SEO news: January 2026
Martin Hill
AI content visibility, personalized AI Mode, ChatGPT ads, or usage statistics of traditional search vs. AI tools. We’ve selected the best of January for you!
Content visibility for AI bots
Do you have content on your website? And could I see it?
A seemingly trivial query can trigger AI bots to hit your server many times more frequently than conventional crawlers. While your site’s content might appear fully visible to a human user, that isn't always the case for automated bots. Before analyzing why your content isn't being cited, ensure there are no technical barriers preventing these bots from accessing your site in the first place.
If you want to be in the ChatGPT or Gemini etc. responses, your site must be crawlable by AI bots. Without technical optimization, your content may remain invisible to AI.
Summary of key points:
Problem with JavaScript
The fundamental difference between Googlebot and AI crawlers lies in the technology. While Google has mastered JavaScript (JS) rendering, AI crawlers largely ignore JS due to the high computational costs involved.
- The Consequence: If your key information (prices, reviews, descriptions) is hidden within JavaScript, an AI bot will only see a blank page or bare HTML code.
- Solution: The most critical content must be part of the clean HTML source code rather than being loaded only after script execution.
Similarly, this applies to elements such as FAQs, which are frequently hidden behind click-to-expand actions (onclick events).
Strategies for enhancing AI visibility
To increase the chances of AI quoting you, focus on these areas:
- Structured Data (Schema Markup): Use schemas for articles and products. They enable bots to immediately grasp the context of your content.
- Authority and Freshness: Cite specific and, ideally, well-known authors. AI models actively look for signals of expertise and trustworthiness. Additionally, ensure your content is updated regularly.
- Semantic Blocks: Adopt an "Answer–Explanation–Elaboration" writing style. AI models favor clear definitions placed at the very beginning of paragraphs.
- User Experience and Core Web Vitals: For AI-driven search engines, it is essential that they link to fast, well-structured, and intuitive websites. Poor UX significantly reduces the likelihood of being cited.
Don't neglect continuous monitoring. Regularly verify that technical updates or site changes aren't inadvertently blocking AI bots from accessing your content.
New crawler: Google Messages
Google has officially expanded its list of crawlers to include a new assistant called Google Messages. This bot falls into the category of "user-triggered fetchers", meaning it does not activate autonomously while browsing the web, but is instead triggered by a specific user action.
Key information:
- Purpose: The bot is utilized solely to generate link previews for URLs shared between users within Google Messages chat sessions.
- Identification (User-Agent): You will find it in your server logs identified as GoogleMessages.
The primary reasons why Google created this specific bot and separated it from the standard Googlebot are:
Readability of logs for website owners
Before the introduction of this specific bot, it wasn't always clear what was generating requests to your server. When you saw "Google access" in your logs, it could have been a crawler indexing your site for search, or simply a technical request from an application.
- The Reason: Google wants you to clearly distinguish in your server logs: "This isn't an indexing attempt; someone just shared a link to my website in a chat."
Enhanced chat user experience (Link Previews):
When someone sends you a link to an article, you want to see a preview (headline, image, snippet), not just a plain URL.
- The Reason: The GoogleMessages bot is an optimized fetcher that quickly visits a website, downloads metadata (e.g., Open Graph tags), and generates a visual card within the Google Messages app. By maintaining its own distinct identity, it can communicate to the web server: "I only need the metadata for a preview; there is no need to download the entire page for indexing purposes."
Separation from the search algorithm (SEO):
This bot is a so-called "user-triggered fetcher"—it is activated by a specific user action (such as inserting a link into a message).
- The reason: By doing this, Google signals clearly that this visit has no impact on search rankings. If the standard Googlebot handled these requests, a website owner might mistakenly believe their page is being re-indexed, leading to incorrect SEO conclusions.
Google Messages is the official Google application for sending SMS, MMS, and modern RCS messages on Android devices—essentially serving as the counterpart to Apple's iMessage.
Google Personal Intelligence in AI Mode
Google is expanding its AI Mode capabilities to include what it calls personal intelligence. The primary objective is to bridge global information from the web with your private context, ensuring that results are not just generic, but tailored specifically to your life.
Imagine you are planning a vacation and searching for hotel accommodations. You don't even need to provide the AI with specific details. Based on your permission to access your Google Workspace history, the AI will already know that you’ve purchased flights to Boston and that you're traveling with your family. Consequently, it can proactively offer family-friendly lodging in a prime location that aligns perfectly with your existing itinerary.
