Voice Search & Conversational Search Statistics UK 2026

Statistics 21–30 — Conversational Search & Generative AI Adoption

The most important change in conversational search is no longer confined to smart speakers or traditional voice assistants.

Generative AI has introduced natural-language interaction into mainstream information discovery. Users can ask complex questions, request comparisons, refine previous answers and continue a search journey without repeatedly reformulating individual keyword queries.

The UK data now shows that this behaviour has moved well beyond a small early-adopter audience.

21. 54% of UK adults now use AI tools such as ChatGPT, Copilot or Gemini

This is an important threshold because generative AI is no longer limited to a specialist technology audience.

More than half of UK adults now report using tools capable of conversational interaction, natural-language information retrieval and AI-assisted discovery.

Search implication: Conversational interfaces have reached sufficient adoption to influence mainstream search and research behaviour.

Source: Ofcom, Adults’ Media Use and Attitudes 2026.

22. 79% of UK 16–24-year-olds use AI tools

Younger adults are therefore substantially ahead of the total adult population in adoption.

This matters strategically because search habits developed among younger audiences can influence expectations about how information systems should work more broadly.

Users accustomed to conversational AI may be less willing to reduce complex needs into a sequence of short keyword queries.

Audience implication: Conversational search behaviour is particularly advanced among younger UK adults.

Source: Ofcom, Adults’ Media Use and Attitudes 2026.

23. 74% of UK adults aged 25–34 use AI tools

The high adoption rate extends conversational AI well beyond students and the youngest adult demographic.

The 25–34 age group contains substantial numbers of working professionals, business buyers, consumers, renters, homeowners and parents.

Their adoption therefore increases the commercial relevance of conversational search across multiple sectors.

Commercial implication: AI-assisted discovery is increasingly relevant to mainstream consumer and professional decision journeys.

Source: Ofcom, Adults’ Media Use and Attitudes 2026.

24. 59% of UK adults used generative AI within the previous three months

The same research found that 56% had recently used AI capable of producing human-like text or speech in response to queries, while 54% had used AI-powered digital assistants capable of understanding natural language.

This independently supports the wider picture of conversational interaction becoming common across the UK population.

Behaviour implication: Natural-language interaction with AI is becoming an established digital behaviour rather than an occasional novelty.

Source: UK Department for Science, Innovation and Technology, Public Engagement Survey 2025/2026.

25. 15.8 million UK online adults visited an AI chatbot in June 2025

That represented 32% of the UK online adult population.

The measured services included ChatGPT, Copilot, Gemini, DeepSeek, Perplexity, Claude.ai and Grok.

This is measured website and app reach rather than a survey asking whether people remember using AI.

However, Ofcom cautions that visiting an AI chatbot does not prove that every visit involved search activity.

Measurement implication: AI-chatbot reach is now significant, but AI-tool usage and AI search usage should still be distinguished.

Source: Ofcom, The Era of Answer Engines, citing Ipsos iris, June 2025.

26. ChatGPT alone reached 13 million UK online adults in June 2025

It was substantially ahead of other standalone AI chatbot services in the same dataset.

Microsoft Copilot reached approximately 2.6 million UK online adults, while Google Gemini reached approximately 1.7 million.

Platform implication: ChatGPT had already established a materially larger standalone UK audience than competing AI chatbot services by mid-2025.

Source: Ofcom, The Era of Answer Engines, citing Ipsos iris.

27. ChatGPT’s measured UK adult reach increased by approximately 194% in one year

That represents growth of approximately 194%, taking the measured audience to almost three times its size one year earlier.

The scale of the increase demonstrates how quickly AI-assisted information discovery can develop once a conversational interface becomes widely known.

Growth implication: Conversational AI adoption has progressed at a pace substantially faster than many established digital behaviours.

Source: Ofcom, The Era of Answer Engines, citing Ipsos iris.

28. ChatGPT received 1.8 billion UK visits in the first eight months of 2025

The equivalent period in 2024 generated approximately 368 million visits.

That means measured visit volume was approximately 4.9 times higher, an increase of roughly 389% year on year across the comparison period.

Visit counts differ from unique-user reach because one user can visit a service repeatedly, but the growth still demonstrates substantial increases in usage intensity.

Usage implication: Growth is occurring not only in the number of people exposed to conversational AI but also in the volume of interactions with those services.

Source: Ofcom, Online Nation 2025.

29. Around 30% of UK searches now display an AI Overview

This is strategically important because conversational and generative search behaviour is not limited to people actively choosing an AI chatbot.

AI-generated responses are increasingly being inserted into conventional search journeys.

Users can therefore experience answer-led search without consciously deciding to switch from a traditional search engine to an AI tool.

Search implication: Generative search adoption is occurring through both active chatbot use and passive integration into existing search interfaces.

Source: Ofcom, Online Nation 2025.

30. 53% of UK adults say they often see AI-generated search summaries

In many cases, users are not actively requesting an AI experience. The summaries are presented automatically within the search interface.

This represents a fundamental change from the earlier smart-speaker model.

Conversational search is no longer an alternative interface that users must consciously choose. Generative answers are increasingly being integrated into mainstream discovery itself.

Strategic implication: Businesses need to consider visibility inside answer-led search even if their target audience does not actively identify itself as an AI-search user.

Source: Ofcom, Online Nation 2025.

Statistics 31–40 — Query Behaviour, Local Discovery & Commercial Intent

Conversational search changes more than the interface used to enter a query.

It changes how much information users can provide at the beginning of the search, how easily they can refine a request and how many requirements can be included within a single interaction.

Instead of reducing a need to a short phrase such as Italian restaurant London, a user can describe party size, location, dietary requirements, preferred atmosphere, budget and timing in one request.

The same pattern applies to ecommerce, travel, professional services and product research.

31. AI Mode queries are now approximately three times longer than traditional Google searches

Earlier AI Mode testing in 2025 had already shown queries around two to three times longer than conventional searches.

The change is significant because longer queries allow users to express considerably more context, restrictions and intent before the system returns an answer.

A conventional keyword query may identify the category being searched.

A conversational query can identify the category, location, budget, preferences, exclusions and desired outcome simultaneously.

Search implication: Search optimisation increasingly needs to address complete user requirements rather than isolated keyword strings.

Source: Google Search, AI Mode behavioural data, 2025–2026.

32. Google AI Mode has surpassed one billion monthly users globally

This scale demonstrates that conversational search is no longer restricted to standalone AI chatbot platforms.

It is being incorporated directly into one of the world’s largest existing search ecosystems.

Platform implication: Conversational search is increasingly becoming a layer within mainstream search rather than a separate category of technology.

Source: Google, A New Era for AI Search, May 2026.

