📊 Google Q2 Earnings & AI Usage Surge

Alphabet’s Q2 2025 earnings report revealed a 14% year-over-year revenue increase, hitting $96.4 billion. Google Search alone brought in $54 billion, underscoring its enduring dominance. A key highlight was the aggressive expansion of AI services across Google platforms. AI Overviews now reach over 2 billion monthly users globally. Google’s AI Mode is actively used by more than 100 million users in the U.S. and India, and the Gemini mobile app has surpassed 450 million monthly active users.

Capital expenditure has been increased dramatically. Alphabet will now spend up to $85 billion in 2025, with a significant portion going to AI infrastructure, including data centre expansion and TPU upgrades.

🧠 Google Launches Web Guide: AI-Powered Topic Clustering

In mid-July, Google introduced a new Search Labs experiment called Web Guide, which uses Gemini to cluster results into thematic groups. Instead of presenting a standard vertical list of links, it organises results by intent or topic (e.g. “beginner tips,” “advanced strategies,” “video tutorials”). This could mark a major evolution in how search results are navigated – and how SEOs structure their content.

Early feedback has been mixed. While some users praise the clarity, others argue it adds cognitive load. For SEOs, it reinforces the value of well-structured, topic-cluster content strategies.

🧪 Project Mariner Enters Agent Mode Testing

DeepMind’s Project Mariner is expanding with a new Agent Mode integration. This advanced AI assistant can now:

• Analyse and interpret screen layouts
• Navigate through websites
• Fill out forms and simulate user behaviour
• Conduct basic research tasks autonomously

Developers can now access Agent Mode via the Gemini API. Google is expected to merge this functionality into AI Mode within Search – paving the way for real-time task execution in the browser and further automation of search-related actions.

📉 AI Overviews Decimate Publisher Traffic

New research from Authoritas and Pew confirms what publishers feared: AI Overviews drastically reduce organic traffic. Key findings include:

• Click-through rates on featured sources under AI Overviews average just 0.94%
• Some sites report traffic drops of up to 79% for queries now answered via AI summaries
• Users increasingly rely on AI answers without clicking links, undermining traditional organic visibility

In response, UK publishers and digital watchdogs have called for regulatory reviews of Google’s dominance and transparency in AI citation practices.

🔍 AI Summaries Dominate Longer Queries

According to Pew, AI Overviews appear in about 18% of total Google queries. However, for longer or more complex queries (10+ words), this rises to over 52%. Single-word searches rarely trigger AI responses, suggesting that informational intent still guides summary activation.

For SEOs, this data underscores the importance of optimising for conversational, question-based queries – and ensuring content appears in AI training sets and structured results.

📱 Google Search UI Tests Continue

Several users and developers have reported UI changes in Search, including:

• AI Mode returning only link cards (possibly a bug)
• Testing of expanded “Helpful Highlights” for product queries
• New scrollable “Knowledge Snapshots” for definitions and explanations

Google also updated merchant policy documentation and revised local service ad guidance inside Business Profiles – signalling continued integration between Search and Maps-based commercial queries.

🤝 Google Cloud to Support OpenAI Infrastructure

In a surprising collaboration, Google Cloud has partnered with OpenAI to provide compute resources. This follows a broader industry trend where competitors collaborate on foundational infrastructure despite product rivalry.

Google Cloud, AWS, and Microsoft Azure are all aggressively investing in hosting AI workloads – and search providers are becoming compute providers in a race to dominate AI development ecosystems.

🌪️ June Core Update Fallout Lingers

Although the June 2025 Core Update completed rolling out in mid-July, volatility continues to impact many sites. SEO forums report:

• Delayed recoveries for “helpful content” pages
• Increased weight on E-E-A-T signals
• Penalisation of sites with thin or unoriginal content
• Sites with strong internal linking and structured content faring better overall

Google has not released any new documentation, but many SEOs believe the core update and Helpful Content System are now fully merged under the “Page Quality System.”

💼 Job Posting Rich Results See Manual Action Wave

Google Search Central confirmed a spike in manual actions affecting sites using JobPosting schema. This appears tied to policy enforcement on:

• Expired or misleading job ads
• Duplicate job listings across domains
• Abuse of non-structured content elements within job markup

If you run recruitment content, double-check your structured data and monitor GSC for manual action notifications.

🔎 Google Lens Updates Visual Search Priorities

Google Lens is now being used in over 12 billion monthly visual queries. A July update improves object segmentation and integrates Gemini-generated labels, making it easier to identify specific components in clothing, décor, or food images.

Implication for SEOs: Image SEO (including filenames, structured data, and alt text) is more important than ever. Visual search is rapidly converging with traditional and AI search pathways.

💬 What the Experts Are Saying

“AI Overviews are increasing query volume by over 10 percent and reshaping user engagement.”
— Sundar Pichai (Alphabet CEO)

“Web Guide’s clustering system could change how we structure content – context is the new king.”
— Aleyda Solís, International SEO Consultant

“Don’t assume that AI-generated traffic equals real interest. Track engagement deeper than ever.”
— Lily Ray, SEO Director at Amsive Digital

“SEO in 2025 isn’t just about ranking pages. It’s about inserting your value into every AI-generated conversation.”
— David Roche, The SEO Guide Book

❓ FAQs

What is Web Guide in Google Search?

Web Guide is a Search Labs experiment that uses Gemini to organise search results by topic clusters rather than a flat list.

Why are publishers losing traffic with AI Overviews?

AI Overviews often answer questions directly in the search results, meaning users no longer need to click through to websites.

How can SEOs respond to the June Core Update?

Focus on improving E-E-A-T signals, remove thin content, and enhance internal linking. Use structured data to help Google interpret your site accurately.

Are AI summaries more likely on certain queries?

Yes. Longer, question-based queries are far more likely to trigger an AI Overview than short or branded keywords.