Key features and connections
- Gmail and Photos Integration: Users can grant AI Mode access to their emails and photo library. This allows the AI to actively process your upcoming plans, purchase history, and visual memories to provide a truly personalized experience.
- Smarter Recommendations: If you are searching for a day trip, the AI will cross-reference the hotel confirmation in your Gmail and analyze your photo library (perhaps noting frequent visits to bakeries or sweet shops) to proactively suggest a specific itinerary designed to entertain the entire family.
- Personal Shopping Assistant: When searching for apparel, the AI will account for your preferred brands based on past purchase history and cross-reference them with the real-time weather forecast for the destination found in your emailed flight itinerary.
- Creative Inquiries: You can even pose abstract questions, such as "Suggest a movie title about my life,"based on your personal data and history.
Privacy and security
- Opt-in Principle: This feature is entirely voluntary. Users must explicitly choose to link their applications and can revoke access at any time.
- No Training on Private Data: Google states that the Gemini 3 model, which powers this feature, does not learn directly from the content of your emails or photos. Only anonymized prompts and model responses are used to improve the system.
- Error Correction and Feedback: Since this is an experimental feature, the AI may occasionally correlate unrelated information. Users have the ability to correct responses or provide direct feedback to improve accuracy.
Availability
- The feature is currently available as part of the Labs program for Google AI Pro and AI Ultra subscribers.
- It is currently limited to personal accounts within the United States and is available exclusively in English.
- Not available for corporate or school Workspace accounts.
ChatGPT and ads
It’s official: ChatGPT will begin testing ads in the United States over the coming weeks for users on the "Free" and "Go" tiers.
Principles for advertising:
- Independent Responses: Advertisements do not influence ChatGPT's responses. Answers are optimized based on what is most useful to the user. Ads are always kept separate and clearly labeled.
- Conversation Privacy: Conversations with ChatGPT will be protected from advertisers, and user data will not be sold to third-party advertisers.
- User Control: Users can opt out of ad personalization or delete the data used for targeting. A paid version will always remain completely ad-free.
- No "Time-mining": The goal is not to maximize the time a user spends within the application (as seen with social media), but rather to ensure that advertising serves only as a useful supplement.
- Protecting the Vulnerable: Advertisements will not be displayed to individuals under the age of 18, nor will they appear alongside sensitive topics such as health, politics, or mental well-being.
Who is affected by ads:
- Ads Will Be Included In: The Free version and the ChatGPT Go tier.
- Remaining Ad-Free: The Plus, Pro, Business, Enterprise, and Education plans will remain entirely ad-free.
What it's gonna look like:
- Placement: Advertisements will begin appearing in the United States over the coming weeks, positioned primarily below the generated response.
- Interactivity: Unlike regular banners, ads in ChatGPT will be conversational. You will be able to ask the ads directly for product details, helping you make a better informed buying decision.
- Context: Advertisements will only be displayed when they are strictly relevant to the current topic of your conversation.
For now, this change only affects users in the United States, but it represents a fundamental shift in how the service operates. It is highly likely that this new feature will reach Europe in the near future.
Shopping with an agent
Google is introducing an open standard for agentic shopping and new AI tools designed to help e-commerce businesses reach customers and boost sales.
Google, in collaboration with giants like Shopify, Walmart, and Target, has established the UCP standard. This serves as a universal communication bridge, allowing any AI agent (such as your personal assistant) to enter an e-shop, check shipping options, and pay securely without requiring a specific integration for every individual brand.
- Universal Commerce Protocol (UCP): A new open standard that enables AI agents and systems to collaborate seamlessly throughout the entire shopping journey. Instead of requiring a unique connection for every agent, UCP allows for universal interaction between e-shops, agents, and payment providers.
Users in the US can now purchase goods instantly within AI Mode or the Gemini app. Thanks to the AP2 protocol, payments are secure and processed via Google Pay or PayPal, while the e-shop itself remains the merchant of record.
- Agentic Checkout: A feature within AI Mode in Search and the Gemini app that enables purchases directly during product research, utilizing payment information stored in Google Wallet."
E-shops (e.g., Lowe’s or Reebok) can activate their own Business Agent within the Merchant Center. This functions as a digital sales assistant that learns from brand data and can advise customers on their selection or offer alternatives for out-of-stock items directly within the chat.
- Business Agent: A virtual sales assistant directly within Search that answers inquiries in a specific brand's voice and helps close sales.