33. AI Mode query volume more than doubled every quarter after launch

This measures usage growth rather than simple availability.

A product can be technically accessible without becoming part of users’ normal behaviour. Rapid query growth provides stronger evidence that people are returning to the interface and expanding the types of questions they ask.

Adoption implication: The growth of conversational search is visible in interaction volume as well as user reach.

Source: Google Search, May 2026.

34. More than one in six AI Mode searches in the US now use voice or images

This is an important evolution of the original voice-search concept.

Users are no longer restricted to choosing between typing and speaking. They can combine natural-language questions with visual input and continue the search conversationally.

Multimodal implication: Future search optimisation increasingly needs to consider how products, places, objects and entities can be understood across text, voice and visual inputs.

Source: Google, How AI Mode Is Changing and Expanding the Way People Search, May 2026.

35. Image searches in AI Mode were growing by more than 40% month on month

Visual search allows users to begin with something they see rather than something they can easily describe.

The user can then add natural-language context through follow-up questions.

For retailers, property businesses, travel organisations, hospitality companies and other visually driven sectors, this creates new discovery pathways that do not begin with a conventional keyword query.

Visual discovery implication: Image quality, product information, structured data and clear entity association become increasingly important as search becomes multimodal.

Source: Google Search AI Mode behavioural data, May 2026.

36. Planning-related AI Mode queries grew 80% faster than AI Mode queries overall

Planning is particularly suited to conversational search because it usually involves multiple criteria and several stages.

Examples can include:

• Planning a holiday.
• Finding a restaurant.
• Organising an event.
• Comparing products.
• Creating an itinerary.
• Researching a major purchase.

Traditional search often requires several separate queries to complete these tasks. Conversational systems can maintain context across the journey.

Journey implication: Search is increasingly moving from finding individual pieces of information towards helping users complete multi-stage decisions.

Source: Google Search Trends analysis, May 2026.

37. Brainstorming queries in AI Mode grew 30% faster than overall AI Mode usage

This type of search differs substantially from classic navigational or informational search.

The user may not know exactly what they want when the interaction begins.

Instead, the system becomes part of the discovery process — helping the user generate ideas, refine criteria and compare possibilities.

Intent implication: Conversational search increasingly operates before a user has formed a precise keyword or final purchase requirement.

Source: Google Search AI Mode trends, May 2026.

38. UK searches for “when to book a table” surged 140% in 2026

The finding was published alongside the rollout of AI Mode’s restaurant-booking capabilities in the UK.

Users can now provide multiple conditions inside a single restaurant request — including date, time, party size, cuisine, dietary requirements and location — before being shown relevant available options.

This is a strong example of conversational search moving beyond information retrieval into local commercial action.

Local-search implication: Local discovery increasingly depends on complete, accurate business information that can satisfy multiple user constraints simultaneously.

Source: Google UK, Booking Restaurants in the UK Just Got Easier with AI in Search, April 2026.

39. Google’s Shopping Graph contains more than 50 billion product listings

AI Mode can use this product data to respond to conversational requests in which users describe what they want using natural language rather than fixed ecommerce filters.

For example, a user can specify fit, style, budget, colour or intended use inside one conversational query and then refine the results further.

Ecommerce implication: Product visibility increasingly depends on accurate machine-readable product attributes as well as conventional category and keyword optimisation.

Source: Google, AI Mode Shopping update, September 2025.

40. More than two billion Google product listings are refreshed every hour

Those updates can include pricing, availability, promotions and other product information.

This highlights an increasingly important difference between conversational commercial search and static content discovery.

When users ask for a recommendation based on budget, availability, location or another immediate constraint, freshness becomes part of relevance.

Commercial implication: Product feeds, inventory accuracy, price data and structured product information increasingly contribute to whether an organisation can satisfy highly specific conversational queries.

Source: Google Shopping / AI Mode product infrastructure data.

Statistics 41–50 — AI Search Interfaces, Follow-Up Questions & Future Discovery

The final ten statistics show how conversational discovery is becoming embedded inside mainstream search rather than remaining a separate activity confined to dedicated AI chatbots.

AI summaries now appear directly within search results. Users can continue from an initial answer into follow-up questions, combine text with images, voice or camera input and increasingly use AI systems to discover information that would previously have required several separate searches.

The strategic significance is therefore broader than “voice search”.

The emerging model is an interconnected search environment in which users move between traditional search, generated answers, conversational follow-ups and multimodal discovery.

41. 75% of UK online adults read AI-generated search summaries at least some of the time

Ofcom’s 2026 research presented participants with an example of an AI-generated search summary and asked about their engagement with this type of result.

The finding demonstrates that generated answers have already become part of ordinary information discovery for a large majority of UK internet users.

Search implication: AI-mediated discovery now reaches substantially more people than those who deliberately choose to use standalone AI chatbots.

Source: Ofcom, Adults’ Media Use and Attitudes 2026.

42. 42% of UK online adults read AI search summaries often or always

This is more significant than occasional exposure.

It indicates that generated summaries are becoming a regular part of information-seeking behaviour for a substantial proportion of the population.

Visibility implication: Organisations increasingly need to consider whether their information can be understood and selected within generated answers as well as conventional organic listings.

Source: Ofcom, Adults’ Media Use and Attitudes 2026.

43. 54% of people who do not use AI chatbots still read AI search summaries

This is one of the most strategically important findings in the report.

It shows that measuring ChatGPT, Gemini or Copilot usage alone significantly understates exposure to AI-mediated search.

A user can avoid standalone AI tools entirely and still interact regularly with generated answers embedded inside conventional search.

Measurement implication: AI-search adoption should not be estimated solely from chatbot usage.

Source: Ofcom, Adults’ Media Use and Attitudes 2026.

44. 62% of UK 16–24-year-olds read AI search summaries often or always

Younger users therefore show materially higher routine engagement than the 42% recorded across online adults overall.

This supports the wider evidence that younger audiences are adopting AI-mediated discovery particularly quickly.

Audience implication: Businesses targeting younger consumers should expect generated answers and conversational interfaces to play an increasingly prominent role in information discovery.

Source: Ofcom, Adults’ Media Use and Attitudes 2026.

45. 60% of UK 25–34-year-olds read AI search summaries often or always

This means regular use remains high beyond the youngest adult demographic.

The 25–34 age group includes substantial numbers of consumers making decisions around careers, housing, travel, financial products, professional services and major purchases.

Commercial implication: AI-mediated search is becoming relevant across high-value consumer and professional decision categories rather than remaining concentrated among students or technology enthusiasts.

Source: Ofcom, Adults’ Media Use and Attitudes 2026.

