Go back to all blogs

Digital news to watch: Top Stories roll out in Google AI Overviews

5 min read
Digital news to watch: Top Stories roll out in Google AI Overviews

In this week’s digital news, Google has rolled out news updates and top stories within the AI Overviews section in Google Search. Reddit execs are “intensely debating” whether feeding ChatGPT is eating its own ad business.

Google’s John Mueller and Martin Splitt discuss how quality is more than just text, that the user experience can have something to do with it. Meta has published research on a new ads system called Hierarchical Interest Representation, designed to better match users with relevant ads by mapping latent interests across a large graph of users, advertisers and products.

Top Stories roll out in Google AI Overviews

Google has begun integrating “Top Stories” directly into its AI Overviews, a move that fundamentally changes how news content is surfaced and consumed within the generative search interface. This rollout, which is now live for mobile users in the United States, effectively bridges the gap between real-time news reporting and AI-synthesised answers. By positioning reputable journalism inside the AI response, Google is creating a new primary surface for publishers to maintain relevance in an environment that often prioritises static summaries over dynamic news feeds.
For search visibility, this development is a critical lever for news publishers. Previously, appearing in the AI Overview required being cited as a source within the generated prose. Now, publishers have a dedicated visual carousel to capture user attention. For SEO professionals, this reinforces the importance of being “Top Stories” eligible, a status that depends on Google’s News policies, freshness, and topical authority, as it is now the most reliable way to secure a prominent, visual footprint within the increasingly dominant AI Overview experience.
Read more here.
Reddit’s ad chief admitted at Cannes that the company is still internally debating whether its content licensing deals with OpenAI and Google undermine its own ad targeting business. Reddit’s advertising proposition rests on the uniqueness of its community data, yet that same content is being sold to train large language models. The executive’s working distinction is that raw content is what gets licensed, while ad targeting relies on engagement signals Reddit does not share. However, he stopped short of calling the issue resolved. The admission is notable given Reddit’s ad products are pitched explicitly on the premise that its data advantage is unique and replicable nowhere else.
Read more here.

Google explains SEO connection of site quality to non-indexed pages

Google’s John Mueller has provided critical clarification regarding why pages remain “crawled – currently not indexed,” suggesting that the root cause is often an issue with overall site quality rather than a technical SEO error. Mueller emphasised that quality is holistic, encompassing not just the uniqueness of the text, but the entire user experience, including intrusive ads, excessive interstitials, page performance, and whether the content is genuinely “valuable” compared to existing alternatives. He noted that if a site’s quality crosses a threshold of concern, Google’s systems may proactively limit indexing to protect the quality of its index.
This insight directs practitioners away from the common pitfall of endlessly searching for “technical bugs” when content fails to rank. Instead, brands should conduct quality audits that look beyond the text to the usability and utility of the page. By addressing “filler” content and ensuring a seamless page experience, publishers can resolve the underlying quality signals that often cause Google to deprioritise their pages, ultimately paving the way for improved indexing and visibility.
Read more here.

Exploring hierarchical interest signals in Meta Ads optimisation

Meta has published research on a new ads system called Hierarchical Interest Representation, designed to better match users with relevant ads by mapping latent interests across a large graph of users, advertisers and products. Using a transformer-based architecture trained on billions of interactions, the system learns unified embeddings that connect inferred user interests with advertiser offerings, even where engagement signals are sparse. It enriches sparse data using multimodal content such as text, images and video processed through LLMs. The outputs are intended to improve deep funnel performance across Meta’s existing ads infrastructure including its Generative Ads Model and Andromeda retrieval system.
Read more here.

The shift from SEO to social content

Search intent is being intercepted before it reaches a company’s website. AI overviews answer the query directly inside Google, and GenAI assistants like ChatGPT and Perplexity now guide most B2B buyers through research and shortlisting without a single visit to a vendor’s blog. That interception is why SEO leads that once arrived weekly now arrive monthly, and why the fix has to happen on the social platforms and communities where B2B buyers now spend that research time instead. The way through is repurposing what SEO already built, turning the case studies, reports, and articles into video, carousel, and thought leadership formats.

Read more here.

Google Gemini launch delayed as tech falls short of internal goals

Google has months behind schedule on delivering Gemini 3.5 Pro. The company has been taking its time trying to improve its capabilities, particularly in coding. The delay has frustrated Google engineers, AI researchers, and managers, many of whom are concerned that the company risks losing an edge. The model is currently being tested with partners, and Google is productively engaged with the US government on model testing and broader frameworks.

Read more here.

If you need more information on any of these stories, or you’d like support on any of your digital marketing campaigns. Get in touch with our team of experts by dropping us an email at [email protected] 

Author

Stay in the loop
Share post
Copy Link

hi

Other posts you might like

How to protect your affiliate programme from fraud

How to protect your affiliate programme from fraud

TLDR: Affiliate fraud is more common than most programmes want to admit, but it’s also more preventable than most teams
Digital news to watch: New research shows AI visibility rankings aren’t stable

Digital news to watch: New research shows AI visibility rankings aren’t stable

In this week’s digital news to watch, AI search is reshaping how brands earn visibility, with earned media and PR

Popular topics

[other_categories]
Get In Touch
Get In Touch
Email Us
Email Us