The SEO, SEM and AI news for the week of 12 October 2026 brings several changes affecting campaign management, measurement and search visibility. Google has completed its spam update, Search Console has introduced a useful improvement for international projects, and Google Ads is adding new tools to control automation and evaluate results. OpenAI has also announced a visual advertising format for ChatGPT alongside new measurement integrations.
This article covers fifteen confirmed developments between 5 and 11 October. It continues the analysis from the previous week’s SEO, SEM and AI news, with the same objective: separating changes that deserve action in real accounts from limited tests and announcements that do not yet change anything for a business in Spain.
There are four issues I would place ahead of everything else: evaluating the Spam Update now that it has finished, reviewing Google Ads automated options, identifying changes that alter attribution figures and strengthening the security of advertising accounts. The remaining developments provide useful context, particularly for ecommerce and visibility strategies in generative search engines.
Google has confirmed the completion of a ranking update and published new guidance on content, while also expanding analysis capabilities and the way certain platforms can provide fresh information.
Google confirms in its official status dashboard that the update began on 24 September and ended on 8 October. The rollout affected all languages and lasted almost fourteen days. This changes the approach we were taking in the previous briefing: we are no longer dealing with an active rollout.
I would now compare the period before the update, the rollout itself and the subsequent results in Search Console, separating queries, pages and directories. I would not attribute every fall in clicks to a penalty. Seasonality, changes in demand, competition and CTR variations can also play a role. If the decline is concentrated in repetitive pages or low-value content, then a deeper audit would be justified.
On 6 October, an expansion of Google’s helpful content guidance was reported. It discourages the creation of fictitious profiles, AI-generated photographs presented as real people and qualifications that do not exist in order to make it appear that an expert has written or endorsed a particular article.
The change is especially relevant to websites publishing content at scale and using artificial author profiles to communicate authority that does not actually exist. It does not amount to a new requirement for extensive author biographies. In a professional project, I would review editorial attribution, verifiable information and the consistency of the people and entities being cited. This is also relevant to GEO positioning and AI visibility, where a consistent brand identity makes information easier to interpret correctly, although it does not guarantee being cited.
Since 6 October, it has been possible to select more than one country within Search Console’s geographic filter. The new functionality simplifies the analysis of grouped markets, something that previously required exporting data or building external reports.
For example, a business serving several European countries can compare their combined performance without necessarily including all international traffic. It does not change rankings, but it does make reporting more consistent. For clients with several language versions, I would test the new filter and establish stable country groups for comparing equivalent periods.
On 8 October, Google documented the UGC Fresh Data Program. Approved platforms will be able to submit recently created user content and engagement signals through a channel separate from standard crawling and the Indexing API.
It is intended for forums and social platforms with large amounts of public content, stable individual URLs, moderation, author profiles and the technical ability to send structured data. It does not guarantee indexing or visibility. There is nothing to configure for a conventional corporate website. Presenting it as a new way to index any blog within minutes would simply be noise.
Most of this week’s changes are concentrated in advertising. Some make it easier to evaluate performance, while others modify creative assets or attribution criteria, so they are worth monitoring.
On 7 October, self-service access to user-based Conversion Lift studies for Search and Performance Max was announced for eligible accounts. The published information includes substantial requirements, including at least one thousand observed conversions and a participating budget of at least $5,000.
The value of the tool lies in incrementality. Traditional attribution tells us which conversions are assigned to advertising; a study with a control group attempts to estimate how many would not have happened without that exposure. It can be useful for large accounts, but it requires sufficient volume, a clear hypothesis and a reliable business objective. It is not a feature that should be recommended indiscriminately for small local campaigns.
Google Ads is testing a tool that proposes a Target ROAS based on average margin before advertising costs. According to the demonstration published on 6 October, a 15% margin would correspond to a ROAS of approximately 667% in order to cover advertising costs.
The interesting part is moving bidding decisions closer to profitability, although an average margin can conceal differences between products, returns and logistics. When managing ecommerce accounts, I would review the conversion value being sent to Ads and the actual profit remaining on each order. The calculator is still in beta and its result should not be confused with a recommendation that is automatically appropriate for every business.
On 8 October, Google introduced the ability to preview up to ten headlines and descriptions generated from a URL before enabling AI Max text customisation. One day earlier, additional columns had also been announced for reviewing locations of interest, optimised targeting and brand inclusions.
Both improvements relate to control. We can identify unsuitable advertising messages before activating them and locate automation settings within complex accounts. In a Consultor SEO Alicante
My experience with local ChatGPT Work and Work Cloud: files, browsers, autonomy and context in SEO, website migrations and development… Leer más
Leave a Comment