46. 57% of UK adults aware of AI would trust AI-written news less than human-written news

A further 27% said they would trust AI-generated news about the same, while only 7% said they would trust it more.

The finding highlights an important tension within conversational discovery.

People are increasingly consuming AI-generated answers while remaining cautious about AI-generated information.

Trust implication: Source transparency, identifiable evidence and authoritative attribution are likely to remain important as generated answers become more common.

Source: Ofcom, Adults’ Media Use and Attitudes 2026.

47. Even among AI users, 50% trust AI-written news less than human-written news

Regular use therefore does not automatically remove scepticism.

Users can find AI tools useful while still wanting evidence, recognisable sources and the ability to verify important claims independently.

Authority implication: AI visibility and trust are different objectives. Being surfaced in a generated response does not by itself establish user confidence.

Source: Ofcom, Adults’ Media Use and Attitudes 2026.

48. Google AI Overviews now reaches more than 2.5 billion monthly active users

The scale is significant because AI Overviews are integrated within conventional Google Search rather than requiring users to visit a separate chatbot interface.

For publishers and businesses, this creates a discovery environment where visibility can occur inside a generated answer, through cited links or through conventional organic search results on the same journey.

Ecosystem implication: AI-generated answers now operate at mainstream search-engine scale.

Source: Google Search, website-owner update, 2026.

49. AI Overviews have increased Google usage by more than 10% for the query types where they appear

The company has also reported growth in overall and commercial query volumes alongside expansion of its AI search experiences.

This challenges the assumption that generated answers necessarily eliminate further searching.

For some query types, Google reports that users search more after becoming accustomed to AI-assisted results.

Behaviour implication: Conversational and generated search can expand the number and complexity of searches rather than simply replacing existing keyword queries one for one.

Source: Google Search / Alphabet, AI Overviews usage data.

50. One in five UK adults now use AI tools to keep up with news

This is particularly significant because news has traditionally been accessed through publishers, broadcasters, search engines and social-media feeds.

AI assistants are now becoming another intermediary through which audiences discover and interpret current information.

The same Ofcom research found that UK adults encounter approximately 12 different news sources in an average month, rising to 18 among 25–34-year-olds.

Discovery implication: AI systems are becoming part of the information-distribution layer itself, not merely tools people use after reaching a publisher’s website.

Source: Ofcom, News Consumption in the UK 2026, September 2026.

CGO Media Analysis — What the 50 Voice & Conversational Search Statistics Tell Us

Taken together, the 50 statistics show that the most important change is not simply an increase in the number of people speaking searches into devices.

The larger transformation is the movement from short, isolated queries towards natural-language, contextual and increasingly multi-stage discovery.

Voice assistants helped establish the behavioural foundation. Generative AI and AI-integrated search are now extending that behaviour across text, voice, images and conventional search interfaces.

1. Voice Search Is No Longer the Whole Story

The data confirms that smart speakers and voice assistants have achieved substantial penetration in the UK.

However, the most strategically important development is that conversational behaviour has escaped the dedicated voice-device category.

Users can now interact naturally with:

• Smart speakers.
• Smartphones.
• Search engines.
• AI chatbots.
• Smart TVs.
• Vehicles.
• Visual-search interfaces.
• Multimodal AI systems.

This means organisations should avoid building a strategy around the narrow idea of optimising specifically for Alexa-style searches.

2. Natural-Language Search Is Becoming Mainstream

The adoption figures for ChatGPT, Gemini, Copilot and AI-generated search summaries show that conversational interaction is no longer limited to early adopters.

A majority of UK adults now use AI tools, while younger age groups report particularly high usage.

The effect on search behaviour is significant because these systems allow people to formulate requests in a way that more closely resembles ordinary speech.

Instead of learning how to phrase a search for the machine, users increasingly expect the machine to understand how they naturally describe a problem.

3. Queries Are Becoming Richer Expressions of Intent

Google’s AI Mode data shows that conversational queries are materially longer than conventional searches.

Longer queries matter because they allow users to provide more context before the system begins evaluating possible answers.

A traditional search might be:

best hotel Manchester

A conversational query might instead ask:

Which central Manchester hotels are suitable for a two-night business trip, have reliable Wi-Fi, parking nearby and good restaurants within walking distance?

The second query contains far more information about the user’s actual decision criteria.

4. Search Is Moving From Retrieval Towards Evaluation

Traditional search primarily helped users locate documents.

Conversational search increasingly helps users evaluate options.

A system may be asked to:

• Compare several providers.
• Recommend suitable products.
• Filter businesses according to constraints.
• Explain trade-offs.
• Construct itineraries.
• Summarise evidence.
• Identify the best match for a specific requirement.

This changes the role of optimisation.

It is no longer enough for a webpage simply to contain the phrase used in the search.

The organisation needs to provide enough evidence for a system to understand what it offers, who it serves, where it operates and why it may be relevant to a particular requirement.

5. Direct Answers Reduce the Number of Visible Choices

Voice assistants established an important characteristic of answer-led discovery: the user may receive one selected response instead of a page containing multiple alternatives.

AI-generated answers extend the same principle into conventional search.

A generated response can synthesise information from several sources, highlight a small number of recommendations or answer the question without requiring the user to open every underlying page.

This increases the strategic value of being selected as a source or candidate.

6. Entity Clarity Becomes More Important

Ofcom’s voice research shows that devices can misinterpret requests and deliver the wrong station or service.

Conversational AI faces a related challenge at much larger scale: it must determine which organisation, product, person, service or location the user actually means.

Clear entity signals can therefore become increasingly important.

These may include:

• Consistent organisation names.
• Clear service definitions.
• Accurate location information.
• Structured data.
• Consistent third-party profiles.
• Strong About and organisation information.
• Author and expert attribution.
• Reliable external references.

7. Structured Data Becomes More Commercially Important

Conversational search increasingly involves constraints that machines need to evaluate directly.

Examples include:

• Price.
• Availability.
• Opening hours.
• Product specifications.
• Location.
• Delivery options.
• Ratings.
• Event dates.
• Service areas.

Google’s Shopping Graph illustrates the scale of this transition, with tens of billions of product listings and billions of updates every hour.

The implication extends beyond ecommerce.

Any organisation whose suitability depends on attributes or availability benefits from making those facts clear, consistent and machine-readable.

8. Local Search Becomes More Conversational and Transactional

The evolution of restaurant booking provides a useful example.

A local search no longer needs to end with a list of businesses.

The user can potentially specify location, party size, time, cuisine and dietary requirements before the system identifies suitable options and moves directly towards a booking.

The same model can extend to:

• Hotels.
• Healthcare providers.
• Professional services.
• Property.
• Retail.
• Travel.
• Events.

9. Trust Becomes More Important as Answers Become More Synthesised

The Ofcom trust data reveals an important contradiction.

