Google Tests ‘Dive Deeper’ Feature in Discover, Expanding Generative AI Integration for Video Content

MOUNTAIN VIEW, California — In its ongoing effort to weave artificial intelligence deeper into everyday search and content discovery experiences, Google has launched a new experimental feature for its Discover feed. Dubbed “Dive Deeper,” the functionality is designed to provide users with a rapid, AI-generated synthesis of trending topics accompanied by curated links to original journalism, community reactions, and related media.

Robby Stein, Vice President of Product for Google Search, formally announced the experiment on X (formerly Twitter). Initially rolling out exclusively for video content, the feature represents the latest step in Google’s aggressive strategy to leverage large language models (LLMs) to keep users engaged within its ecosystem while simultaneously directing traffic to web publishers.

As digital marketers, SEO professionals, and publishers pore over the implications of the test, questions remain regarding geographical rollouts, user adoption rates, UI/UX iterations, and how these AI-driven impressions will ultimately be tracked within Google Search Console.

1. Main Facts: What Is the ‘Dive Deeper’ Feature?

The core premise of the “Dive Deeper” test is to reduce friction when a user wants to expand their understanding of a specific topic they encounter via their Google Discover feed.

When users scroll through Discover, they may encounter specific content cards—starting initially with video content—featuring an accompanying AI-generated summary. Below this summary sits the newly introduced “Dive Deeper” button. Tapping this button opens a dedicated interface that includes:

• A concise, AI-crafted overview of the broader topic or event.
• Prominent source cards and direct links to original reporting from various publishers.
• Related community reactions and discussions across the web.
• Sub-topic headlines grouped under an “Explore more stories” banner, offering further avenues for exploration.

Google has made it clear that the feature is in its infancy. The company is actively testing multiple design configurations and interface layouts to determine what yields the best user experience. Because the feature relies heavily on generative AI, instances utilizing these summaries carry the standard disclaimer: "Generated with AI, which can make mistakes."

2. Chronology: The Evolution of Discover’s AI Integration

To understand the significance of the “Dive Deeper” experiment, it is essential to trace how Google has progressively integrated generative AI into the Discover feed over the past year:

• October 2025: Google rolls out AI previews on trending topics within the Discover feed in select international markets, including the United States, South Korea, and India. These previews are designed to dynamically expand, giving users quick context paired with web links before they even click through to a publisher’s site.
• August 31: Google officially launches a generative AI performance report globally within Google Search Console. This report is designed to help site owners track visibility originating from generative features inside Discover.
• Late September: Google updates its official Search Console documentation to clarify that the new AI report captures impressions generated by links shown within Discover’s generative AI features, though Search Labs experiments are intentionally excluded.
• Friday (Current Announcement): Robby Stein takes to X to announce the public-facing experiment of the “Dive Deeper” button, kicking off the test with video-based Discover cards and promising iterative design changes over the coming weeks.

3. Supporting Data and Visual Breakdown of the Test

While Google’s official announcement was brief, the accompanying visual assets—consisting of a three-screen mockup shared in Stein’s post—provide profound clarity on how the feature functions in practice:

Screen One: The Entry Point

The user interacts with a standard Google Discover feed. In the provided example, the feed displays a YouTube video card originating from local news outlet KCRA 3, focusing on a Northern California heat wave. Directly beneath the video card, a brief AI-generated summary contextualizes the event, accompanied by the mandatory AI disclaimer and the new “Dive Deeper” button.

Screen Two: The Intermediate Overview

Tapping the button transitions the user to a secondary screen featuring the Google corporate branding at the top. The interface retains the core video content but supplements it with a much more robust, expanded summary. Below this expanded text, a horizontal carousel of source cards appears, prioritizing direct attribution—with the first prominent card linking directly to kcra.com.

Screen Three: Deep Exploration

The third screen shifts focus toward a broader exploratory experience, sporting the header “Explore more stories.” Here, the AI organizes the broader narrative into sub-topic headlines. Each sub-topic is paired with targeted source links and a brief snippet detailing what the linked coverage entails, effectively transforming a single video card into a multi-faceted journalistic hub.

4. Official Responses and Industry Context

Google’s product team has framed the test as a user-centric initiative designed to make exploratory research smoother and more intuitive.

"Today we’re experimenting with a new way to make exploring a topic and discovering related web content easier in Discover," Robby Stein wrote in his announcement post on X.

Elaborating on the functionality, Stein added:

"By tapping ‘Dive Deeper,’ you’ll see a short topic overview with prominent links to related stories, community reactions & original reporting. We’ll test this experience on videos to start, and experiment with multiple designs in the weeks ahead as we iterate to learn what works best."

Despite these details, Google has remained tight-lipped regarding several critical logistical metrics. The company has not disclosed:

• The specific countries or linguistic markets currently included in the test pool.
• The exact percentage of global or regional users seeing the button in their feeds.
• Whether the test is being administered strictly through Google’s Search Labs testing ground or deployed broadly to standard consumer accounts.

5. Implications for Publishers, SEOs, and Analytics

The rollout of “Dive Deeper” carries substantial implications for the digital publishing ecosystem, search engine optimization (SEO) strategies, and performance analytics.

Traffic Shifts and User Engagement

For publishers, any feature that keeps users within an aggregated Google interface—even if it provides prominent outbound links—creates a complex dynamic. On one hand, the inclusion of direct source cards and original reporting links in the expanded overview offers a vital referral traffic channel. On the other hand, a satisfying AI-generated summary might occasionally satisfy a user’s immediate informational needs without requiring them to visit the underlying website.

Because the initial rollout targets video content (such as YouTube clips), publishers who successfully tie their written journalism to multimedia packages may find themselves well-positioned to capture visibility within these deeper exploration screens.

Tracking and Search Console Complications

A pressing concern for webmasters centers on measurement. Google Search Console recently introduced a global performance report dedicated to generative AI traffic within Discover. However, the current reporting mechanics present notable limitations:

• The report aggregates data globally by page, country, and date, relying strictly on impressions as the primary metric.
• The reporting interface lacks a granular breakdown by specific features. Consequently, if impressions are generated via a “Dive Deeper” overview, they are currently lumped together with all other generative AI Discover features.
• It remains ambiguous whether impressions derived from this specific test are being tracked under the standard Search Console AI report or if they are walled off as Search Labs experiments (which are officially excluded from standard reporting).

As search analytics professionals attempt to audit traffic fluctuations, the inability to isolate “Dive Deeper” traffic from broader Discover AI previews makes performance attribution exceptionally difficult.

6. Looking Ahead

As Google continues to iterate on the design and user experience of the “Dive Deeper” button over the coming weeks, publishers and SEO experts will be watching closely for expansion signals.

While the test is currently confined to video cards, the foundational architecture—summarization paired with multi-source attribution and sub-topic exploration—could easily be scaled to standard web articles, image posts, and breaking news alerts. For now, the digital marketing community must monitor their Search Console data closely, adapt to shifting user behaviors within the Discover feed, and prepare for a future where generative AI acts as the primary gatekeeper and curator of mobile web traffic.