News · July 10, 2026
Predicting search used to mean guessing the next Google update. In 2026 the question is bigger: what happens when the dominant interface for finding anything stops being a list of links and becomes an answer, and then becomes an agent acting on your behalf? The data now gives a reasonably honest picture. Here is what UK businesses should expect from search this year and next, and what to do about each part of it.
The structural trend is settled. Zero-click searches on Google rose from 56% to 69% in the year following AI Overviews’ launch, and Gartner projects traditional search volume declining 25% through 2026, with organic traffic falling as much as 50% by 2028 for many categories, figures compiled at Mersel AI and Digital Agency Network. Informational queries are absorbed first: the quick definitions, comparisons and how-tos that once fed blog traffic are increasingly answered on the results page itself. Any UK business whose growth model assumes ever-rising organic sessions needs a new model, this year.
The traffic that AI surfaces do send behaves differently. Independent measurements repeatedly find AI-referred visitors converting at several times organic rates: Seer Interactive logged ChatGPT visitors converting at 15.9% against 1.76% for organic, and Ahrefs found AI visitors producing 12.1% of signups from just 0.5% of traffic. SimilarWeb counted over a billion AI referral visits in a single month of 2025, growing triple digits year on year. The strategic read: stop judging AI search by session counts. It is a shortlist channel; by the time someone clicks, the machine has already argued your case. Being on the shortlist is the new page one, which is the entire premise of Generative Engine Optimisation.
Google is not being replaced so much as replacing itself. AI Overviews reach a claimed near-billion searchers, AI Mode extends the conversational surface, and Google’s May 2026 Search Central guidance confirmed the operating rule: its AI surfaces lean heavily on underlying organic ranking. On Google, classic technical and authority SEO is the floor that AI visibility stands on, which is why writing off traditional SEO is exactly wrong. The nuance is behavioural: even where blue links persist, AI Overviews sharply cut click-through on them, so ranking well now often pays in citation rather than click.
ChatGPT drives 87.4% of AI referral traffic per Conductor’s 2026 benchmarks, retrieves through Bing’s index, and cites commercial pages far more readily than rivals. Perplexity is citation-forward and freshness-obsessed. The overlap between top Google results and AI-cited sources has fallen from 70% to under 20% by Brandlight’s measurement, meaning each surface must be earned somewhat separately. Practical consequence: Bing Webmaster Tools is suddenly non-optional, product and pricing pages deserve optimisation attention, and a monthly citation check across engines belongs in your reporting stack. Our ChatGPT optimisation guide covers the tactics.
The frontier shift is agentic search: shopping assistants, research agents and booking tools that query the web on a person’s behalf and return conclusions, not links. Early consumer data points the same way, with 58% of users telling Capgemini they have replaced classic search with AI tools for product and service discovery, and a majority expressing willingness to buy what an AI suggests. Agents read structure, schema, clean APIs and conventions like llms.txt; they are indifferent to hero animations. By 2027 a meaningful share of searches will not be typed by a human at all, and the brands that are machine-legible will be the ones agents can choose.
The uncomfortable thread through all of this is attribution. Zero-click influence produces buyers who arrive as direct traffic carrying a recommendation your analytics never saw. The 2026 toolkit is pragmatic: AI-referral segments in GA4, monthly brand-citation tracking across the major engines, branded search volume as a proxy for machine-made awareness, and the humble “how did you hear about us” field, which routinely reveals that a fifth or more of inbound leads were AI-influenced. Boards will need educating that share of answer is a leading indicator the same way rankings once were.
Put together, the honest strategy is a portfolio. Defend organic where it still pays, especially commercial-intent queries where clicks persist. Invest in GEO so the answer engines cite and recommend you, which mostly means comparison-rich, statistically grounded, freshly maintained content on a technically clean site. Convert ruthlessly, because scarcer clicks raise the price of every leak, making CRO the multiplier on the whole system. And keep building owned channels, especially email, that no interface shift can confiscate. None of these is optional; the weighting is what differs by business.
Three moves. First, benchmark reality: measure your current AI citation share on your twenty most valuable queries and your AI-referral conversion rate, so 2026 planning starts from data rather than anxiety. Second, ship the technical floor: schema, server-side rendering, Bing indexation, crawler access. Third, commit to one content cluster built to be cited, with original numbers in it. Businesses that did the equivalent in early SEO eras compounded for a decade; the same window is open now, and most UK competitors have not stepped through it.