Consumers are increasingly using AI-generated answers while remaining cautious about information generated by AI.

This means visibility alone is insufficient.

Users and systems both benefit from information supported by:

• Identifiable sources.
• Original research.
• Expert authorship.
• Evidence.
• Dates.
• Clear methodology.
• Consistent facts.
• Independent corroboration.

This is especially important in higher-trust sectors such as healthcare, financial services, legal services and professional advice.

10. Voice, AI Search and Traditional SEO Are Converging

The evidence does not support treating voice search, conversational AI and traditional search as completely separate disciplines.

The same underlying information may now need to serve:

• A conventional Google result.
• An AI Overview.
• An AI Mode response.
• A ChatGPT answer.
• A voice-assistant response.
• A visual-search result.
• A local recommendation.

The interface changes, but many of the underlying requirements remain connected:

clear information, strong entities, reliable sources, useful content, technical accessibility and consistent factual signals.

What UK Businesses Should Do in 2026

The evidence in this report shows that voice search, conversational AI and traditional search are increasingly converging.

For UK businesses, the practical objective should therefore not be to create a separate “voice SEO” strategy in isolation.

The stronger approach is to make the organisation, its services, products, locations, expertise and evidence sufficiently clear for search engines and AI systems to understand, evaluate and recommend across multiple interfaces.

1. Write for Complete Questions, Not Just Keywords

Conversational queries are longer and contain more context than many conventional searches.

Businesses should therefore expand important pages beyond a narrow target phrase and answer the wider questions a potential customer is likely to ask.

That can include:

• Who the service is suitable for.
• Where the organisation operates.
• What the service includes.
• What it costs.
• How long it takes.
• What alternatives exist.
• What makes one option more suitable than another.
• What evidence supports the claims being made.

2. Strengthen Entity Clarity Across the Website

Conversational systems need to understand exactly which organisation, service, person, location or product they are evaluating.

Businesses should therefore improve consistency around:

• Organisation name.
• Brand descriptions.
• Office and service locations.
• Team members and experts.
• Products and services.
• Parent and subsidiary relationships.
• Sector specialisms.
• Contact information.

The same information should remain consistent across the website and important third-party profiles.

3. Use Structured Data Where It Describes Real Page Content

Structured data can help machines understand important attributes and relationships.

Depending on the organisation, this can include:

• Organisation information.
• Local business data.
• Products.
• Offers.
• Prices.
• Events.
• Articles.
• Authors.
• Reviews where eligible.
• Breadcrumbs.

Structured data should reflect content genuinely visible and supported on the page rather than being added purely to create additional machine-readable claims.

4. Make Product and Service Attributes Explicit

Conversational search increasingly works through combinations of user requirements.

A system may need to determine whether a product or service satisfies conditions involving:

• Price.
• Location.
• Availability.
• Size.
• Features.
• Eligibility.
• Delivery.
• Opening hours.
• Specialisms.
• Service area.

If these attributes exist only inside vague marketing language, they can be harder for users and machines to evaluate.

5. Build Strong Question-and-Answer Coverage

FAQs remain useful when they answer genuine customer questions rather than being added mechanically for SEO.

Good FAQ content should address:

• Specific objections.
• Common comparisons.
• Eligibility questions.
• Costs.
• Timelines.
• Service boundaries.
• Local availability.
• Implementation questions.

These sections can help support conversational search because they often mirror the natural form of user questions.

6. Strengthen Local Information for Conversational Discovery

Local conversational queries can contain multiple constraints at once.

Businesses should make important local information easy to identify, including:

• Full address.
• Service area.
• Opening hours.
• Telephone number.
• Booking options.
• Parking or accessibility information.
• Nearby areas served.
• Location-specific services.

Local profiles should remain consistent with the information presented on the website.

7. Create Evidence That Can Support Recommendations

Conversational search frequently asks systems to compare or recommend.

Recommendation confidence is likely to be stronger where an organisation can demonstrate evidence beyond its own marketing claims.

Useful evidence can include:

• Original research.
• Statistics.
• Case studies.
• Named experts.
• Independent coverage.
• Industry references.
• Methodology pages.
• Documented outcomes.

This is especially important where the query asks which organisation is suitable, credible or experienced.

8. Improve Source Transparency and Authorship

Ofcom’s research shows that users remain cautious about AI-generated information.

Content therefore benefits from clear evidence about who produced it and why the source should be trusted.

Important research and advice pages should make authorship, methodology, dates and supporting sources easy to identify.

9. Optimise for Follow-Up Questions

Traditional SEO often treats each query as a separate event.

Conversational search allows the user to continue from the previous answer.

For example:

Initial question:

“Which accounting firms in Manchester specialise in ecommerce businesses?”

Follow-up:

“Which of those also works with companies selling internationally?”

Further refinement:

“Which one publishes useful guidance about VAT and cross-border payments?”

The organisation therefore needs depth beyond the initial category match.

10. Measure More Than Conventional Organic Rankings

As search becomes more answer-led, ranking reports alone provide an incomplete view of visibility.

Businesses should increasingly monitor:

• Traditional search visibility.
• AI Overview appearances.
• AI citations where measurable.
• Branded search demand.
• Referral traffic from AI platforms.
• Local profile visibility.
• Search Console performance.
• Conversion from AI-assisted journeys.

Some AI-platform measurement remains incomplete, so businesses should avoid inventing precision that current analytics cannot support.

Voice & Conversational Search Trends UK 2026–2027

The evidence reviewed throughout this report points towards a search environment in which the boundaries between traditional search, voice search and conversational AI continue to weaken.

Voice remains an important input method, but the larger development is the emergence of search systems capable of maintaining context, interpreting multiple requirements, combining different forms of input and assisting users through increasingly complex decisions.

The trends below are therefore presented as evidence-led directions rather than guaranteed predictions.

Trend 1 — Voice Search Becomes One Part of Multimodal Search

The original voice-search model focused primarily on spoken questions directed at smartphones or smart speakers.

That model is becoming broader.

Modern AI search can increasingly accept:

• Typed questions.
• Voice input.
• Images.
• Camera input.
• Files.
• Video.
• Existing browser context.

Google’s 2026 Search updates explicitly move in this direction, allowing users to search across several input types inside the same AI-assisted environment.

Trend 2 — Follow-Up Questions Become a Normal Part of Search

Traditional search typically required users to reformulate each query manually.

Conversational search preserves context.

Google now allows users to move directly from an AI Overview into a follow-up conversation in AI Mode, while retaining the context of the original question.

This changes search from a sequence of disconnected queries into a connected research journey.

