Zero-Click Searches in 2026: What the Pew Data Says for SaaS

What are zero-click searches and what does the 2026 data actually show?

Zero-click searches are queries where users get their answer directly inside the search-results page — through AI Overviews, featured snippets, knowledge panels, “People Also Ask” expansions, or the SERP itself — without clicking through to any third-party site. The 2024 baseline measurement from SparkToro and Datos put US zero-click rates at 58.5% of all Google searches; updated measurements through 2025-2026 show the share rising as AI Overviews expand. For SaaS marketing, the strategic shift is from optimising for clicks (a smaller share of total query volume) to optimising for citation eligibility within the answer surfaces, because being mentioned inside an AI Overview is now a top-of-funnel touchpoint independent of whether the user ever visits the site.

TL;DR — Key takeaways

• SparkToro and Datos’s 2024 zero-click study measured 58.5% of US Google searches ending without a click to a non-Google property. That measurement was taken before AI Overviews became widespread, and the share has continued to rise: SparkToro’s 2026 follow-up (Similarweb clickstream data) put US zero-click at 68.01% for January–April 2026, up from 60.45% in 2024.
• Pew Research’s 2025 analysis found that on Google searches where an AI summary appears, users click on a traditional search result 8% of the time; on searches without an AI summary, the click rate is 15%. That is a roughly 47% relative CTR reduction at the user-behaviour level, measured from actual browsing data rather than vendor-aggregated impression metrics — the strongest single piece of evidence available on AI Overview click impact.
• Ahrefs’s AI Overview citation research found 38% of Google AI Overview citations come from pages already ranking in Google’s top 10 — down from ~76% in July 2025, as Google shifted toward query fan-out — meaning ranking well in traditional organic results is still a major driver of being cited in zero-click answers. Zero-click optimisation does not replace SEO; it inherits from it.
• ConvertMate’s analysis of 80M+ AI citations identified the citation-eligibility drivers: 67% schema markup uplift, 41% expert-attribution uplift, 4.1× original-data multiplier, 3.2× freshness multiplier. These are the levers that turn an organic ranking into an AI citation.
• Seer Interactive’s traffic-quality measurement found ChatGPT-referred sessions convert at 15.9%, Perplexity at 10.5%, and Google at 1.76% — meaning the smaller volume of zero-click-to-click traffic is materially higher-intent. The strategic implication is that lost click volume is partially compensated by sharply higher conversion rates on the clicks that remain.
• Zero-click optimisation sits inside Phase 8 (AI citation readiness) of the 12-phase SEO & GEO audit framework, but it depends on Phases 1-7 (technical health, indexability, content quality) being in place first. The hierarchy matters: an unindexed page cannot be cited; an uncached page cannot rank; an unranked page is rarely cited.

How much have AI Overviews actually reduced organic clicks?

The single most useful data point is the Pew Research finding from July 2025: on Google search results that include an AI summary, users click a traditional result 8% of the time; on results without an AI summary, the click rate is 15%. That is roughly a 47% relative reduction in CTR, measured from actual user browsing rather than vendor impression aggregates. The Pew sample was 900 US adults who agreed to share their browsing data — methodologically clean, transparent, and one of the few zero-click measurements grounded in observed behaviour rather than panel surveys or projected estimates.

What that data does not say: it does not say the reduction is uniform across query types. Informational queries (the kind that trigger AI Overviews most often) take a larger CTR hit than navigational or transactional queries. Branded queries are largely unaffected. Commercial-investigation queries — the kind that drive most SaaS evaluation — sit somewhere in between, depending on how complete the AI Overview’s answer is.

The other useful piece of context is volatility. Conductor’s tracking of AI Overview impression share from late 2025 through early 2026 showed the share of queries triggering AI Overviews fluctuating between 23% and 47% across a six-month window (September 2025 through February 2026), depending on Google’s category-specific testing. The 2026 baseline is closer to 35-45% of informational queries triggering an AI Overview, but the number swings 5-10 points monthly as Google adjusts which categories are eligible. For SaaS marketers, this means measurement windows shorter than 30 days are noise; the right monitoring cadence is monthly or quarterly, against a stable baseline measured before AI Overview rollout in the relevant query category.

The mid-tier secondary reporting that cites a flat “34.5% CTR reduction” or “64% organic traffic drop” is usually aggregating Ahrefs, Similarweb, and proprietary panel data into a single headline number. Those numbers are directionally consistent with the Pew finding (large CTR compression, particularly for informational queries) but the precision implied by the specific percentage is rarely defensible. The honest framing is “AI Overviews compress click-through by roughly 40-50% on the queries where they appear, with the magnitude varying by query type and the completeness of the AI Overview’s answer.”

