For more than two decades, looking for information online has been almost synonymous with using Google. In 2026, people still rely heavily on Google, but they also ask ChatGPT, read Google AI Overviews, search YouTube, explore Reddit discussions and compare local businesses through Maps and reviews.
The important change is not that one platform has replaced all the others. Search is fragmenting: people use different tools for different jobs, and more answers are delivered without a visit to a website. For businesses, that changes how visibility should be earned and measured.
This article brings together international clickstream studies, Australian consumer research, usability studies and Google’s own statements. The aim is to separate what the evidence actually shows from claims that AI has already made traditional SEO irrelevant.
Key search statistics in 2026
- 68.01% of US Google searches ended without a click during January – April 2026, according to SparkToro and Similarweb.
- 73.7% of searches across 41 selected major sites on US desktop took place on Google in Q4 2025, according to SparkToro and Datos.
- 3.2% of measured searches and AI prompt sessions across those sites happened on AI tools.
- 13.6 million Australians aged 14+ (58%) used an AI tool in an average four-week period during Q1 2026, according to Roy Morgan.
- 71% of surveyed Australian adults said their search-engine use was about the same as a year earlier, according to Conjointly.
These figures describe different activities and populations. User adoption, search share, revenue and clicks are not directly interchangeable measurements.
1. Google still dominates measured search activity
When someone says “everybody searches on ChatGPT now”, it’s worth asking what they mean by searching. Are they counting people who have tried a product, website visits, individual prompts or genuine information-seeking sessions?
In March 2026, SparkToro and Datos published an analysis of search activity across 41 selected domains in the US and EU/UK, using desktop panel data. In the US Q4 2025 sample, traditional search engines represented around 80% of measured searches, commerce sites around 10%, social platforms around 5.5% and AI tools 3.2%. Google alone accounted for 73.7%.
This is not a global market-share estimate. The analysis excludes mobile-app behaviour, covers an editorially selected set of sites and counts AI prompt sessions alongside other searches. Even with those qualifications, it demonstrates the sheer scale of traditional search compared with standalone AI tools.
It also reminds us that a lot of searching already takes place beyond Google: Amazon, YouTube, marketplaces and social platforms answer different types of user needs.
Google’s own business is still growing
Alphabet reported 17% year-on-year growth in Google Search and Other revenue in Q2 2026. Google also described increasing engagement with AI-powered search experiences. The distinction matters: revenue growth does not prove that search volume grew by 17%, nor that every website received more clicks. But it certainly does not fit the simple narrative of a search business in collapse. See Google’s Q2 2026 CEO remarks.
2. Australians are adopting AI quickly – without necessarily abandoning search engines
Australian adoption figures make the scale of AI clear. Roy Morgan estimated that 13.6 million people aged 14 and over used an AI tool in an average four weeks during the March quarter of 2026. ChatGPT led with 10.5 million users, followed by Gemini (5 million), Copilot (4 million), Canva Magic Studio (1.4 million) and Claude (777,000).
The largest reported usage rates were among people aged 25-34 (74%) and 35-49 (72%), ahead of 18-24 year-olds (68%). This is no longer just a student or early-adopter trend.
But AI tool usage is broader than search. People also use these tools for writing, work, programming, image generation and document analysis. We should not interpret every AI user as a person who has abandoned Google.
A separate Telsyte study, reported by IT Brief, suggested that the proportion of Australians identifying AI as their main method of finding online information rose from 5% to 12%. That is meaningful growth, even if it remains a minority of consumers.
Meanwhile, most Australians say their search habits are unchanged
Conjointly surveyed 1,351 Australian adults in March 2026: 71% said they used search engines about as much as a year earlier, 23% reported using them more, and 7% less. These are self-reported changes, not observed search logs, and the displayed figures are rounded. Still, those reporting more search outnumbered those reporting less. See Conjointly’s March 2026 study.
Both trends can be true simultaneously. People can turn to AI for new activities while continuing to search Google to locate shops, navigate to websites, compare providers and buy products.
3. Search is growing more “zero-click”
For website owners, this may be a more consequential change than the competition between ChatGPT and Google. Many searches now end with an answer, phone number, map, weather result, product information or summary displayed directly on the search platform.
