July 2026 SEO & AIO Industry Update

Home » News & Insights » July 2026 SEO & AIO Industry Update

Search Visibility Is Becoming Broader, Less Predictable And More Dependent On Verifiable Authority

July provided further evidence that AI is changing how people search without removing the importance of traditional search.

Google Search reportedly broke previous usage records during the World Cup, while ChatGPT continued to grow as a referral source and conversational voice products made it easier to search without typing. At the same time, Google expanded Search Console beyond websites, local-business information became more important to AI recommendations and regulators intervened in the data advantage underpinning Google’s dominance.

The clearest message is that search visibility is spreading across websites, social platforms, business profiles, reviews, videos and AI-generated answers. SEO remains the foundation, but rankings and organic website visits now reveal less of the complete customer journey.

In Short:

Google can now report how Instagram, TikTok, X and YouTube content performs in Search.
Google Business Profiles and other Google-owned surfaces are becoming prominent AI Mode sources.
ChatGPT generated 92.4% of the measured AI referral traffic in one major study.
AI citations can change according to reasoning mode and hidden retrieval systems.
Topical authority, recognisable expertise and consistent business information support AI visibility.
Incorrect crawler controls can unintentionally block Googlebot or prevent noindex directives from working.
Regulators and publishers are challenging the exchange of content and traffic underpinning AI search.

Search Visibility Now Extends Beyond Your Website

Business Impact

Customers may encounter a business through a social post, video, Google Business Profile, Yelp listing or AI-generated recommendation before visiting its website.

These surfaces should no longer be treated as separate from SEO. Together, they provide the information search engines and AI assistants use to understand, verify and recommend a business.

Social, Local And Google-Owned Surfaces Are Becoming SEO Assets

What happened

Google introduced Search Console Platform Properties, allowing eligible Instagram, TikTok, X and YouTube accounts to report impressions, clicks, queries and content performance within Google Search. Separate analysis found that citations to Google-owned properties in AI Mode had increased 8.4 times in approximately two months, making Google’s own domain the second-most-cited group in the study, particularly through business and product panels. ChatGPT also gained access to real-time Yelp reviews, ratings, photographs and business information.

Why this matters

The information held about a business outside its website is becoming increasingly important to both traditional and AI-led discovery.

For local businesses, this includes:

Google Business Profile categories and services;
Opening hours and contact information;
Customer photographs;
Google and Yelp reviews;
Responses to customer feedback;
Consistent names, addresses and telephone numbers;
Social and video content;
References from recognised directories and publications.

An audit of 71 real businesses found substantial gaps between what those businesses were in the real world and what AI systems could confidently verify online. Problems included inaccessible websites, inconsistent details, dead domains and insufficient independent trust signals.

This means a business can have a legitimate operation and reasonable Google rankings yet remain difficult for an AI assistant to verify or recommend.

Google Business Profile remains a priority, but it is no longer the only local source that matters. Yelp’s integration with ChatGPT shows why review management and consistent business information need to extend across the wider local-search ecosystem.

Search Console Platform Properties should also make social and video activity easier to incorporate into SEO reporting. We can set this up for eligible clients and will contact relevant clients about implementation by the end of August. Clients wishing to arrange access sooner can contact us directly.

Search Demand Is Moving, While AI Referral Traffic Remains Concentrated

Business Impact

Declining traffic for an established keyword does not necessarily mean that the underlying customer demand has disappeared.

People may be dividing one broad search into several more specific questions, moving between search and AI platforms, or completing more of their research before visiting a website.

Search Behaviour Is Being Redistributed Across More Queries And Interfaces

What happened

A study of more than one million high-volume keywords found that 29% of measured demand was declining, but this was largely offset by growth across other queries. Separately, analysis of 6.77 million trackable AI-referred sessions found that ChatGPT generated 92.4% of measured AI referral traffic. OpenAI also introduced GPT-Live, making spoken conversations, interruptions and follow-up searches more natural within ChatGPT.

Why this matters

The keyword study does not prove that all high-volume demand is moving directly into low-volume long-tail searches. It does show that search behaviour is becoming less concentrated around a stable list of familiar phrases.

A broad query may be replaced by:

A specific comparison;
A location or price-led question;
Several follow-up searches;
A YouTube or social search;
A conversational AI prompt;
A voice-led research session.

SEO research therefore needs to look beyond whether an individual keyword has gained or lost volume. Businesses should examine which questions, formats and stages of intent are growing around the wider topic. The referral study also reinforces the need to report ChatGPT separately rather than combining every assistant into one AI channel. Other platforms remain worth monitoring, but ChatGPT currently accounts for the overwhelming majority of measurable AI referral sessions in the cited dataset.