Trend 3 — Queries Become Longer and More Specific

Google’s AI Mode data already shows users asking questions significantly longer than traditional search queries.

This allows users to specify several requirements at once rather than splitting a decision across multiple searches.

A search for a hotel, product, restaurant or professional service can increasingly contain:

• Location.
• Budget.
• Timing.
• Preferences.
• Restrictions.
• Features.
• Intended use.

This places greater emphasis on content that clearly describes attributes and suitability.

Trend 4 — AI Search Moves Further Into Commercial Action

Conversational search is beginning to move beyond research and recommendation.

Google’s UK rollout of agentic restaurant-booking functionality demonstrates how a conversational query can progress from a detailed requirement directly towards a transaction.

The user can specify party size, cuisine, location, dietary requirements and timing before receiving relevant options and booking links.

This model could become increasingly relevant across other transactional sectors.

Trend 5 — Search Agents Become Part of Ongoing Discovery

Google has begun introducing AI agents within Search that can continue monitoring information after the original query.

This represents a substantial change from conventional search.

Instead of asking the user to repeat the same search later, an agent can potentially monitor changing information across sources and identify relevant updates.

This creates a search environment in which discovery can continue even when the user is not actively conducting another query.

Trend 6 — Visual Search and Conversational Search Converge

Some user needs are difficult to express accurately through words alone.

A person may know the type of furniture, clothing, plant, product or destination they want when they see it but struggle to describe it.

AI Mode increasingly allows users to begin with an image and then refine the discovery process conversationally.

This produces a combined visual and language-based journey.

Trend 7 — AI Search Becomes Embedded Rather Than Optional

Ofcom’s UK data shows that AI-generated search summaries already reach people who do not consider themselves users of AI chatbots.

This is likely to become increasingly important.

Users may experience conversational or generated search without deliberately selecting a separate AI product.

The transition is therefore happening inside existing search behaviour rather than only through migration to new platforms.

Trend 8 — Source Authority Becomes More Important, Not Less

Generated answers create an understandable concern that users may pay less attention to individual source websites.

However, the trust data suggests that identifiable, credible sources remain important.

Users continue to show caution towards AI-generated information, particularly in areas such as news and higher-trust decisions.

At the same time, AI systems need reliable underlying material from which to construct answers.

This increases the strategic value of:

• Original research.
• Primary data.
• Expert authorship.
• Transparent methodology.
• Independent citations.
• Consistent brand signals.

Trend 9 — Entity Understanding Becomes a Core Search Requirement

Conversational queries increasingly ask systems to identify which candidate satisfies multiple conditions.

That requires the system to understand entities and their attributes rather than merely match pages containing similar words.

Businesses with ambiguous names, inconsistent locations, vague service descriptions or weak third-party references may become harder to evaluate confidently.

Trend 10 — “Voice SEO” Declines as a Standalone Discipline

The evidence does not suggest that voice interaction is disappearing.

The opposite is true: voice is becoming integrated into a larger number of interfaces.

What is becoming less useful is the idea that voice search requires an entirely separate SEO discipline with its own isolated optimisation techniques.

The same organisation may now be discovered through:

• A spoken query.
• A typed AI Mode query.
• An AI Overview.
• A ChatGPT conversation.
• A visual search.
• A local search.
• A conventional organic result.

These interfaces increasingly draw on overlapping information, authority and entity signals.

Research Methodology & Limitations

This report was developed by the CGO Media Research Team to provide a structured assessment of voice search, voice-assistant usage, conversational search, generative AI adoption and emerging search behaviour relevant to UK organisations in 2026.

The research deliberately distinguishes between several concepts that are frequently combined in industry commentary despite measuring different behaviours.

Research Scope

The report examines five connected areas:

• UK voice-device adoption and everyday usage.
• Smart speakers, voice assistants and audio behaviour.
• Conversational search and generative AI adoption.
• Query behaviour, local discovery and commercial intent.
• AI search interfaces, follow-up questions and future discovery.

These categories were selected because together they provide a more accurate picture of the transition from traditional voice interfaces towards wider conversational search.

Primary Research Sources

CGO Media prioritised primary research, official platform documentation and large-scale audience measurement wherever possible.

Principal sources include:

• Ofcom.
• UK Department for Science, Innovation and Technology.
• Ipsos iris.
• Edison Research.
• Google Search and Google UK.
• Google Shopping and AI Mode product documentation.

Where platform companies publish behavioural data about their own products, those statistics are attributed directly to the platform and are not presented as independent population research.

Voice Search, Voice Assistants and Smart Speakers

The terms voice search, voice assistant usage and smart-speaker usage are not treated as interchangeable within this report.

A smart speaker may be used to:

• Play music.
• Listen to radio.
• Set alarms.
• Control household devices.
• Hear weather information.
• Ask factual questions.

Only some of these activities represent information-search behaviour.

For this reason, device ownership figures are used to measure potential access to voice technology rather than the number of people conducting voice searches.

Conversational Search Definition

CGO Media uses conversational search to describe information-discovery interactions in which users can communicate in natural language and retain or refine context across a continuing search journey.

The interaction may occur through:

• Voice.
• Text.
• Images.
• Camera input.
• Multimodal AI interfaces.

Conversational search is therefore broader than voice search.

A typed ChatGPT question or an AI Mode follow-up can be conversational even when no spoken input is used.

UK-Specific and International Evidence

The report is written for a UK audience and prioritises UK-specific evidence wherever reliable research is available.

However, some important evidence about AI search interfaces comes from global or US platform data.

Examples include:

• Global AI Mode usage.
• Global AI Overview adoption.
• US multimodal AI Mode behaviour.
• Google Shopping Graph scale.

These figures are included because they document the development of search technologies already available or relevant to UK users.

They are not presented as measurements of UK population behaviour unless the underlying source is specifically UK-based.

Survey Data and Measured Audience Data

The report combines self-reported research with measured audience data.

These methodologies answer different questions.

Survey evidence can reveal whether people say they use AI, voice assistants or particular features.

Audience measurement, such as Ipsos iris, can estimate how many users actually visited measured websites and applications during a specific period.

Neither methodology should automatically be treated as superior in every context.

Self-reported behaviour can be affected by memory or interpretation, while measured audience data may not reveal exactly what the user did during each visit.

AI Chatbot Usage and AI Search Usage

A substantial methodological distinction exists between using generative AI and using generative AI specifically for search.

People may use AI tools for:

• Writing.
• Programming.
• Brainstorming.
• Summarisation.
• Translation.
• Image generation.
• Information discovery.

The first six activities do not necessarily represent search.

This report therefore avoids treating total ChatGPT or generative-AI usage as though every interaction represents a replacement for Google Search.

AI Search Summaries and Embedded AI

Conversational and generative search increasingly appears within conventional search interfaces.

