Nội dung
- ChatGPT Ads begins automatically creating ads for advertisers
- ChatGPT is dominating AI referral traffic
- Website architecture becomes the foundation of SEO in the AI era
- Six SEO priorities to revisit as AI search rises
- Google Open Knowledge Format: a new hint for organizing knowledge on websites
- Google says Cloudflare’s Content Signals has no real effect
- ChatGPT Ads lets AI generate ads, then have them reviewed manually
- Google’s AI Overviews continue expanding the “Further Exploration” feature
- A perspective for the Vietnamese market
- References
This week, SEO continues to shift decisively toward an “AI search first” landscape, from how Google displays results to how users are guided deeper into AI Overviews, and even to platforms like ChatGPT beginning to touch advertising. For Vietnamese marketers, this is not just a matter of tracking tech news, but a signal to adjust content strategy, technical foundations, and resource allocation sooner.
-
Key points:
- ChatGPT Ads has been noted as being able to automatically generate ad variations for users to review and activate.
- Data from 6.77 million sessions shows ChatGPT accounts for the majority of AI referral traffic, reflecting a very high level of concentration.
- SEO is being forced to rethink many priorities as AI search and AI Overviews play an increasingly important intermediary role between users and websites.
- Moves around website structure, knowledge format, and robots.txt show that being “AI-readable” is becoming a practical requirement, no longer just an experimental concept.
ChatGPT Ads begins automatically creating ads for advertisers
Search Engine Roundtable reported that OpenAI’s ChatGPT Ads platform has introduced a feature that suggests prebuilt ad templates. According to the interface description, when users click “add new ad,” the system can generate an ad variation based on the website and campaign settings, after which users only need to review, edit, and approve it before launch.
The notable point is that the “template creation” step is no longer entirely manual. This suggests OpenAI is pushing an AI-assisted advertising model, where machines handle the drafting while humans retain final control. For marketers, this signals that copywriting, variation building, and creative testing may increasingly be automated directly within ad platforms.
Source: Search Engine Roundtable, Search Engine Land.
ChatGPT is dominating AI referral traffic
For Vietnamese businesses, all eight signals point to one common conclusion: SEO is no longer just a game of blue rankings on Google. As AI search and conversational assistants grow, brands need to invest more heavily in information architecture, editorial quality, content clarity, and the ability to be trusted and cited by AI.
According to Search Engine Land, data from more than 6.77 million visits shows that ChatGPT accounts for as much as 92% of AI referral traffic. This figure not only highlights ChatGPT’s popularity, but also reflects an important reality: if your website has not yet been discovered through AI assistants, you may be missing a rapidly growing acquisition channel.
For SEO, the significance of this data lies in the fact that language models are no longer just “passively reading” content; they are becoming traffic distribution gateways. When one source dominates so heavily, dependency risk rises as well: changes in product, interface, or citation behavior can cause major traffic swings. Therefore, content strategy should prioritize being understood, cited, and directed back to the website in AI environments.
Source: Search Engine Land.
Website architecture becomes the foundation of SEO in the AI era
Search Engine Land’s SMX Now continued to emphasize the topic of building better site architecture for SEO, AI, and users. The message here is quite clear: if a website’s structure is fragmented, even strong content will struggle to help search engines and AI systems understand the relationships between pages, topics, and entities.
In the AI search phase, site architecture is not only about making it easier for crawlers to collect data, but also about helping models infer topic, relevance, and information hierarchy more quickly. That makes elements such as content clustering, internal linking, logical navigation, and taxonomy systems far more important. For SEO teams, this is the time to work more closely with product, UX, and web development teams.
Source: Search Engine Land.
Six SEO priorities to revisit as AI search rises
Search Engine Land proposed 6 SEO priorities that should be reconsidered in the context of AI search. While the original content does not frame a single formula, the overall spirit is that modern SEO must serve three audiences at once: users, traditional search engines, and intermediary AI systems.
Instead of optimizing only for rankings, marketers need to think more about whether content can be understood in context, answer search intent correctly, and be trustworthy enough for AI to cite. This is a shift from “keyword optimization” to “optimizing usefulness and clarity.” Brands that do SEO well in this phase are often those with tight content structure, clear expertise signals, and a page experience strong enough to keep readers engaged.