One notable finding was that 28.8% of ChatGPT referrals in the study landed on internal website search-result pages. This may indicate that the assistant identified the correct website but could not confidently identify the best product, service or information page.

Businesses receiving this traffic should review whether their:

Site architecture is clear;
Landing pages address distinct needs;
Important products and services have direct URLs;
Internal links communicate page relationships;
Titles and headings distinguish similar pages;
Key information is available without additional searching.

AI is creating another acquisition channel, but it also exposes weaknesses in website structure that conventional reporting may overlook.

AI Visibility Is Unstable, But Authority Still Compounds

Business Impact

AI visibility, even if being cited by AI introduces another layer of competition. AI visibility should not be treated as a fixed ranking.

The same prompt can generate different brands and sources depending on the model, reasoning mode, retrieval route and time of testing. Businesses need to measure how frequently they appear rather than claiming one permanent ChatGPT or AI Mode position.

Different AI Systems Can Produce Different Sources For The Same Question

What happened

A study of ChatGPT reasoning modes found only 25.6% overlap between the sources cited under minimal and high reasoning. Higher reasoning increased citation rates from 50% to 68% and performed nearly five times as many web searches. Other testing found that 11.6% of repeated prompts changed their underlying retrieval source, substantially reducing the overlap between cited pages and domains. Separate research estimated that around 40% of AI citations leave the source brand unnamed.

Why this matters

A company may appear in a quick answer but disappear when ChatGPT performs more detailed research. It may also provide the information supporting an answer without being named clearly enough for the user to recognise the brand.

AI visibility reporting should therefore distinguish between

Being retrieved;
Being cited;
Receiving a visible link;
Being named;
Being recommended;
Generating a website visit;
Contributing to a conversion.

Repeated testing is more useful than isolated screenshots. Reports should record the platform, model, reasoning mode, prompt and date wherever possible.

Traditional Authority Signals Continue To Matter

What happened

Analysis reported by Search Engine Land found that businesses with stronger topical authority were better able to earn and retain visibility in AI-generated answers. At the same time, Google commentary indicated stronger scrutiny of scaled AI pages that are thin, repetitive or insufficiently useful.

Why this matters

This reinforces established SEO principles rather than introducing a completely separate AI optimisation system.

A business is more likely to be understood as a reliable source when it provides:

Clear core product and service information;
Useful supporting content;
Experienced authors and contributors;
Original evidence or examples;
Logical internal linking;
Consistent terminology;
Reputable third-party references;
A coherent body of information around its area of expertise.

These qualities broadly reflect the principles commonly discussed through experience, expertise, authority and trust. Topical authority does not mean publishing hundreds of loosely differentiated pages. In fact, scaling generic AI content can weaken the very authority a business is attempting to build.

The question should not be whether AI was used to produce a page. The question is whether the finished page contributes something distinctive, accurate and genuinely useful.

Technical Access Remains The Foundation Of Visibility

Business Impact

A business cannot rank or be cited if search and AI systems cannot access and interpret its content.

July showed how controls introduced for security, privacy or AI governance can unintentionally damage conventional SEO.

Crawler Controls Can Quietly Remove Content From Search

What happened

Cloudflare introduced more granular controls separating AI crawler activity into search, agent and training categories, but reporting warned that some configurations could also block Googlebot. Publicly shared Claude conversations also demonstrated why disallow is not the same as noindex: Google discovered the URLs but could not read their noindex instructions because robots.txt prevented crawling. Google additionally advised that incorrect XML sitemap lastmod dates are worse than omitting the dates entirely.

Why this matters

Crawler access should not be changed through a general AI-blocking switch without first understanding which systems will be affected.

Businesses using Cloudflare or another content-delivery and security platform should test crawler changes through:

Server and CDN logs;
Search Console URL Inspection;
Response-code testing;
Robots.txt validation;
Verified crawler information;
Comparisons between crawler and normal-user responses.

The Claude example also reinforces a longstanding technical SEO principle: robots.txt controls crawling, not guaranteed index removal.

A page that must remain private should be protected by authentication or another genuine access control. A page that should remain publicly accessible but excluded from Google should use a noindex instruction that Google is permitted to crawl and read.

For XML sitemaps, lastmod should only be populated when it reflects a meaningful page change. Automatically updating every URL to the current date provides no useful freshness signal and may cause Google to stop trusting the field

The Exchange Between Search Platforms And The Open Web Is Being Renegotiated

Business Impact

Search and AI systems depend on websites for current, trustworthy information. Publishers depend on those systems for discovery, audiences and revenue.