It helps to see the money moving beneath the interface change. Generative AI and agents influenced an estimated 262 billion dollars of global retail revenue in the 2025 holiday season, roughly a fifth of the total, on Salesforce’s measurement, while AI-referred sessions carried higher revenue per visit than organic in Adobe’s data. On the cost side, click scarcity is inflationary: as informational clicks evaporate, competition concentrates on the commercial queries that still pay, pushing both paid auctions and SEO difficulty upward. The rational response is portfolio arithmetic: move investment toward surfaces where value is created before the click, citation share, brand presence in answers, owned channels, and hold organic and paid positions where intent remains monetisable. Businesses running 2019 channel weightings into 2026 economics are the ones experiencing the shift as decline rather than reallocation.
Prediction humility matters, so plan against ranges. The base case: current trajectories continue, zero-click keeps climbing, AI referrals keep growing triple digits from a small base, and Google’s surfaces remain the volume centre. The acceleration case: agentic adoption spikes, and by 2027 a material share of category discovery happens machine-to-machine, rewarding brands with clean structured data, APIs and llms.txt conventions overnight. The friction case: regulation, licensing disputes or answer-quality scandals slow the rollout, extending the life of classic SERPs. The strategic insight is that the sensible preparation is nearly identical across all three: verifiable content, technical retrievability, earned authority, conversion strength and owned audiences pay in every scenario, which is what makes them strategy rather than speculation.
Five signals will tell you which scenario is unfolding. Monetisation of AI surfaces: any move to paid placement inside answers changes GEO economics instantly. Attribution tooling: platforms exposing citation impression data would end the measurement fog. Agent protocols: adoption of standards that let assistants transact directly with businesses. Publisher settlements: licensing deals that shape what content engines can synthesise. And the query mix: whether commercial queries follow informational ones into zero-click, which would raise the stakes on answer presence again. We track these across client reporting; sudden movement in any of them is a rebalancing trigger, not a curiosity.
No, and the claim misreads the mechanics: Google’s AI surfaces lean on organic ranking, and authority signals feed every engine’s trust decisions. What is dying is SEO judged purely on click volume. The discipline survives as the foundation layer of answer-era visibility, with citation share joining rankings as a headline metric.
Usually some, rarely all. Paid retains urgency, precision and guaranteed presence; GEO earns the recommendation layer money cannot currently buy. The audit question is marginal return: saturated paid keywords with flat incremental yield are the natural funding source for citation assets and authority work.
Reframe the KPI set before the trend forces it: present sessions alongside citation share, AI-referral conversion, branded search volume and assisted enquiries. A business can lose informational clicks while gaining answer presence and revenue; boards handle that story well when the metrics arrive together rather than as excuses.
Verifiability. Content a machine can safely repeat, structured so it can be retrieved, on a site it can crawl, from a brand independent sources vouch for. Every surface shift of the past two years has rewarded exactly that, and nothing on the 2026 horizon points anywhere else.
Strategy survives contact with reality only if the team is built for it, so close the loop on capability. The 2026 search function, in-house or agency-side, needs four muscles working together: classic technical and authority SEO, because it remains the floor under Google’s AI surfaces; GEO craft, the answer-first, evidence-dense content discipline the engines cite; conversion competence, because scarcer clicks make every landing experience a revenue decision; and measurement literacy across the new metrics, citation share, AI-referral conversion, branded demand. Most UK teams have the first muscle and are missing at least two of the others, which is a hiring and training agenda more than a crisis. The transition also rewards a temperament: curiosity over certainty, quarterly experiments over annual plans, and a willingness to report new metrics to sceptical stakeholders before the old ones finish declining. The businesses that navigated previous platform shifts well were rarely the biggest; they were the ones that started rehearsing early.
Search in 2026 is a portfolio: shrinking but still valuable blue links, fast-growing AI answers whose clicks convert at multiples of organic, and the first commercial stirrings of agentic discovery. The winning posture is the same in every scenario: be verifiable, be retrievable, be independently vouched for, convert what arrives, and own an audience no interface shift can confiscate. Measure presence in answers alongside positions in lists, rebalance budget toward where value is created before the click, and treat this year’s early-mover window as the strategic gift it is, because it will not be open in 2027.
Prioritise urgency by exposure. E-commerce and consumer brands face the steepest curve, because AI-assisted shopping is already converting at scale and holiday-season data shows the channel compounding fastest there. Professional and B2B services follow closely: buyers now shortlist via assistants long before a form fill, so absence from answers is silent pipeline loss. Local and regional businesses have the happiest news, since entity clarity and reviews translate quickly into AI recommendations with modest effort. Publishers and information businesses face the hardest strategic questions and should be furthest along in diversifying revenue away from raw traffic. Wherever you sit, the sequencing in this piece holds; only the deadline differs.
Search is not dying. It is being renegotiated, and the terms favour brands that are verifiable, structured and genuinely useful. If you want a clear map of where your visibility stands across every surface that matters in 2026, speak with a specialist today.
Smarter Marketing. Real Results.
