Contents
- SEO in the AI era: why rankings no longer tell the whole story
- Changes in visibility and measurement layers — what the SEO team now has to check
- Data costs, copyright and transparency are pushing SEO into a risk-management problem
- SEO in Vietnam will win by moving from page optimization to verifiable trust optimization
- What to do with SEO so you do not get stuck in old rankings
- References
SEO is no longer just about ranking for clicks. As AI Overviews, ChatGPT Search and increasingly automated answer layers intrude on the search journey, brands must compete not only to be seen but also to be chosen as a citation by AI and trusted by users. For Vietnamese marketers, this is a very real shift: traffic can fall even when content is still strong, so both SEO measurement and execution have to change.
Key points
- SEO is being pulled from a ranking race into a race to be cited by AI, trusted by users, and mentioned in the right context.
- Changes in ChatGPT Search, AI Overviews, content watermarking and measurement data all make one thing clear: clicks are no longer a good enough metric.
- Businesses need to manage content provenance, evidence, and new metrics such as brand citations, prompt coverage and AI referrals.
- In Vietnam, SEO will struggle to keep its old role if it still optimizes only for SERPs and ignores AI visibility and brand trust.
SEO in the AI era: why rankings no longer tell the whole story
The common thread in the latest developments is that the search journey is being bent out of shape. Search Engine Journal says ChatGPT Search uses even OpenAI‘s internal index to serve results, and Resoneo data shows hundreds of sites without content agreements are still being served in ways similar to licensed partners. At the same time, Search Engine Journal also stresses that many SEOs still measure success mainly by clicks, impressions and rankings, while AI has added brand citations, prompt coverage and AI referrals to the picture.
On Google’s side, Search Engine Land notes that AI Overviews continue to deepen the zero-click trend: the answer appears right on the results page, so users do not necessarily need to click through to a website. That means a page can still be “good SEO” in the traditional sense while becoming weaker at the final point businesses care about most: attention, recall and conversion. When traffic no longer follows a straight line, SEO can no longer be treated simply as the function that “brings people to the site.”
That is why the most important argument right now is not “Will AI kill SEO?” but “SEO is changing roles.” Its old role was to generate clicks. Its new role is to help a brand appear in AI answer layers, in cited source lists and in users’ minds when they make decisions.
Changes in visibility and measurement layers — what the SEO team now has to check
The update cluster shows that part of the infrastructure around SEO is changing at the same time. These are not minor shifts, because each one directly affects how marketing teams read data and defend the value of content.
ChatGPT Search Index: measuring whether a site is served by AI
Search Engine Journal says ChatGPT Search does not rely only on contracted partner sources, but also uses OpenAI’s internal index to return results for many smaller sites. This matters a great deal for SEO: brands cannot assume that only “big” sites or those with deals will be mentioned by AI. If AI pulls data differently from traditional web search, marketing teams need to track visibility inside the AI ecosystem, not just in SERPs.
For marketers, the task is to look at which query types are easy for AI to answer, which content has a high chance of being selected as a source, and which pages are losing clicks but still retain brand recall value. That is the foundation of the new measurement model.
Brand citations and AI referrals: measuring results before the click happens
Search Engine Journal’s article on metrics in the AI era makes one issue very clear: clicks, impressions and CTR only show the result after a user has already chosen, not why they were persuaded. When AI joins the answer process, metrics such as brand citations, prompt coverage and AI referrals become an earlier layer of measurement. In other words, brands need to know whether AI mentions them, in what context, and whether that drives users back.
The key point is that SEO is not losing its measurement role; it is being forced to measure more broadly. Anyone who still looks only at rankings will easily misread the real performance.
AI Overviews: traffic can fall even when content is still useful
Search Engine Land describes a reality many content teams know well: the question has already been answered by AI on the SERP, so clicks are reduced. For SEO, this changes how a piece of content is evaluated. An article may no longer drive as many sessions as before, but it can still nurture the brand, create authority signals and be chosen by AI as a source.
That is why measurement needs to clearly separate “seen,” “cited” and “clicked.” Bundling all of that into a single traffic line is the easiest way to blur SEO’s real impact.
Data costs, copyright and transparency are pushing SEO into a risk-management problem
The most notable part of this wave of change lies in compliance. SEO is not just about technical and content optimization, but also about managing how data is reused, how content is attributed and how AI handles inputs.
