand name appearances, answer citations, featured snippets, People Also Ask boxes) produces no measurable click in Google Analytics. Measuring only clicks means measuring a shrinking share of organic reach.

This article covers the analytics side of zero-click search: how to quantify it, how to assign value to impressions, and how to detect the specific CTR decay effect that AI Overviews introduce. For the SERP features and search behaviours that produce zero-click outcomes, see zero-click search.

How large is the zero-click problem?

SparkToro’s 2026 analysis of Similarweb clickstream data found that 68.0% of US Google searches between January and April 2026 ended without a click of any kind, and that only 232 of every 1,000 searches sent a click to the open web.1 The UK figure is higher, at 69.5% zero-click and 232 open-web clicks per 1,000 searches, while Germany, Italy, Canada and France sat between 62% and 66%.2 A zero-click search in this data is one with no click at all: the searcher was satisfied by a Featured Snippet, Knowledge Panel, People Also Ask box, AI Overview or local pack, searched again, or left.

When an AI Overview appears, clicks fall further. Pew Research Center tracked 68,879 Google searches by 900 US adults in March 2025 and found that users clicked a traditional result on 8% of visits where an AI summary appeared, against 15% where none did, and clicked a link inside the summary on 1% of visits.3 Semrush’s own keyword-level study points the other way: tracking 200,000 keywords through 2025 with Datos, it found that terms which gained an AI Overview between May and October saw their zero-click rate fall from 33.75% to 31.53%, and concluded that “AI Overviews don’t automatically increase zero-click behavior”, the higher rate on those terms reflecting their informational intent.4 The two studies measure different things, user visits against keyword series, and neither settles it. In Google AI Mode, where the interface is conversational by default, Semrush’s analysis of almost 69 million US desktop Google Search sessions between 1 May and 5 July 2025 found that 6% to 8% of AI Mode sessions led to an external domain, a zero-click rate of 92% to 94%.5

The implication: as AI Overviews expand and AI Mode rolls out, the share of Google traffic that clicks through to external sites is declining. A site maintaining flat click volume while its impression volume grows is, in practice, losing reach relative to query volume. A site whose CTR is falling on previously strong queries may have AI Overview coverage that has absorbed the clicks.

How do you read the GSC impressions vs. GA4 sessions gap?

The gap between GSC impressions and GA4 organic sessions is not a discrepancy to fix: it is signal to interpret.

GSC impressions count how many times a page appeared in search results, including impressions that never generated a click. GA4 organic sessions count visits that reached the site. The difference between them includes:

  • Searches where the user did not click (zero-click)
  • Searches where the user clicked a different result (your page appeared but they chose another)
  • Sessions that reached the site but were not tracked in GA4 (cookie rejection, JavaScript blocking)

A widening impressions-to-sessions gap, controlled for GA4 tracking changes, is a signal that more of your impressions are going unclicked. Tracking this ratio over time, impressions per organic session, on high-visibility pages identifies where zero-click pressure is increasing.

Two Search Console faults distort this ratio

The impressions side of this calculation is not clean. A logging error caused Search Console to over-report impressions from 13 May 2025 to 27 April 2026, dates Google records on its data anomalies page, and Google has confirmed it will not backfill the correction. Separately, the removal of the num=100 parameter in September 2025 stripped out impressions that SEOs attribute to automated crawlers, an explanation Google has not documented.

Both inflate the earlier side of any historical comparison, and both move impressions while leaving clicks and sessions untouched. That matters here more than almost anywhere else, because an impressions-to-sessions ratio with an inflated numerator reads as rising zero-click pressure that is not real. Any ratio trend spanning those dates needs the discontinuity accounted for before it is interpreted as an AI Overview effect.

How do you assign value to impressions?

Impressions have brand exposure value that does not appear in session data. Quantifying that value requires a modelled approach.

The simplest method: apply a per-impression value rate based on the type of search visibility.