If clicks are compressing, how do you get cited inside the answer?

This is where the Ahrefs top-10 finding becomes load-bearing. In Ahrefs’ March 2026 update, 38% of AI Overview citations came from pages already ranking in Google’s top 10 for the query — down from ~76% in July 2025, as Google shifted toward query fan-out. The strategic implication still holds: ranking is a major lever for being cited. Zero-click optimisation does not replace SEO; it inherits from it. The order of operations is rank in the top 10, then optimise for citation among the top 10 candidates.

Once a page is in the top 10, the citation-eligibility levers identified in ConvertMate’s analysis of 80M+ AI citations stack roughly as:

• Schema markup (67% citation uplift). Pages with valid Article, FAQPage, HowTo, or Product schema are cited at materially higher rates than equivalent pages without schema. The schema must be in the server-rendered HTML — AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended) do not execute JavaScript. The implementation pattern is covered in the structured data for AI search piece.
• Expert attribution (41% uplift). Content quoting named experts with verifiable credentials is cited more than content making the same claims unattributed. For SaaS, this means quoting your CTO, customer-facing engineers, or named product researchers rather than presenting their analysis as institutional voice.
• Original data (4.1× multiplier). Content with first-party data — your own product usage stats, your own customer cohort analyses, your own survey results — is cited at over four times the rate of content rehashing the same secondary sources every competitor cites.
• Freshness within 30 days (3.2× multiplier). AI engines weight recently updated content heavily. The cadence is closer to monthly editorial review than annual content audits.
• Direct-answer formatting. Answer the question in the first paragraph below the H2 that poses it. AI engines extract opening paragraphs into AI Overviews; pages that bury the answer 800 words in are rarely cited even when the answer is correct. This is the foundation of answer engine optimisation — structuring each section to answer its query directly rather than building toward a conclusion.

The AI Overview ranking piece covers the citation-mechanics layer in more depth; the ChatGPT citation piece covers the equivalent levers for the AI assistants that crawl independently of Google.

What does this mean specifically for B2B SaaS funnels?

B2B SaaS buying journeys involve a long discovery and evaluation phase before sales contact. Zero-click search reshapes that phase in three concrete ways.

Top-of-funnel exposure increasingly happens without site visits. A prospect researching “best CRM for series A startups” reads the AI Overview, sees three or four vendors mentioned in the summary, and forms an initial consideration set without clicking through. The vendor that gets cited in the summary has been mentioned to the prospect at the same level of authority as the others, even though no visit occurred. Traditional first-touch attribution misses this entirely — the prospect’s first measurable touchpoint becomes a brand search or direct visit weeks later.

Mid-funnel evaluation queries shift toward AI assistants. Once a prospect has a shortlist, deeper questions like “how does X handle multi-tenant billing” increasingly route through ChatGPT, Perplexity, or Claude rather than Google. These assistants build their answers from your documentation, your blog posts, and your case studies. Sites with comprehensive technical content, named-engineer authorship, and current freshness are cited; sites that don’t update their content quarterly are increasingly invisible at this funnel stage.

The clicks that do happen carry higher intent. Seer Interactive’s measurement showed ChatGPT-referred traffic converting at 15.9% versus Google’s 1.76% — roughly 9× the rate. The smaller volume is partially compensated by the higher conversion rate. The strategic reframing: optimise for citation eligibility (volume goes down, but intent rises) rather than treating the click-rate compression as pure loss.

The B2B-specific addition is covered in the B2B AI search buying-journey piece — particularly the multi-touchpoint attribution and the relationship between AI-influenced sessions and downstream conversion.

What should you actually be measuring in 2026?

Three measurement layers, in priority order:

Layer 1 — Citation visibility. Track which AI engines (ChatGPT Search, Perplexity, Claude, Google AI Overviews) cite your content for the prompts that matter to your business. This is qualitative monitoring: a prompt matrix of 30-50 prospect-equivalent queries, run monthly, with citation patterns recorded. Tools like Otterly, Profound, and Ahrefs Brand Radar automate this; for early-stage SaaS the manual approach is fine. The metric is share-of-citation within your competitive set, not absolute citation count.

Layer 2 — AI-referred session quality. The clicks that do happen from AI assistants are higher-intent. Track AI-referred sessions in GA4 using the channel grouping pattern in the AI referral traffic tracking piece. The metrics that matter: conversion rate, demo-request rate, and time to conversion. Compare against organic traffic baseline. If AI referrals are 10× the conversion rate of organic (consistent with Seer’s published data), that 10× multiplier matters in the funnel attribution even at low absolute volume. Connecting those sessions to downstream revenue rather than leaving them as an isolated GA4 row is the job of full-funnel tracking.