SparkToro and Similarweb estimated that 68.01% of US Google searches ended without a click during January – April 2026. An earlier SparkToro/Datos estimate for 2024 was 60.45%.
Not every zero-click search is caused by AI. Google has answered simple questions directly through maps, featured snippets, knowledge panels and other search features for years. Some users also reformulate their query or stop browsing for reasons unrelated to AI.
AI Overviews are associated with fewer traditional clicks
A Pew Research Center study of US browsing behaviour in March 2025 found that users clicked a traditional Google result in 15% of visits without an AI summary, compared with 8% of visits with an AI summary. Just 1% of visits with a summary led to a click on a source link inside that summary.
For example, a person researching why a tap is dripping may get enough of an answer from a summary to avoid visiting a plumber’s article. By contrast, someone who urgently needs a plumber has to select an actual provider. That difference in intent is crucial.
It’s now entirely possible for a website to maintain visibility for informational questions while receiving fewer visits from those questions. The appropriate response isn’t to dismiss all falling traffic as “AI”, but to investigate which queries and pages lost clicks and what those searches were supposed to achieve.
4. Search intent matters more than the platform
One of the most useful questions a business can ask is not simply “Where do people search?” but “What are they trying to accomplish?”
| Intent | Example | Common destinations |
|---|---|---|
| Informational | Why is my hot water system leaking? | Google, AI Overviews, ChatGPT, Gemini |
| Local provider | Emergency electrician near me | Google Search, Maps, reviews, AI assistants |
| Product research | Best solar panels for a Brisbane home | Google, YouTube, Reddit, AI tools |
| Transaction | Buy replacement pool pump | Google Shopping, retailers, marketplaces |
| Inspiration | Small outdoor kitchen ideas | Pinterest, Instagram, TikTok, YouTube |
| Navigation | Business opening hours | Search, Maps, official websites |
These are typical patterns, not rigid divisions. People can ask an AI assistant for local recommendations, then verify businesses on Maps; they can also research a product on YouTube before returning to Google to purchase it.
Nielsen Norman Group’s user research provides a useful qualitative perspective: participants often used AI-generated answers alongside conventional results, product details, images and websites. A University of Melbourne-linked SIGIR 2026 eye-tracking study offers another way of examining attention and scanning behaviour. Neither should be turned into a universal claim about all consumers.
Why local business searches are different
A person choosing a dentist, physiotherapist, accountant or electrician needs more than a generic answer. They may want a nearby provider, evidence of expertise, recent reviews, opening hours and a reliable way to make contact. Google Search and Google Maps are especially useful because they combine these practical details in one place.
AI assistants increasingly participate in local discovery, but a recommendation may simply start the process. The customer may still validate the recommendation through the Google Business Profile, independent reviews or the website. BrightLocal’s local consumer review research is useful context for the continuing importance of trust and reviews.
5. How AI search tools find and use information
It helps to distinguish what a language model learned during training from what an AI-powered search retrieves when a question is asked. The model’s training teaches patterns and associations; it is not a permanently current directory of every business. Search-enabled systems may retrieve fresh material from webpages and other sources to help ground their answers.
This retrieval process is sometimes described as retrieval-augmented generation (RAG). Different services use different infrastructure, indexes, search providers and citation methods; not every answer uses live retrieval.
This is why core SEO work remains relevant. Search systems need to be able to find important pages, understand what they describe and decide whether the information is pertinent. Google explicitly says that existing SEO fundamentals continue to apply to its AI features. See Google Search Central: AI features and your website.
Entities, schema and business identity
An entity is a recognisable thing: a company, person, location, product or service. A business might be represented by a website, Google Business Profile, professional directories, LinkedIn profile, media references and customer reviews. Clear, consistent information helps systems associate those references with the same real-world organisation.
Appropriate Schema.org structured data – such as Organization, LocalBusiness, Person, Service and Product – can make selected facts machine-readable. The sameAs property can identify appropriate external references to an entity. Google’s organisation structured-data documentation explains that this markup can assist its understanding and disambiguation.
Schema does not guarantee an AI citation or recommendation. Google says no special AI-specific markup is required for AI Overviews or AI Mode. Structured data supports clarity, but isn’t a substitute for crawlable content, genuine expertise, relevance and reliable third-party information.