News · July 7, 2026
Predicting search used to mean guessing the next Google update. In 2026 the question is bigger: what happens when the dominant interface for finding anything stops being a list of links and becomes an answer, and then becomes an agent acting on your behalf? The data now gives a reasonably honest picture. Here is what UK businesses should expect from search this year and next, and what to do about each part of it.
The structural trend is settled. Zero-click searches on Google rose from 56% to 69% in the year following AI Overviews’ launch, and Gartner projects traditional search volume declining 25% through 2026, with organic traffic falling as much as 50% by 2028 for many categories, figures compiled at Mersel AI and Digital Agency Network. Informational queries are absorbed first: the quick definitions, comparisons and how-tos that once fed blog traffic are increasingly answered on the results page itself. Any UK business whose growth model assumes ever-rising organic sessions needs a new model, this year.
The traffic that AI surfaces do send behaves differently. Independent measurements repeatedly find AI-referred visitors converting at several times organic rates: Seer Interactive logged ChatGPT visitors converting at 15.9% against 1.76% for organic, and Ahrefs found AI visitors producing 12.1% of signups from just 0.5% of traffic. SimilarWeb counted over a billion AI referral visits in a single month of 2025, growing triple digits year on year. The strategic read: stop judging AI search by session counts. It is a shortlist channel; by the time someone clicks, the machine has already argued your case. Being on the shortlist is the new page one, which is the entire premise of Generative Engine Optimisation.
Google is not being replaced so much as replacing itself. AI Overviews reach a claimed near-billion searchers, AI Mode extends the conversational surface, and Google’s May 2026 Search Central guidance confirmed the operating rule: its AI surfaces lean heavily on underlying organic ranking. On Google, classic technical and authority SEO is the floor that AI visibility stands on, which is why writing off traditional SEO is exactly wrong. The nuance is behavioural: even where blue links persist, AI Overviews sharply cut click-through on them, so ranking well now often pays in citation rather than click.
ChatGPT drives 87.4% of AI referral traffic per Conductor’s 2026 benchmarks, retrieves through Bing’s index, and cites commercial pages far more readily than rivals. Perplexity is citation-forward and freshness-obsessed. The overlap between top Google results and AI-cited sources has fallen from 70% to under 20% by Brandlight’s measurement, meaning each surface must be earned somewhat separately. Practical consequence: Bing Webmaster Tools is suddenly non-optional, product and pricing pages deserve optimisation attention, and a monthly citation check across engines belongs in your reporting stack. Our ChatGPT optimisation guide covers the tactics.
The frontier shift is agentic search: shopping assistants, research agents and booking tools that query the web on a person’s behalf and return conclusions, not links. Early consumer data points the same way, with 58% of users telling Capgemini they have replaced classic search with AI tools for product and service discovery, and a majority expressing willingness to buy what an AI suggests. Agents read structure, schema, clean APIs and conventions like llms.txt; they are indifferent to hero animations. By 2027 a meaningful share of searches will not be typed by a human at all, and the brands that are machine-legible will be the ones agents can choose.
The uncomfortable thread through all of this is attribution. Zero-click influence produces buyers who arrive as direct traffic carrying a recommendation your analytics never saw. The 2026 toolkit is pragmatic: AI-referral segments in GA4, monthly brand-citation tracking across the major engines, branded search volume as a proxy for machine-made awareness, and the humble “how did you hear about us” field, which routinely reveals that a fifth or more of inbound leads were AI-influenced. Boards will need educating that share of answer is a leading indicator the same way rankings once were.
Put together, the honest strategy is a portfolio. Defend organic where it still pays, especially commercial-intent queries where clicks persist. Invest in GEO so the answer engines cite and recommend you, which mostly means comparison-rich, statistically grounded, freshly maintained content on a technically clean site. Convert ruthlessly, because scarcer clicks raise the price of every leak, making CRO the multiplier on the whole system. And keep building owned channels, especially email, that no interface shift can confiscate. None of these is optional; the weighting is what differs by business.
Three moves. First, benchmark reality: measure your current AI citation share on your twenty most valuable queries and your AI-referral conversion rate, so 2026 planning starts from data rather than anxiety. Second, ship the technical floor: schema, server-side rendering, Bing indexation, crawler access. Third, commit to one content cluster built to be cited, with original numbers in it. Businesses that did the equivalent in early SEO eras compounded for a decade; the same window is open now, and most UK competitors have not stepped through it.
It helps to see the money moving beneath the interface change. Generative AI and agents influenced an estimated 262 billion dollars of global retail revenue in the 2025 holiday season, roughly a fifth of the total, on Salesforce’s measurement, while AI-referred sessions carried higher revenue per visit than organic in Adobe’s data. On the cost side, click scarcity is inflationary: as informational clicks evaporate, competition concentrates on the commercial queries that still pay, pushing both paid auctions and SEO difficulty upward. The rational response is portfolio arithmetic: move investment toward surfaces where value is created before the click, citation share, brand presence in answers, owned channels, and hold organic and paid positions where intent remains monetisable. Businesses running 2019 channel weightings into 2026 economics are the ones experiencing the shift as decline rather than reallocation.