Layer 3 — Brand search lift. Zero-click exposure creates brand awareness without a measurable click event. The downstream signal is direct traffic and branded-query volume. Track branded GSC impressions and direct-traffic baseline monthly; an inflection upward without a corresponding paid-media spike is the most reliable signal that zero-click exposure is converting to active consideration.

Traditional metrics — organic sessions, keyword rankings, page-1 share — remain useful but become lower in the priority order. They measure inputs (rankings) rather than the outcome that matters (citation share + conversion).

What’s the worst thing SaaS marketers do when they react to zero-click data?

Three predictable failure modes:

Disinvesting from SEO because “clicks are dying.” The Ahrefs top-10 finding is the counter-argument: 38% of AI Overview citations still come from pages already ranking in the top 10 (down from ~76% in July 2025, as Google shifted toward query fan-out). Disinvesting from the rank position that earns a large share of citations is the wrong direction. The right adjustment is to maintain SEO investment for the rank, and add schema, expert attribution, original data, and freshness on top — the levers that turn the rank into a citation.

Optimising for AI Overview inclusion at the expense of human readability. The pages that get cited consistently are also good pages for human readers. Lead paragraphs that directly answer the H2’s question serve both audiences; gimmicky formatting that helps AI extraction but reads poorly to humans hurts engagement metrics that AI engines indirectly weight.

Treating zero-click as a transient phenomenon to wait out. The Pew measurement of 8% vs 15% click rates, the SparkToro 58.5% baseline, and the Conductor volatility tracking all point in the same direction: zero-click is the structural state of 2026 search, not a temporary anomaly. SaaS marketing strategies that assume click rates will recover to 2022 levels will be uncompetitive within 18 months. The right horizon is to build the citation-eligibility infrastructure now, while the competitive set is still adjusting.

Frequently Asked Questions

What percentage of Google searches actually end without a click in 2026?

SparkToro and Datos’s 2024 study measured 58.5% of US Google searches ending without a click to a third-party website. SparkToro’s 2026 follow-up (using Similarweb clickstream data) put US zero-click at 68.01% for January–April 2026 — up from 60.45% in 2024 — confirming the share has continued to rise as AI Overviews expand. Pew Research’s July 2025 user-behaviour study found that on searches where an AI summary appears, users click traditional results only 8% of the time vs 15% on searches without an AI summary — a 47% relative CTR reduction at the query-by-query level.

Should SaaS marketers stop investing in traditional SEO because of AI Overviews?

No. Ahrefs’s AI Overview citation research found 38% of AI Overview citations come from pages already ranking in Google’s top 10 for the query — down from ~76% in July 2025, as Google shifted toward query fan-out. Top-10 ranking remains a major driver of citation eligibility — disinvesting from SEO removes a substantial part of the foundation that earns the citation. The right adjustment is to maintain SEO investment, then layer citation-eligibility levers (schema markup, expert attribution, original data, 30-day freshness) on top of the rank.

What content format gets cited in AI Overviews?

Content with a direct answer in the first paragraph below each H2, valid Article or FAQPage schema in the server-rendered HTML, named-expert attribution where claims are non-obvious, and first-party data where available. ConvertMate’s analysis of 80M+ AI citations measured the levers: 67% citation uplift from valid schema, 41% from expert quotes, 4.1× from original data, 3.2× from freshness within 30 days. Direct-answer formatting (answering the question before adding context) is what makes the content extractable.

Does zero-click search hurt SaaS lead generation?

Volume goes down; conversion rate of the clicks that remain goes up. Seer Interactive measured ChatGPT-referred sessions converting at 15.9%, Perplexity at 10.5%, and Google at 1.76% — roughly 9× the conversion rate for AI-referred traffic. Net impact on pipeline depends on the trade between the click-rate compression and the intent-quality lift. Most SaaS businesses with comprehensive technical documentation see net-neutral or net-positive pipeline impact when they shift measurement and optimisation toward citation share rather than click volume. Lead-attribution windows need extending to 60-90 days to capture the longer path from zero-click exposure to brand search to conversion.

How should I measure zero-click impact on my SaaS site?

Three layers, in priority order: (1) Citation visibility — a monthly prompt-matrix of 30-50 prospect-equivalent queries run against ChatGPT, Perplexity, Claude, and Google AI Overviews, with share-of-citation tracked against competitors. (2) AI-referred session quality — GA4 channel grouping to separate AI-referred sessions, with conversion rate and demo-request rate as the key metrics. (3) Brand search lift — monthly tracking of branded GSC impressions and direct-traffic baseline; an inflection upward without paid-media spike indicates zero-click exposure converting to active consideration. Traditional metrics (organic sessions, rankings) remain useful but drop in the priority order — they measure inputs, not the citation outcome that matters.