- New Data Attributes: Dozens of new fields in Merchant Center for conversational search (e.g., compatible accessories, answers to frequent questions).
Direct Offers: This new pilot program in Google Ads allows the AI to detect when a customer is hesitating. In such cases, it can offer a real-time discount (e.g., 20%) to motivate them to complete the transaction immediately within the AI interface.
With this move, Google is transforming search from a tool that merely links to products into a platform where purchases take place directly under the guidance of AI agents.
AI vs. Google Search
Similarweb's report for early 2026 clearly shows that ChatGPT's lead is beginning to narrow in favor of Google.
Key information:
- Market Share Shift: Google Gemini is significantly gaining ground. While ChatGPT controlled 86% of the web chatbot direct visit market a year ago, it has now fallen to 64.5%. During the same period, Gemini grew from 5% to 21.5%.
- Other Players: Other tools such as DeepSeek (3.7%), Grok (3.4%), Perplexity (2.0%), and Claude (2.0%) remain significantly behind the two main giants in terms of direct web visits.
- Decline of Specialized Tools: Platforms focused purely on copywriting and content generation have seen a 10% decline. Tools like Jasper or Writesonic are reporting double-digit drops, suggesting that users are migrating to universal assistants (like Gemini/ChatGPT) or utilizing direct in-app integrations.
- Traditional Search: Classic search engines (Google Search) are not yet experiencing a drastic slump; their traffic has decreased by only 1% to 3% year-over-year.
The market is thus stabilizing around two main poles (OpenAI and Google), while specialized services must find new ways to reach their users.
SEO vs. GEO – Google's pragmatic approach
John Mueller's perspective from Google on the current debate regarding traditional SEO. John Mueller's main message is the transition from theory to practice. While the marketing community enthusiastically debates the emergence of a new field called GEO, Mueller recommends staying grounded and focusing on real data.
- The Real Impact Is Still Small: A study on SEJ states that the average website receives only 0.19% of its traffic from ChatGPT. Overall, AI assistants account for less than1% of traffic for most publishers.
- The End of "SEO Tricks": The debate about GEO essentially forces content creators to return to their roots—creating content so unique and valuable that AI tools not only cite it, but users also have a real reason to actually click through to the source website.
- Individual Strategies: There is no one-size-fits-all guide. If you run a website for programmers (who use AI extensively), GEO will be critical for you. If you sell goods to mainstream consumers, traditional channels still dominate.
Mueller’s message is that there is no need to panic and rename all SEO departments to GEO. However, it is essential to monitor where your customers are coming from and adapt to the fact that AI has become a permanent part of the digital ecosystem.
ChatGPT, health section
ChatGPT Health is not just a new feature, but a completely separate and secure space within the app designed for analyzing your personal health data. The goal is to help users better understand their bodies and prepare them for a qualified discussion with a doctor.
With this step, OpenAI is responding to the fact that hundreds of millions of people are already using AI to consult on their lifestyle and health status.
Connection with user data
The biggest innovation is the ability to securely connect your real health records and apps:
- Medical Records (EHR): Importing laboratory test results, clinical history, and physician reports (currently primarily in the USA via the b.well network).
- Wellness Apps: Integration with Apple Health, MyFitnessPal, Peloton, Function, or Weight Watchers.
- Practical Use: AI can explain trends in your cholesterol levels, suggest a diet when taking specific medications (e.g., GLP-1), or summarize your medical history for a new doctor.
Maximum emphasis on privacy and security
Due to the sensitivity of the data, this space operates in a high-protection mode:
- Isolation: Information from "Health" does not flow into regular chats. The health section has its own memory and specific encryption.
- No Training: Conversations in this section are not used to train OpenAI's underlying models.
- Control: The user has the option to delete data or disconnect linked apps at any time.
Created by doctors, not as a replacement for them
OpenAI emphasizes that the tool is not a replacement for clinical care and does not provide diagnoses:
- Over 260 physicians from 60 countries participated in the development.
- The model is tuned according to clinical standards (HealthBench) to communicate clearly and safely, and to know when to strongly refer the user to a professional.
Availability
- It is currently being launched for a small group of test users.
- In the coming weeks, access will be extended to the web and iOS for the general public.
- Electronic health record integration is limited to the US for now.
Instead of leafing through confusing PDF files with blood results or tracking charts in five different apps, you can ask ChatGPT for a comprehensive overview. For example: "Based on my Apple Health data and my latest blood tests, what questions should I ask my doctor at my check-up tomorrow?"