Prediction humility matters, so plan against ranges. The base case: current trajectories continue, zero-click keeps climbing, AI referrals keep growing triple digits from a small base, and Google’s surfaces remain the volume centre. The acceleration case: agentic adoption spikes, and by 2027 a material share of category discovery happens machine-to-machine, rewarding brands with clean structured data, APIs and llms.txt conventions overnight. The friction case: regulation, licensing disputes or answer-quality scandals slow the rollout, extending the life of classic SERPs. The strategic insight is that the sensible preparation is nearly identical across all three: verifiable content, technical retrievability, earned authority, conversion strength and owned audiences pay in every scenario, which is what makes them strategy rather than speculation.
Five signals will tell you which scenario is unfolding. Monetisation of AI surfaces: any move to paid placement inside answers changes GEO economics instantly. Attribution tooling: platforms exposing citation impression data would end the measurement fog. Agent protocols: adoption of standards that let assistants transact directly with businesses. Publisher settlements: licensing deals that shape what content engines can synthesise. And the query mix: whether commercial queries follow informational ones into zero-click, which would raise the stakes on answer presence again. We track these across client reporting; sudden movement in any of them is a rebalancing trigger, not a curiosity.
No, and the claim misreads the mechanics: Google’s AI surfaces lean on organic ranking, and authority signals feed every engine’s trust decisions. What is dying is SEO judged purely on click volume. The discipline survives as the foundation layer of answer-era visibility, with citation share joining rankings as a headline metric.
Usually some, rarely all. Paid retains urgency, precision and guaranteed presence; GEO earns the recommendation layer money cannot currently buy. The audit question is marginal return: saturated paid keywords with flat incremental yield are the natural funding source for citation assets and authority work.
Reframe the KPI set before the trend forces it: present sessions alongside citation share, AI-referral conversion, branded search volume and assisted enquiries. A business can lose informational clicks while gaining answer presence and revenue; boards handle that story well when the metrics arrive together rather than as excuses.
Verifiability. Content a machine can safely repeat, structured so it can be retrieved, on a site it can crawl, from a brand independent sources vouch for. Every surface shift of the past two years has rewarded exactly that, and nothing on the 2026 horizon points anywhere else.
Strategy survives contact with reality only if the team is built for it, so close the loop on capability. The 2026 search function, in-house or agency-side, needs four muscles working together: classic technical and authority SEO, because it remains the floor under Google’s AI surfaces; GEO craft, the answer-first, evidence-dense content discipline the engines cite; conversion competence, because scarcer clicks make every landing experience a revenue decision; and measurement literacy across the new metrics, citation share, AI-referral conversion, branded demand. Most UK teams have the first muscle and are missing at least two of the others, which is a hiring and training agenda more than a crisis. The transition also rewards a temperament: curiosity over certainty, quarterly experiments over annual plans, and a willingness to report new metrics to sceptical stakeholders before the old ones finish declining. The businesses that navigated previous platform shifts well were rarely the biggest; they were the ones that started rehearsing early.
Search in 2026 is a portfolio: shrinking but still valuable blue links, fast-growing AI answers whose clicks convert at multiples of organic, and the first commercial stirrings of agentic discovery. The winning posture is the same in every scenario: be verifiable, be retrievable, be independently vouched for, convert what arrives, and own an audience no interface shift can confiscate. Measure presence in answers alongside positions in lists, rebalance budget toward where value is created before the click, and treat this year’s early-mover window as the strategic gift it is, because it will not be open in 2027.
Prioritise urgency by exposure. E-commerce and consumer brands face the steepest curve, because AI-assisted shopping is already converting at scale and holiday-season data shows the channel compounding fastest there. Professional and B2B services follow closely: buyers now shortlist via assistants long before a form fill, so absence from answers is silent pipeline loss. Local and regional businesses have the happiest news, since entity clarity and reviews translate quickly into AI recommendations with modest effort. Publishers and information businesses face the hardest strategic questions and should be furthest along in diversifying revenue away from raw traffic. Wherever you sit, the sequencing in this piece holds; only the deadline differs.
Search is not dying. It is being renegotiated, and the terms favour brands that are verifiable, structured and genuinely useful. If you want a clear map of where your visibility stands across every surface that matters in 2026, speak with a specialist today.
Smarter Marketing. Real Results.
Everything we write about, from AI Overviews to advanced SEO, is what we put to work for our clients every day. Let's talk about what it could do for your business.