Source: Search Engine Land.
Google Open Knowledge Format: a new hint for organizing knowledge on websites
Search Engine Journal analyzed the Open Knowledge Format (OKF) that Google introduced as a way to turn a body of knowledge into a directory of interconnected markdown files. Initially, this format was designed for internal enterprise data, but the article argues that it also suggests how websites can organize knowledge so AI can read and understand it better.
The core idea is that each concept can have its own page, along with a concise YAML description and logical links to related concepts. Compared with a “read each page in isolation” model, this knowledge-graph-style structure helps AI see the connections between ideas. For content websites, this is a noteworthy cue for in-depth knowledge pages, resource centers, or content hubs.
Source: Search Engine Journal.
Google says Cloudflare’s Content Signals has no real effect
Search Engine Roundtable cited John Mueller (Google) as saying that the Content Signals directive in robots.txt introduced by Cloudflare “has no effect” on crawlers or LLMs. He also said it only makes robots.txt more cumbersome and increases maintenance costs later on.
The message is quite practical for technical SEO professionals: not every proposal that sounds AI-relevant will deliver real results. If a directive is not yet used by crawlers or models, rolling it out broadly may add complexity without proportional benefit. Therefore, instead of chasing every new standard, SEO teams should prioritize proven signals and focus on structure, content quality, and accessibility.
Source: Search Engine Roundtable.
ChatGPT Ads lets AI generate ads, then have them reviewed manually
In another Search Engine Roundtable article, the “Generated ads for you” feature is described as a step forward in the ChatGPT Ads platform: the system pre-generates an ad variation based on the website and campaign settings, then the user reviews, edits, and confirms it. This works like an “AI suggests – humans approve” workflow.
In practice, this is a model that can shorten message-testing time, especially for smaller marketing teams. However, it also raises the bar for input quality: the website, offer, product description, and campaign content need to be clear enough for AI to generate a draft that closely reflects reality. If the input data is poor, even fast-generated ads will struggle to achieve good quality.
Source: Search Engine Roundtable.
Google’s AI Overviews continue expanding the “Further Exploration” feature
Search Engine Roundtable reported that users have begun seeing “Further Exploration” appear in Google Search AI Overviews. This is an additional guide that helps searchers go deeper into a topic after reading the initial AI answer. In other words, Google is turning AI Overviews from a short answer block into a transit point toward multiple layers of information.
For SEO, this means search results are no longer a static ranking table. They are a guided discovery journey. Websites with strong internal linking, clear explanations, and multi-layer topic coverage will have more opportunities to be pulled into that discovery flow. This is why hub content, FAQs, concept explainers, and in-depth comparisons need serious investment.
Source: Search Engine Roundtable.
A perspective for the Vietnamese market
In the short term, marketing teams should review website structure, core content clusters, and the foundational data on each page: titles, descriptions, internal links, FAQs, schema, and message consistency. In the long term, the advantage will belong to brands that build content like an organized knowledge system, rather than simply producing disconnected articles to chase keywords.
This is also the time for Vietnamese marketers to work more closely across SEO, paid media, content, and dev. If ChatGPT Ads can automatically generate ad variations, while Google continues expanding AI Overviews, then the real value will lie in content strategy and input data that are strong enough for both humans and AI to understand and trust.
See more marketing news and guides at https://marketing365.vn.
Follow more updates from Marketing365 to stay up to date with the latest marketing trends.
Read more articles in the same category Digital Trends.
This article focuses on SEO trước làn sóng AI Search with a perspective for the Vietnamese market.
References
- OpenAI can generate ChatGPT Ads for you
- ChatGPT commands 92% of AI referral traffic. Here’s what 6.77 million sessions reveal.
- SMX Now: Build better site architecture for SEO, AI, and users
- 6 SEO priorities to rethink for AI search
- Google’s Open Knowledge Format Could Work For Websites, Too
- Google: Cloudflare Content Signals Robots.txt Directive Has No Effects
- ChatGPT Ads Can Generate Ads For You
- Further Exploration Found In The Wild Within Google AI Overviews
