TL;DR

In February 2024 Gartner predicted that by 2026 traditional search engine volume would drop 25%, as search marketing lost share to AI chatbots and other virtual agents. Public data available in October 2026 does not show Google search volume collapsing, but it does show fewer clicks per search and steady growth in AI tool usage. That makes visibility inside AI answers a metric marketing teams need to track. A practical measurement stack combines a repeated prompt panel for share of voice in AI, AI-assistant referral traffic in analytics, citation data from Bing Webmaster Tools, and brand-demand signals such as branded search.

What Gartner actually predicted

The prediction is widely quoted, often loosely. The original Gartner press release, issued from Stamford, Connecticut on 19 February 2024, is headlined Gartner Predicts Search Engine Volume Will Drop 25% by 2026, Due to AI Chatbots and Other Virtual Agents. Its opening sentence says that by 2026, traditional search engine volume will drop 25%, with search marketing losing market share to AI chatbots and other virtual agents. The analyst quoted is Alan Antin, Vice President Analyst at Gartner, and the release draws on Gartner's research Predicts 2024: How GenAI Will Reshape Tech Marketing.

Three details are worth noting. The prediction concerns search engine volume, not website traffic. It was framed for technology marketers. And its advice was not to abandon search but to rethink channel strategy, produce unique and useful content, and demonstrate expertise, experience, authoritativeness and trustworthiness.

How the prediction compares with the data

With 2026 nearly over, the public evidence is mixed. Search volume has not visibly fallen by a quarter:

  • In Alphabet's second-quarter 2026 results (22 July 2026), Sundar Pichai said AI features were driving Search query growth, and Google Search and other revenue grew 17% year on year.
  • Datos's Q1 2026 State of Search report, summarised by PPC Land, found that AI tools accounted for 1.65% of US desktop events in Q1 2026, up from 1.31% a year earlier, and 1.72% in the EU and UK, up from 0.98%. Google searches per US searcher rose between December 2025 and March 2026. The data is desktop-only.

Clicks, however, have fallen. SparkToro's June 2026 analysis of Similarweb data found that 68% of US Google searches between January and April 2026 ended without a click. The fair reading: Gartner was right that answers are moving into AI, but so far the shift shows up more in fewer clicks and AI answers inside Google than in a collapse of search itself. Either way, traffic alone no longer tells you whether you are visible.

What to measure: four layers

Aleyda Solis's AI Search Optimization Checklist (updated May 2026) offers one of the most rigorous public frameworks for this. It separates observed data, owned proxies, third-party proxies and modelled estimates, and insists they are reported separately with stated confidence. The four measurement layers below follow that logic.

Layer 1: share of voice in AI answers

Build a prompt panel: the questions your buyers actually ask at each stage, weighted towards non-branded prompts, since that is where discovery happens. Solis suggests a minimum viable start of 30 to 50 commercially relevant prompts run across the two or three AI platforms your audience uses most. Because AI answers vary between runs, repeat prompts and treat each output as a sample, not a fixed ranking. Her five "Presence KPIs" are a strong template:

  • Prompt coverage: the share of tracked prompts in which your brand appears.
  • Recommendation rate: the share of appearances where the AI actually recommends you.
  • Linked citation rate: the share of appearances that include a clickable link to you.
  • Comparative win rate: how often you are preferred when compared with competitors.
  • Representation accuracy: how often you are described correctly.

"Are we endorsed, or just mentioned?" — Aleyda Solis, The AI Search Optimization Checklist

Solis also warns against blending platforms such as AI Overviews, ChatGPT, Perplexity, Gemini and Copilot into one score, because the average hides real differences.

Layer 2: referral traffic from AI assistants

Clicks from ChatGPT, Gemini, Perplexity, Copilot and similar tools can be tracked in analytics through their referrer domains. According to reporting by TechWyse and PPC Land, Google added an "AI Assistant" channel to GA4's default channel group in May 2026. Treat these numbers as a floor: visits from apps often arrive without a referrer and are logged as direct, and clicks from Google's own AI Overviews and AI Mode usually appear as Google organic traffic. PPC Land cites a Similarweb panel study of US desktop users from July to December 2025 in which only 8.8% of AI-influenced visits came through a direct AI referral; more arrived later via search.

Layer 3: citation data from the platforms

Google's AI features documentation says traffic from AI Overviews and AI Mode is included in the Search Console Performance report under the "Web" search type, without a separate breakdown. Microsoft goes further: since February 2026, Bing Webmaster Tools' AI Performance report (in public preview) shows how often your content is cited in Microsoft Copilot, Bing's AI summaries and select partner integrations, which pages are cited and a sample of the grounding queries used. Microsoft notes that citation counts do not indicate placement or importance within an answer.

Layer 4: brand demand and mentions

Because AI influence often converts through another channel, track the outcomes it should move: branded search volume in Search Console, direct traffic to pages that AI answers surface, enquiries that mention AI tools (add the option to your "how did you hear about us" field), and mentions of your brand on the review sites, media and communities AI systems draw on.

A 90-day measurement playbook

  1. Weeks 1–2: Agree the 30–50 priority prompts with sales and customer service, grouped by product, market, language and funnel stage.
  2. Weeks 2–3: Run each prompt several times on your two or three key platforms and record the five Presence KPIs as a baseline.
  3. Week 3: Check GA4 for AI assistant traffic, verify Bing Webmaster Tools and Search Console access, and add AI as a lead-source option in your CRM.
  4. Weeks 4–10: Fix the biggest gaps: inaccurate descriptions, missing comparison content, weak third-party presence.
  5. Weeks 11–12: Re-run the panel and compare like with like by platform and prompt group.
  6. Every quarter: Report observed, proxy and modelled figures separately, with confidence levels.

Common measurement mistakes

The biggest mistake is false precision: declaring victory or crisis from a single prompt run, a single platform or a blended visibility score. Another is presenting modelled revenue from AI search as observed fact. A third is relying solely on referral traffic, which systematically undercounts AI influence. Finally, quoting Gartner's 25% figure as if it were a measured outcome misrepresents it: it was a forecast, and the data so far tells a more nuanced story.

Key takeaways
  • Gartner predicted in February 2024 that traditional search engine volume would drop 25% by 2026; public data does not yet show that collapse.
  • Clicks per search have fallen and AI tool usage is growing, so traffic alone understates or misreads visibility.
  • Track share of voice in AI with a repeated prompt panel and clear KPIs.
  • Use GA4 AI referrals, Bing's AI Performance report and Search Console, knowing each undercounts.
  • Report observed, proxy and modelled numbers separately.

FAQ

Did search volume really drop 25% by 2026?
Not according to the public data we found. Alphabet reported Search query growth in mid-2026, and Datos data showed AI tools still accounting for under 2% of desktop activity in Q1 2026. What has clearly fallen is the share of searches that lead to a click.
Can we buy a tool that shows our AI visibility?
Several tools sample AI answers for you, which saves time. Whatever you use, check how many prompts, platforms and repeats it runs, and treat the output as a sample rather than a ranking, as Aleyda Solis advises.
Why is our AI referral traffic so low?
AI assistants send relatively few clicks, app traffic often loses its referrer, and Google's AI features report as organic search. Low referral numbers do not necessarily mean low visibility; check your prompt panel and branded search trends too.

Sources

Know where you stand in AI answers

Talk to JRK about setting up a prompt panel and reporting that shows how visible your brand really is across AI assistants.

Book a free growth session →

Want this working for your brand?

Book a free 30-minute growth session. We'll look at your funnel and tell you what we'd do first — honestly.

Book a growth session →