Google Trends and Gemini
In mid-January, Google announced a major update to its Google Trends tool. The new "Explore" page now utilizes Gemini AI to make it easier for users to interpret search data.
How Gemini improved Google Trends:
- Automated Comparison: If you enter a topic (e.g., dog breeds), Gemini automatically identifies and plots up to 8 relevant terms (e.g., Golden Retriever, Beagle) on a graph, allowing you to instantly see their relationship to one another.
- Contextual Suggestions: The panel offers broader categories (e.g., "hypoallergenic breeds") to help you explore the topic in greater depth.
- Interactive Prompts: Gemini suggests specific questions you can ask the AI to gain deeper insights about a given trend.
Previously, it was up to the user to come up with related queries for comparison. Now Gemini does it for you.
Guide for e-shops to AEO and GEO
In January 2026, Microsoft released a strategic playbook titled "From discovery to influence: A guide to AEO and GEO," designed to help e-commerce businesses optimize for AI agents, assistants, and generative search engines.
Three paths to visibility in AI
To ensure products appear in recommendations, companies must manage data on three levels:
- Indexed (crawled) Data: Information that the AI learned during training or extracted from websites. These form the basic perception of the brand and its reputation.
- Product Feeds and APIs: Structured data that companies actively send to platforms. They ensure accuracy in comparison tools and recommendations.
- Live Website Data: What AI agents see in real-time directly on your website (reviews, dynamic pricing, stock availability). If the website doesn't function correctly, the AI agent will not complete the purchase.
The manual defines three key areas of action to improve discoverability:
Structured data
- Schema Implementation: The basics such as Product, Offer, AggregateRating, Review, Brand, ItemList, and FAQ.
- Dynamic Fields: Ensure that the markup includes the current price, availability, color, size, and SKU/GTIN codes.
- Real-time Synchronization: Align the data in your product feeds with the live website data so that AI agents do not encounter conflicting information.
Content enrichment
- Benefit-Focused Descriptions: Product titles should include key differentiators (e.g., “Waterproof 3-Season Jacket” instead of just “Jacket”).
- Modular and Quotable Content: Create Q&A blocks that AI can easily extract and quote.
- Multimodal Signals: Write detailed alternative texts (alt text) for images and provide video transcripts so that AI can understand visual content as well.
Authority
- Verified Reviews: Use rating schemas and highlight the ratio of verified purchases.
- Authoritative Identity: Link your products to expert reviews and present certifications (e.g., sustainability) as factual entities.
- Content Integrity: Avoid exaggerated or unverifiable claims; AI systems may penalize language with low trust scores.
The manual explains the shift from traditional SEO focused on click-through rates to new disciplines that build brand clarity and credibility within the AI ecosystem.
Sources
- https://openai.com/cs-CZ/index/our-approach-to-advertising-and-expanding-access/
- https://blog.google/products-and-platforms/products/search/google-trends-explore-with-ai/
- https://about.ads.microsoft.com/content/dam/sites/msa-about/global/common/content-lib/pdf/from-discovery-to-influence-a-guide-to-aeo-and-geo.pdf
- https://blog.google/products/ads-commerce/agentic-commerce-ai-tools-protocol-retailers-platforms/
- https://www.searchenginejournal.com/google-gemini-gains-share-as-chatgpt-declines-in-similarweb-data/564690/
- https://www.similarweb.com/corp/wp-content/uploads/2026/01/attachment-Global-AI-Tracker-6.pdf
- https://www.searchenginejournal.com/googles-mueller-weighs-in-on-seo-vs-geo-debate/564583/
- https://openai.com/cs-CZ/index/introducing-chatgpt-health/
- https://blog.google/innovation-and-ai/products/gemini-app/personal-intelligence/
- https://www.seroundtable.com/new-googlebot-google-messages-40798.html
- https://developers.google.com/crawling/docs/crawlers-fetchers/google-user-triggered-fetchers#google-messages
- https://www.womenintechseo.com/knowledge/ai-crawlability-for-ai-search/
Martin Hill
Martin has been doing SEO since 2012. When he’s not in the woods, on his bike, or with a fishing rod by the water, he’s the head of the Keypers SEO team, where he’s been since 2017. Starting out on his own websites, he now focuses on strategy, technical SEO and project management. He enjoys finding ways to do things better — whether it’s processes, deliverables, or connecting SEO to other channels. SEO is a strategic detective for him — finding what’s not working, understanding the context and designing a solution that makes business sense.