Estimated approach:

  1. For each query where you have impressions, calculate the expected CTR at your average position. First Page Sage’s 2026 table, built from Search Console data on 1,184 client and panel domains between March and August 2026, splits each position by whether an AI answer appears: position 1 at 22.6% without one and 3.6% with one, position 3 at 5.8% and 0.8%, and position 10 at 0.5% and 0.1%. Use the column that matches the query’s results page rather than the blended figure, and treat both as an agency’s estimate from its own panel.6
  2. The gap between expected CTR and actual CTR, multiplied by your average value per organic session, estimates the revenue or value equivalent of clicks that would have arrived at a higher CTR.
  3. Impressions with zero clicks but high exposure on brand-relevant queries can be assigned a brand exposure CPM: the equivalent cost to generate the same branded exposure through display advertising.

This produces a directional figure, not a precise attribution. Its value is framing impressions as a reportable metric with a monetary proxy.

How do AI Overviews affect CTR?

The CTR effect of AI Overviews is query-specific. Queries where AI Overviews appear consistently show CTR decay on the organic results below the AI answer.

To identify AI Overview impact on specific pages in GSC:

  1. Open the Performance report.
  2. Filter to the page in question.
  3. Switch to the Queries tab and sort by impressions.
  4. Compare the CTR trend over the past 12 months for queries where impressions are flat or growing but CTR is declining.

A falling CTR with stable impressions is the signature pattern of AI Overview coverage absorbing clicks for those queries. The page is appearing in results (impressions unchanged), but a higher proportion of users are reading the AI answer instead of clicking through.

This is directional, not definitive. CTR changes have other causes (position changes, title tag changes, seasonal shifts), so cross-reference the timeline of CTR decline against the dates when AI Overviews were confirmed to have appeared for those queries. Since 31 August 2026 that date can be read rather than inferred: Search Console’s Search Generative AI performance report shows impressions from AI Overviews and AI Mode by page and date for every site, though it carries no click or query data, so the CTR comparison still has to be made in the standard Performance report.7

How do you report zero-click value to stakeholders?

A reporting framework that includes impression value alongside click data:

Clicks and sessions. Traditional SEO metrics, direct revenue attribution.

Impressions. Total search exposure; trend over time indicates reach expansion or contraction.

Impression-to-session ratio. The click yield from impressions; declining ratio signals zero-click pressure.

Estimated brand exposure value. Impressions on brand-relevant queries × CPM equivalent, as a proxy for non-click brand value.

AI-attributed sessions. From the GA4 AI Assistant channel group; the portion of AI-driven traffic that does generate a site visit. For AI citation rate and share of voice measurement across ChatGPT, Perplexity, and Gemini, the AI visibility that does not produce a trackable session, see AI visibility measurement.

Adding these rows to a standard monthly reporting template makes zero-click impact visible without requiring a separate reporting process or a significant measurement build.

Footnotes

  1. In 2026, less than one third of Google searches still send a click — SparkToro, 8 June 2026. Similarweb desktop and mobile panel, US, January to April 2026. ↩

  2. Zero-click searches: highest in the UK, lowest in Germany — SparkToro, 16 June 2026. Same panel and window for the UK, France, Canada, Italy and Germany. ↩

  3. Google users are less likely to click on links when an AI summary appears in the results — Pew Research Center, 22 July 2025. ↩

  4. Semrush AI Overviews study: what 2025 SEO data tells us about Google’s search shift — Semrush, refreshed December 2025. Zero-click analysis of over 200,000 keywords with Datos, January to October 2025. ↩

  5. Google AI Mode’s early adoption and SEO impact — Semrush. Clickstream data for almost 69 million US desktop Google Search sessions, 1 May to 5 July 2025. ↩

  6. Google Click-Through Rates (CTRs) by Ranking Position in 2026 — First Page Sage, 22 September 2026. Agency panel of 1,184 domains, US-weighted, desktop and mobile blended, March to August 2026. ↩

  7. Introducing Search Generative AI performance reports in Search Console — Google Search Central. Carries Google’s note: “As of August 31, 2026, we’ve rolled out these insights to all websites worldwide.” ↩