Google Is Testing A New Dive Deeper Button In Discover To Streamline AI-Powered Content Exploration

Google has officially begun testing a new feature within its Discover feed called the Dive deeper button, an innovative addition designed to provide users with a comprehensive topic overview accompanied by direct links to related web content, original journalism, and community reactions. The initiative marks the latest step in Google’s ongoing integration of generative artificial intelligence into its discovery ecosystems, fundamentally changing how users interact with content streams on mobile devices.

The experiment was publicly announced on Friday by Robby Stein, Vice President of Product for Google Search, through a post on the social media platform X. According to Google’s initial rollout parameters, the test is launching specifically with video content before potentially expanding to other media formats. By clicking the new prompt, users are taken away from a solitary piece of media and dropped into an AI-generated summary hub that contextualizes the underlying topic, offering a richer, multi-faceted look at trending subjects.

The move comes at a time when digital discovery platforms are heavily leaning into generative AI to keep users engaged longer by providing instant syntheses of complex news events. However, this transition also raises critical questions for publishers, SEO professionals, and digital marketers who rely on referral traffic from Google Discover. As the tech giant continues to iterate on these AI-driven interface designs, understanding the mechanics, tracking metrics, and broader implications of the Dive deeper feature is essential for stakeholders across the digital publishing landscape.

Anatomy Of The Test: What Google Is Deploying

According to the official announcement shared by Robby Stein, the core intent behind the experiment is to make exploring a complex topic and discovering related web content significantly smoother for everyday users of the Discover feed. Stein described the functionality of the new button in detail, noting that once tapped, users will be presented with a short topic overview prominently featuring links to related stories, original reporting, and relevant community reactions.

In the initial phase of the rollout, Google is deliberately narrowing the scope of the test to focus exclusively on videos. This targeted approach allows the product team to evaluate user engagement, click-through rates, and overall utility within a controlled environment. Stein noted that Google will experiment with multiple design variations over the ensuing weeks to iterate and learn what works best for the consumer audience. At present, Google has not disclosed the specific geographic regions, language markets, or the exact percentage of the user base included in this initial testing phase.

A visual breakdown of the mockups shared alongside the announcement illustrates how the user journey unfolds. On the first screen, a standard Discover feed presents a local news weather video sourced from YouTube regarding a Northern California heat wave. Beneath the video card sits a concise AI-generated summary accompanied by a mandatory disclaimer label reading, Generated with AI, which can make mistakes. The newly minted Dive deeper button is positioned directly beneath this summary block.

Upon interacting with the button, the user transitions to a secondary screen bearing the Google corporate logo, the original video, a more extensive narrative summary, and a horizontal carousel of source cards linking back to the originating publishers, such as local news outlets. The third and final screen in the demonstration sequence is titled Explore more stories. This section groups short summaries under distinct sub-topic headlines, each followed by direct source links and brief explanatory descriptions of the linked coverage.

The Evolution Of AI In Google Discover

The introduction of the Dive deeper button does not happen in a vacuum; it represents the continuation of a broader strategic push by Google to infuse generative artificial intelligence across its entire suite of discovery products. In October 2025, Google rolled out AI previews on trending topics within the Discover sports feed. That initial implementation allowed users to expand topic previews to access deeper contextual information coupled with direct links to the wider web. Those features went live simultaneously across the United States, South Korea, and India, setting a precedent for how Google handles dynamic content feeds.

The current test builds upon this foundation by moving beyond sports and general trending queries into broader multimedia content, specifically video cards. By attaching summaries directly to video assets and pairing them with an interactive expansion button, Google is attempting to bridge the gap between passive viewing and active content consumption. Users no longer have to perform a manual search to find out what other publications are saying about a breaking news story featured in a video; the synthesis and the citation pathways are served directly to them within the Discover interface.

Concurrently, the backend infrastructure supporting these features has been evolving. Google recently introduced a generative AI performance report within Search Console specifically tailored for Discover traffic. According to official company statements, this specialized reporting system rolled out to websites worldwide on late August, though webmasters have noted discrepancies regarding property access and metric definitions.

As of mid-September, official Google Search Console help documentation clarified that the new performance report specifically tallies impressions for links displayed within generative AI features inside Discover. The data can be filtered and grouped by page, country, and date. Notably, impressions currently stand out as the sole metric detailed on the help page, which does not explicitly itemize every individual feature that contributes to the AI impression count, though it does explicitly exclude experimental features running through Search Labs. It remains unconfirmed whether impressions generated by the Dive deeper overview will be isolated or lumped into the broader pool of generative AI Discover impressions.

Industry Implications And Publisher Concerns

For publishers and digital content creators, the deployment of features like Dive deeper carries significant strategic implications. On one hand, the inclusion of prominent links to original reporting and related stories within the AI-generated overview provides a potential pathway for referral traffic. Unlike pure zero-click search experiences that keep users entirely within the search results page, the second and third screens of the Dive deeper interface actively surface source cards and direct links to publisher websites such as local news stations.

However, the reality of AI-driven content aggregation remains a double-edged sword. When a user reads an AI-synthesized summary of a news event directly inside the Discover feed, the immediate informational need may be satisfied without the user ever clicking through to the primary source. Even when links are provided, click-through rates on secondary summary cards historically lag behind direct visits to original articles. Furthermore, the presence of disclaimers warning users that AI can make mistakes highlights the ongoing challenge of factual accuracy in automated summaries, placing the onus on publishers to ensure their original reporting is accurately represented by machine learning models.

The lack of granular reporting within Google Search Console further complicates matters for SEO and analytics teams. Because the Discover AI performance report primarily measures impressions without providing a clear breakdown of specific interface elements—such as distinguishing between a standard Discover card, an AI preview, or a Dive deeper overview—attributing traffic shifts and visibility changes to this specific feature will prove challenging. Publishers will need to monitor their analytics closely for organic fluctuations once the feature moves past the initial testing phase and receives a wider global rollout.

The Road Ahead: What To Expect Next

As Google continues its iterative testing process, product managers and engineers will analyze user telemetry to determine the optimal layout, tone, and functionality of the Dive deeper experience. Robby Stein’s public statements emphasize that multiple design iterations are planned for the coming weeks, meaning the visual presentation and user flow observed in the initial mockups are subject to change based on consumer feedback.

At this stage, Google has not provided a definitive timeline for moving the feature beyond video content or expanding the test to a broader demographic and linguistic audience. Much will depend on user reception during the initial trial period and how regulatory bodies scrutinize the integration of AI-generated summaries with publisher content.

As artificial intelligence continues to redefine the boundaries of search and content discovery, tools like the Dive deeper button underscore Google’s commitment to building a more conversational, synthesized web experience. Whether this evolution ultimately empowers independent journalism through structured attribution or concentrates user attention within Google’s proprietary interface will remain a central debate across the digital media landscape for the foreseeable future.