Why off-site references matter
Independent references may help users and information systems establish that a business exists and has expertise beyond what it claims on its own website. Useful examples include relevant industry associations, genuine reviews, expert contributions and editorial coverage.
SparkToro’s commentary on appearing in AI answers explores links between brand visibility and external mentions, but the evidence is still developing. No universal formula guarantees inclusion in ChatGPT, Gemini or another model. Manufactured reviews, irrelevant directory spam and mass-produced content remain poor strategies.
6. Being visible across platforms does not mean posting everywhere
Search now happens through Google, Maps, ChatGPT, YouTube, Reddit, marketplaces, social platforms and many other services. Yet a small business does not need to spend equally across all of them.
A Brisbane plumber will usually have different priorities from a fashion retailer or specialist B2B manufacturer. For the plumber, commercially important service pages, local visibility and reviews may be pivotal. For a retailer, product availability, marketplace visibility, customer reviews and demonstration videos may matter more. For a technical B2B company, clear specifications, case studies and credible industry references can be especially useful.
Rather than pursue every platform, identify the places customers use when they first notice a problem, research their options, compare providers and decide to enquire or buy. Then invest where visibility can genuinely influence that journey.
7. How should businesses measure search performance now?
Rankings, impressions, clicks and organic sessions remain important, but they are not all direct measures of commercial value. In an environment with AI answers, Maps interactions and zero-click discovery, the complete picture requires more than a traffic chart.
A sensible measurement framework includes qualified enquiries, leads and sales; organic performance of relevant service or product pages; visibility for commercially important searches; Google Business Profile calls and website interactions; brand-search demand; and identifiable AI referral traffic where it exists.
I would also be cautious about two popular reporting shortcuts. A whole-site average position in Google Search Console may combine numerous irrelevant or low-intent queries. And traffic estimates from tools such as Ahrefs or Semrush are useful modelling exercises, not observed analytics, especially for small local businesses with limited keyword-volume data.
AI mention-monitoring tools can add directional insight, but repeated prompts can produce different recommendations. A brand’s presence in a set of test questions does not establish the exact proportion of real customers who encounter it.
Most importantly, a decline in informational clicks and a decline in high-intent enquiries are different business problems. They should be investigated and explained separately.
8. Practical priorities for businesses in 2026
- Protect the SEO foundation. Keep important pages accessible, fast enough to use, logically organised and clear about the services or products offered.
- Strengthen local signals where relevant. Maintain the Google Business Profile, accurate contact and location information, genuine reviews and useful local landing pages.
- Clarify entities and relationships. Use consistent business information and appropriate structured data; identify real people, expertise, products and services.
- Be useful beyond your own website. Prioritise credible industry references, demonstrations, reviews and channels where customers actually conduct research.
- Measure business outcomes. Evaluate leads, calls, sales and commercially important visibility alongside overall traffic and rankings.
Conclusion: Search is not disappearing – it is becoming more distributed
I’ve worked in SEO since 2010, through major changes in Google’s algorithms, mobile search, local results and the rise of featured answers. AI is a significant new development because it changes not only how information is found, but how much of it people consume before ever visiting a website.
The research doesn’t support abandoning Google, nor does it support pretending nothing has changed. Traditional search remains dominant in the broad search-activity studies, AI adoption is growing rapidly, and more Google searches end without a website click.
Businesses don’t need to be everywhere. They need to be present, credible and discoverable wherever their potential customers search. Good SEO remains central to that effort, but strong business information, independent reputation and thoughtful platform selection matter more than ever.
Will Mullins is a Brisbane SEO consultant specialising in technical SEO, local SEO, ecommerce SEO and content strategy.
Research and methodology notes
Research discussed here spans 2025-2026. Results come from different markets and methods: desktop panel studies, mobile-and-desktop clickstream panels, browser observation, consumer surveys, qualitative usability research and company financial disclosures. They cannot be combined into a single global market-share figure. The zero-click comparison uses different data providers; AI-tool usage includes non-search activities; and observational CTR comparisons do not establish causality. Statistics were reviewed in October 2026.
Leave a Reply