Anthropic’s watermarking: AI-generated content must leave a trace
Search Engine Land and Search Engine Journal both report that Anthropic has added watermarking and metadata to content and files created by Claude at scale, while linking this to transparency commitments under the EU AI Act. For marketing, this is a very clear signal: AI content will increasingly need identifiable markers and clearer provenance. If SEO teams use AI to produce content, keeping a record of how it was created, checked and edited will no longer be a side task.
This is especially important when a brand wants to prove it is not mass-producing content while still maintaining editorial standards, authenticity and accountability.
The legal response to AI Overviews: publishing rights and the right to attribution
Search Engine Land says around 300 French newspapers have filed a complaint over Google AI Overviews. Although the article does not go into the final ruling, the move shows a clear flashpoint: journalism content and web content are being pulled into the same value chain by AI, but the way effort and value are recognized is not equal. For SEO, that creates new pressure around content provenance, usage rights and the ability to be transparently cited.
Businesses doing SEO in Vietnam should read this as an early warning: the more you rely on AI to summarize, the more you need a process to prove whose content it is, where it came from and how it was edited.
Ads and import data: the link between search, paid media and reporting
Google has added a Campaign Data Import Validation Report and is testing an advertiser list above mobile Search ads. These changes are not purely SEO, but they show that the search ecosystem is being tightened at the data and visibility layers. When ads, organic and reporting all pass through an increasingly complex measurement system, SEO teams have to work more closely with analytics and paid media so they do not misread performance.
Simply put: SEO no longer stands alone. If import data, ad visibility and the AI layer are all shifting, SEO’s data-risk problem becomes much larger than before.
SEO in Vietnam will win by moving from page optimization to verifiable trust optimization
In Vietnam, this shift will arrive more slowly than in some larger markets, but that does not make it any less significant. Many businesses still see SEO as a cheap traffic channel, while users are gradually getting used to receiving answers directly from AI or from pre-summarized results. When that happens, a well-optimized article that lacks brand credibility can still be overlooked.
The difference in the Vietnamese market is that marketing teams often have to do many things at once: content, SEO, social and performance. That can be an advantage if used well, because a brand can reinforce the same message across multiple channels to increase the chance of being recognized by both AI and users. But without a process for attribution, factual accuracy checks and unified metrics, SEO can easily slide back into a model of producing lots of content while struggling to prove impact.
For Vietnamese marketers, the question should be: does my brand count as a source worth mentioning to AI, and if traffic falls, do I still know where SEO’s value lies?
What to do with SEO so you do not get stuck in old rankings
- Measure brand citations, prompt coverage and AI referrals in addition to rank and traffic.
- Review content likely to be summarized by AI and add evidence, sources and real expertise signals.
- Work with analytics and paid media to separate what is seen, what is cited and what is clicked.
- Set up a process to verify the provenance of AI content and keep an editing trail before publishing.
SEO is not disappearing. It is changing roles from “driving people to the website” to “helping the brand appear in answers, have evidence to be trusted, and have data to explain itself.” Anyone who still optimizes only for rankings will see traffic decline as a shock. Anyone who changes measurement and execution early will still preserve the real value of SEO.
See more marketing analysis and how-to guides at https://marketing365.vn.
Follow more analysis from Marketing365 to stay updated on the latest marketing trends.
Read more articles in the Digital Trends category.
References
- Search Engine Land — SMX Now: The Target bidding apocalypse and how we’ve been here before
- Search Engine Land — Anthropic adds AI text watermarking to Claude models worldwide
- Search Engine Journal — ChatGPT’s Search Index Serves Small Sites Too, Data Shows via @sejournal, @MattGSouthern
- Search Engine Journal — The Metric Most SEOs Are Still Not Measuring In The Age Of AI via @sejournal, @SequinsNsearch
- Search Engine Journal — Organic Traffic Dropped? What to Tell Clients & the AI Visibility Metrics to Report Instead
- Search Engine Land — 300 French newspapers file complaint over Google AI Overviews
- Search Engine Land — Google tests advertiser list above mobile Search ads
- Search Engine Land — Google adds Campaign Data Import Validation Report
- Search Engine Journal — Anthropic To Mark Claude Text & Files Under EU AI Act Code via @sejournal, @MattGSouthern
- Google AI Blog — AMIE, our research medical AI system, demonstrates real-time clinical video consultation capabilities in a first-of-its-kind study.