That relationship is becoming increasingly contested, with potentially significant consequences for which sources remain available to search engines and AI assistants

Regulators And Publishers Are Challenging Google’s Position

What happened

The European Commission issued binding measures requiring Google to share anonymised search information with eligible competing search engines and AI chatbots, including the types of data Google uses to optimise its own services. Google simultaneously said AI Mode and AI Overviews send publishers billions of clicks each week. However, publishers including Reddit, USA Today, Politico, The Economist, Reuters and People Inc. were reported to be reassessing their relationships with Google as AI-generated answers reduced referral traffic.

Why this matters

Access to Google’s search data could help competing engines understand queries, ranking outcomes, clicks and user satisfaction at a scale previously unavailable to them.

This is unlikely to transform the market immediately, but it could help alternative search and answer engines improve their products. SEO may gradually become less synonymous with optimising exclusively for Google.

The publisher dispute highlights the other side of the issue.

Google can send billions of clicks across the web while individual publishers still experience material traffic losses. Aggregate growth does not show how those clicks are distributed, whether they replace earlier traffic or which sectors benefit.

Reuters and Politico were among the publishers reported to be considering tighter access controls, while Reddit was reassessing its content agreement with Google. This does not mean ordinary businesses should block Google. For most organisations, the potential loss of conventional search visibility would outweigh any immediate benefit.

It does mean crawler access is becoming a commercial and strategic decision rather than a purely technical one.

Businesses should understand:

Which crawlers access their content;
Whether access supports search, live answers or model training;
How much traffic each platform returns;
Where their brand is cited without receiving a click;
Which independent sources corroborate their claims;
What would happen if an important publisher or data platform restricted access.

As more established publishers protect their content, AI systems may become more dependent on licensed partners, Google-owned properties and sources that remain technically open. That could make third-party authority and accurate structured business information even more important.

AI Is Also Changing How SEO Work Is Delivered

Anthropic added the ability for Claude Cowork to learn repeatable processes from recorded workflows. A user can demonstrate a task and allow Claude to reproduce the sequence later.

For SEO teams, this could support recurring work such as:

Assembling reports;
Collecting competitor changes;
Organising content inventories;
Conducting standard technical checks;
Preparing research;
Comparing templates or page groups.

The opportunity is greater consistency and less repetitive manual work. The risk is reproducing a flawed process more efficiently.

Automation should support expert decision-making, not remove the review needed to determine whether the output is accurate, relevant and appropriate for a particular client.

What This All Means For Brands

July’s developments point towards several practical priorities

Connect Social And Video Accounts To Search Console
- Eligible Instagram, TikTok, X and YouTube accounts can now provide first-party evidence of how off-site content performs in Google Search
Treat Local Profiles And Reviews As AI Discovery Assets
- Google Business Profile remains essential, while Yelp’s integration with ChatGPT demonstrates the importance of maintaining accurate information and authentic reviews across multiple platforms.
Measure Search Demand Across Topics, Not Only Keywords
- Identify where demand is declining, where related questions are growing and whether customers are moving into video, social, local or conversational search.
Report AI Visibility As A Frequency
- Use repeated testing and distinguish between citations, brand mentions, recommendations and actual referral visits.
Build Genuine Topical And Brand Authority
- Prioritise useful coverage, recognisable expertise, original evidence and third-party corroboration rather than scaling generic AI-generated pages
Audit Crawler Controls
- Review Cloudflare, robots.txt, noindex, authentication and sitemap configuration before changing how AI systems access a website.
Broaden Search Reporting
- Website visits remain important, but they should be considered alongside social-search impressions, local actions, branded searches, AI referrals and assisted conversions.

July’s Key Learnings

July did not show traditional search being replaced. Google reached record usage during a major global event while AI assistants continued to expand how people research, compare and discover businesses.

What is changing is the number of surfaces involved.

A customer may move from ChatGPT to Yelp, from Google AI Mode to a Business Profile, or from a Google search to a social video without following a conventional website-first journey.

Strong SEO still provides the foundation for these experiences. Technical accessibility, useful content, clear business information and recognised expertise help both search engines and AI systems understand whether a business is relevant and trustworthy.

The difference is that those signals must now remain consistent across a much wider discovery ecosystem.

Success is no longer defined solely by holding a ranking or generating a website click. It increasingly depends on whether a business can be found, understood, verified and remembered wherever its customers choose to search.