This creates another measurement challenge.

A user may encounter an AI Overview without intentionally choosing an AI search product.

For this reason, the report distinguishes between:

• Standalone AI chatbot usage.
• AI Mode usage.
• AI Overview exposure.
• Traditional search activity.

These experiences can overlap within the same customer journey.

Platform-Reported Behavioural Data

Several statistics in this report originate from Google statements about AI Mode, AI Overviews and Shopping Graph behaviour.

Examples include query length, multimodal usage, planning behaviour and product-data scale.

These figures provide valuable evidence about the behaviour observed within Google’s own products.

However, they remain platform-reported statistics and should be interpreted as such.

They do not automatically demonstrate that identical behaviour occurs across ChatGPT, Perplexity, Gemini or every other conversational-search system.

Search Intent and Commercial Behaviour

Conversational search can reveal more detailed user intent because queries may contain several criteria at once.

However, a longer query does not automatically mean stronger commercial intent.

Users can ask long questions for research, learning, planning, entertainment or purchasing.

Commercial implications should therefore be assessed from the nature of the task rather than query length alone.

Voice Search and Local Search

Voice and conversational interfaces can be particularly relevant to local discovery because users can specify location and immediate requirements naturally.

However, this report avoids repeating older unsourced claims that a fixed percentage of all voice searches are local.

Instead, local-search conclusions are based on documented platform capabilities and measurable UK behaviour where available.

AI Search Visibility and Citations

AI systems do not expose all aspects of their source-selection or recommendation processes publicly.

This limits the certainty with which any organisation can claim that a single optimisation technique directly causes citation or recommendation.

CGO Media therefore distinguishes between:

• Observable source citations.
• Documented platform guidance.
• Empirical research.
• Reasoned strategic interpretation.

The report does not treat assumptions about hidden model behaviour as established ranking factors.

Correlation and Causation

Growth in AI adoption, longer queries and increased exposure to generated answers demonstrate changing search behaviour.

They do not prove that every observed change in website traffic, conversion or traditional search usage is caused by conversational AI.

Search behaviour is influenced by multiple factors, including:

• Platform changes.
• Device adoption.
• Demographics.
• Market conditions.
• Content availability.
• User habits.
• Product design.

Research Limitations

The principal limitations of this research include:

• Not every statistic is UK-specific.
• Voice-device ownership does not equal voice-search usage.
• AI chatbot usage does not necessarily equal AI-search usage.
• Survey responses are self-reported.
• Measured visits do not reveal the purpose of every interaction.
• Platform-reported statistics may use proprietary methodologies.
• AI search products are evolving rapidly.
• Product availability and functionality can differ by country.
• Search systems do not reveal every source-selection mechanism.
• New evidence may supersede individual findings after publication.

These limitations are part of responsible interpretation rather than reasons to disregard the evidence.

Frequently Asked Questions — Voice Search & Conversational Search Statistics UK 2026

The questions below address some of the most common issues surrounding voice search, conversational search, AI-generated answers and the changing way UK consumers discover information online.

1. What is voice search?

Voice search is the use of spoken language to request information or perform a search through a digital device.

The interaction may take place through a smartphone, smart speaker, vehicle, smart TV or another voice-enabled system.

Voice search should be distinguished from general voice-assistant usage because people also use voice assistants to play music, set alarms, control devices or perform tasks that do not involve information search.

2. What is conversational search?

Conversational search allows users to search using natural language and continue refining the request through follow-up questions.

Unlike traditional keyword search, the system can retain context from earlier stages of the interaction.

Conversational search can take place through voice or text and increasingly includes images, camera input and other multimodal information.

3. How common are smart speakers in the UK?

Smart-speaker ownership is now widespread.

Edison Research reported that 45% of UK people aged 16 and over owned a smart speaker in 2025, while Ofcom separately reported smart-speaker availability in 41% of UK households.

The measures use different populations, so the percentages should not be treated as directly equivalent.

4. Does smart-speaker ownership mean the same percentage of people use voice search?

No.

Smart speakers are commonly used for music, live radio, reminders, weather information and other tasks.

Ofcom found that 38% of smart-speaker owners used their device to search for answers to questions, which is more directly related to information-search behaviour.

5. Is voice search still growing?

Voice-enabled technology remains widely used, but the larger growth story is increasingly conversational and multimodal search rather than voice alone.

Spoken input is being integrated into wider AI-search interfaces where users can also type, upload images and continue with follow-up questions.

This means voice is becoming part of a broader interaction model rather than necessarily expanding as an isolated search category.

6. How common is generative AI usage in the UK?

Generative AI has reached mainstream adoption.

Ofcom reported in 2026 that 54% of UK adults use AI tools such as ChatGPT, Copilot or Gemini.

Usage is significantly higher among younger adults, including 79% of 16–24-year-olds and 74% of 25–34-year-olds.

7. Is ChatGPT replacing Google Search?

The available evidence does not support treating the relationship as a simple one-for-one replacement.

ChatGPT has built a large UK audience, but Google continues to operate at enormous scale and is integrating AI-generated answers and conversational interfaces directly into Search.

Users may therefore use conventional search, AI Overviews, AI Mode and standalone AI assistants within the same research journey.

8. What are AI Overviews?

AI Overviews are AI-generated summaries displayed within Google Search for certain queries.

They can synthesise information and provide links to supporting sources alongside conventional search results.

This means users can encounter an AI-generated answer without deliberately switching to a separate AI chatbot.

9. What is Google AI Mode?

AI Mode is Google’s conversational search experience designed for more complex questions, follow-up queries and multimodal search.

Users can ask detailed natural-language questions, continue the interaction and use different forms of input.

Google reports that AI Mode queries are substantially longer than traditional search queries, suggesting that users provide more context and constraints.

10. Why are conversational search queries longer?

Users no longer need to compress a complex requirement into a few keywords.

They can include several criteria in one request, such as:

• Location.
• Budget.
• Timing.
• Preferences.
• Features.
• Restrictions.
• Desired outcome.

The system can then use these constraints together when constructing an answer or recommendation.

11. Does conversational search matter for local businesses?

Yes.

Local conversational queries can combine several conditions such as location, opening hours, availability, services, price and customer requirements.

Search systems can increasingly use these attributes to identify businesses that fit the request.

12. Does conversational search matter for ecommerce?

Yes.

Conversational product discovery allows users to describe the product they want using natural language and refine the results according to attributes such as price, size, style, features and availability.

This increases the importance of accurate product feeds, structured attributes, pricing and inventory information.

13. Do businesses need separate voice-search pages?

Usually not.

Creating large numbers of thin pages specifically targeting spoken variants of existing keywords is unlikely to represent the strongest strategy.

A better approach is to build authoritative pages that answer genuine natural-language questions and provide enough detail to support several search interfaces.

14. Are FAQs useful for conversational search?

They can be.

FAQs naturally mirror question-based user behaviour and can provide concise answers to specific customer concerns.

However, FAQs should answer real questions and should not be created simply to repeat keywords or manufacture large numbers of near-identical queries.

15. Does schema markup guarantee AI citations?

No.

Structured data can help search systems understand entities and attributes, but there is no public evidence that adding a particular schema type guarantees selection or citation by an AI system.

Structured data should therefore support accurate machine understanding rather than be treated as an AI-ranking shortcut.

16. How can a business improve its chances of appearing in conversational search?

There is no guaranteed formula, but businesses can strengthen the information search systems have available to evaluate them.

Priority areas include:

• Clear entity information.
• Detailed product and service attributes.
• Strong topic coverage.
• Original research.
• Expert authorship.
• Structured data.
• Consistent local information.
• Independent authoritative references.

Conclusion — Voice Search & Conversational Search Statistics UK 2026

The evidence reviewed throughout this report shows that voice search remains an established part of UK digital behaviour, but it is no longer the most important way to understand the wider shift taking place in search.

The larger transformation is conversational discovery.

Users increasingly expect search systems to understand natural language, retain context, evaluate several requirements at once and help them progress from an initial question towards a recommendation, comparison or action.

Voice is becoming one input into that wider system alongside text, images, camera input and other multimodal interfaces.

Voice Interfaces Are Already Mainstream

Smart-speaker ownership and voice-assistant usage show that spoken interaction with technology is already familiar to a substantial proportion of UK consumers.

Voice interfaces are used for music, radio, weather, reminders, questions and many other everyday tasks.

This matters because it helped normalise a fundamental behavioural change:

users can speak naturally to technology and expect the system to interpret the request.

Conversational AI Has Expanded That Behaviour Beyond Voice

ChatGPT, Gemini, Copilot, AI Overviews and AI Mode have extended natural-language interaction into a much broader discovery environment.

The UK adoption evidence now shows mainstream usage of generative AI, especially among younger adults.

At the same time, AI-generated summaries are reaching users who may never deliberately choose to open a standalone chatbot.

This means conversational search is becoming part of ordinary search behaviour rather than remaining a separate technology category.

Queries Are Becoming More Expressive

Traditional search encouraged users to simplify their requirements into short phrases.

Conversational search allows the opposite.

Users can describe:

• What they want.
• Where they want it.
• What they can spend.
• Which features matter.
• Which restrictions apply.
• What outcome they are trying to achieve.

This creates richer intent and gives search systems more information with which to evaluate possible answers.

Search Is Moving Towards Recommendation and Action

The search journey increasingly extends beyond locating a webpage.

Systems can now compare options, interpret constraints, suggest suitable candidates and in some cases help users progress towards booking or purchase.

This shifts visibility from a simple ranking problem towards a wider question:

Can this organisation be understood, evaluated and confidently selected for the user’s requirement?

Entity Understanding Becomes More Important

Conversational systems need to know which business, service, product, person or location they are evaluating.

That increases the importance of clear entity information, consistent naming, structured attributes and reliable external references.

Ambiguity becomes more costly when a system is expected to provide one recommendation rather than ten blue links.

Authority and Trust Remain Critical

The Ofcom evidence shows that AI usage is increasing faster than trust in AI-generated information.

This makes source credibility more important, not less.

Organisations that publish identifiable expertise, original research, transparent methodology, reliable data and externally corroborated information can provide stronger evidence for both users and search systems.

Traditional SEO Remains Part of the System

Conversational search does not make conventional SEO irrelevant.

Search engines and AI systems still need accessible, understandable and useful information.

Technical quality, internal linking, content relevance, structured data, authority, page experience and crawlability remain important foundations.

What changes is the scope of optimisation.

The objective increasingly extends beyond ranking one page for one keyword towards ensuring the organisation can participate across multiple discovery interfaces and stages of the customer journey.

Research Usage, Citation & Press

CGO Media publishes research, statistics, frameworks and analysis to support businesses, journalists, researchers and organisations examining how search behaviour is changing across Google, AI search platforms, voice interfaces and conversational discovery.

The statistics and analysis contained within this report may be referenced in editorial coverage, research papers, presentations, industry reports and commercial analysis, provided appropriate attribution and research context are retained.

Using Individual Statistics

Where an individual statistic originates from Ofcom, Edison Research, Ipsos iris, the UK Government, Google or another external organisation, CGO Media recommends citing the original research as the primary source whenever possible.

CGO Media may be cited for the synthesis, interpretation and UK-focused analysis surrounding that evidence.

This distinction helps preserve the difference between original evidence and CGO Media’s own research conclusions.

Using CGO Media Analysis

The analytical sections of this report — including the interpretation of the 50 statistics, implications for UK businesses, the Conversational Discovery Model and 2026–2027 trends — represent CGO Media Research Team analysis.

Journalists, publishers, researchers and organisations are welcome to quote or summarise these findings with attribution to:

Press & Media Enquiries

Journalists, editors, broadcasters, researchers and industry publications requiring additional commentary or clarification can use the dedicated CGO Media Press & Media area.

Relevant enquiry topics include:

• Voice search in the UK.
• Conversational search.
• ChatGPT and AI search adoption.
• Google AI Overviews.
• Google AI Mode.
• Smart speakers and voice assistants.
• Generative Engine Optimisation.
• AI citation and source selection.
• Entity authority.
• Search behaviour and consumer discovery.

Research Methodology

CGO Media publishes a dedicated Research Methodology explaining how sources are selected, evaluated and interpreted across the wider research programme.

The methodology covers:

• Primary and secondary evidence.
• Source selection.
• Geographic limitations.
• Platform-reported data.
• AI-search evidence.
• Research limitations.
• Responsible interpretation.

View the CGO Media Research Methodology

Research Updates

Voice technology, generative AI and conversational search are developing rapidly.

Search interfaces, AI models, platform capabilities and user adoption can change substantially over relatively short periods.

CGO Media therefore reviews its search and AI research as new evidence becomes available.

Readers citing individual statistics should check the original publication date, population and geographic scope, particularly where a figure relates to a fast-moving AI platform or product feature.

Editorial & Research Attribution

This report was prepared and reviewed by the CGO Media Research Team as part of CGO Media’s wider programme examining search behaviour, AI search, GEO, technical SEO, entity authority, citation authority and digital discovery.

The report combines external evidence with CGO Media analysis to examine how voice interfaces and conversational AI are changing the way users search, compare, evaluate and act.

Sources & References

The Voice Search & Conversational Search Statistics UK 2026 report draws primarily on official UK research, independent audience measurement and first-party platform data.

CGO Media prioritised original sources wherever possible rather than relying on statistics reproduced across marketing, SEO or technology websites without clear methodology.

1. Ofcom — Audio Listening in the UK 2025

Publisher: Ofcom
Publication: Audio Listening in the UK 2025
Geography: United Kingdom
Research area: Audio behaviour, voice assistants, smart speakers and radio consumption

This report provides several of the principal UK statistics used within the voice-assistant and smart-speaker sections of this research.

Evidence drawn from the report includes smart-speaker usage, voice-assistant behaviour, audio consumption through smart speakers, source awareness, default-provider behaviour and problems with voice requests being interpreted incorrectly.

2. Edison Research / SSRS — The Infinite Dial UK 2025

Research organisation: Edison Research
Publication platform: SSRS
Publication: The Infinite Dial UK 2025
Published: May 2025
Geography: United Kingdom

The Infinite Dial UK provides evidence on smart-speaker ownership and technology adoption among people aged 16 and over.

The report documents the growth of Amazon Alexa, Google Home and Apple HomePod ownership between 2021 and 2025 and provides an important longitudinal view of smart-speaker adoption in the UK.

3. Ofcom — Adults’ Media Use and Attitudes 2026

Publisher: Ofcom
Publication: Adults’ Media Use and Attitudes 2026
Published: April 2026
Population: UK adults aged 16+
Research area: Media behaviour, AI adoption and attitudes towards AI-generated information

This is one of the most important UK-specific sources in the report.

It provides evidence on generative-AI usage among UK adults, age differences in adoption, engagement with AI-generated search summaries and attitudes towards AI-generated information.

4. Department for Science, Innovation and Technology — Public Engagement Survey 2025/2026

Publisher: UK Department for Science, Innovation and Technology
Publication: DSIT Public Engagement Survey 2025/2026
Published: 16 July 2026
Population: UK adults aged 16+
Fieldwork: November 2025 to March 2026

The nationally representative survey provides independent UK Government evidence about public engagement with digital technology and generative AI.

It is used in this report as an additional source for assessing the scale of generative-AI and natural-language assistant adoption in the UK.

5. Ofcom — The Era of Answer Engines

Publisher: Ofcom
Publication: The Era of Answer Engines: Generative AI’s Impact on Search Experiences and Online Safety
Published: November 2025
Research area: Generative-AI search, AI chatbots, AI search summaries and UK search behaviour

Ofcom’s discussion paper examines how generative AI is changing information retrieval from a system primarily designed to locate webpages towards one that can generate answers directly.

It also draws on Ipsos iris audience measurement and Ofcom qualitative research into UK adults’ search experiences.

6. Ipsos iris — UK Online Audience Measurement

Organisation: Ipsos
Dataset: Ipsos iris
Period cited: June 2024 and June 2025
Population: UK online adults

Ipsos iris audience data cited by Ofcom provides measured reach estimates for major AI chatbot services rather than relying exclusively on self-reported adoption.

The data is used to compare UK reach for services including ChatGPT, Microsoft Copilot and Google Gemini.

7. Ofcom — Online Nation 2025

Publisher: Ofcom
Publication: Online Nation 2025
Published: December 2025
Geography: United Kingdom
Research area: Online behaviour, platforms, AI services and search

Online Nation provides broader evidence about how people in the UK use online services and how AI tools are becoming integrated into digital behaviour.

The report supports analysis of the rapid expansion of ChatGPT usage in the UK and the growing presence of AI-generated search experiences.

8. Google — AI Mode Now Available in the UK

Publisher: Google
Publication: Google Search: Introducing AI Mode in the UK
Published: 28 July 2025
Geography: United Kingdom

Google’s UK launch documentation describes AI Mode’s advanced reasoning, multimodal capabilities and ability to support follow-up questions.

Google also reported that early AI Mode users were asking questions approximately two to three times longer than traditional Search queries.

9. Google — A New Era for AI Search

Publisher: Google Search
Publication: A New Era for AI Search
Published: 19 May 2026
Geography: Global platform data

Google reported that AI Mode surpassed one billion monthly users approximately one year after launch and that AI Mode queries more than doubled every quarter after launch.

The update also documents Google’s continued development of conversational, agentic and multimodal search functionality.

10. Google UK — AI Mode Restaurant Booking

Publisher: Google UK
Publication: Booking Restaurants in the UK Just Got Easier with AI in Search
Published: 10 April 2026
Geography: United Kingdom

The update documents the introduction of agentic restaurant-booking capabilities into AI Mode for UK users.

It demonstrates how conversational search can combine requirements such as location, date, time, party size, cuisine and dietary preferences before presenting suitable booking options.

Google also reported a 140% rise in UK searches for “when to book a table” during 2026 at the time of publication.

11. Google — Shopping Graph & AI Mode Shopping

Publisher: Google
Research area: Shopping Graph, AI Mode and conversational product discovery
Geography: Global platform data

Google reports that its Shopping Graph contains more than 50 billion product listings, covering information such as price, availability, reviews and product characteristics.

More than two billion of these listings are refreshed every hour.

The data illustrates the importance of current, structured product information when conversational search systems are required to evaluate products against highly specific user criteria.

12. Google — AI Overviews & AI Search Usage

Publisher: Google Search
Period: 2025–2026
Geography: Global and selected-market platform data

Google reported in 2026 that AI Overviews had surpassed 2.5 billion monthly active users globally, while AI Mode had surpassed one billion monthly users.

Google has also reported increased Search usage for query categories where AI Overviews appear.

These figures are treated in this report as first-party platform data rather than independent audience research.

13. Ofcom — News Consumption in the UK 2026

Publisher: Ofcom
Publication: News Consumption in the UK 2026
Published: 15 September 2026
Geography: United Kingdom

Ofcom’s 2026 news research examines how UK audiences discover and consume news across publishers, search engines, social platforms and AI intermediaries.

The research found that approximately one in five UK adults used an AI app or service to access news during the previous month, demonstrating the growing role of AI systems as information-discovery intermediaries.

Source Hierarchy Used by CGO Media

When reviewing evidence for this report, CGO Media applied the following source hierarchy:

Secondary statistics websites and marketing articles were not treated as primary evidence where the original source could be identified.

This is particularly important within voice-search research because a number of older statistics have been repeatedly reproduced online without adequate information about their original study population, date or methodology.

Research Review Date

Research reviewed: September 2026

CGO Media Research Team

Voice search, generative AI and conversational search remain fast-moving research areas. Platform functionality, audience reach and user behaviour should therefore be reviewed periodically as new evidence becomes available.