Conference updates put AI Overviews and conversational search at the centre of Google’s interface, tightening the loop between answers and commerce while reshaping referral traffic, advertising economics, and investor research workflows.

This week, Google is putting artificial intelligence at the centre of search, and the move reads as a structural shift rather than a feature update. Thomas Gardner, who oversees private equity at Sunnov Investment Pte. Ltd., characterises the shift as “a move from links to decisions, where the answer sits in front of the click and the transaction sits beside the answer

AI Overviews now reaches more than 2.5 billion users in an average month, while the conversational AI Mode experience clears 1 billion monthly users, scale that lands directly in advertising because search remains a core driver of Alphabet’s latest reported annual revenue base of $402.8 billion.

Google’s I/O developer conference details a redesigned search box built for longer prompts and richer follow-ups, alongside multimodal inputs such as images, files and video. The interface expands as a user types, nudging behaviour away from keywords towards intent, and keeping the conversation in one place. For publishers and retailers, the more immediate question becomes how much traffic stays on-platform when summaries, comparisons and next steps appear before a user reaches an external site.

In Sunnov Investment’s briefing, the unification of AI Overviews and conversational mode is the organising principle, preserving context across an exchange and using query fan-out to search subtopics in parallel. The system draws on sources such as the Knowledge Graph, live finance data and shopping catalogues that span billions of products, then surfaces links in support of a generated answer. “Context persistence turns search into a workflow rather than a page view, and that changes how investors source information, how brands buy attention, and how publishers price their influence” is Gardner’s assessment.

That workflow runs on Gemini 3.5 Flash as the default model in the current upgrade, with benchmark comparisons accompanying the change putting output speed at roughly four times comparable systems. The same stack enables generative user interfaces that assemble interactive charts and layouts in real time, a shift that compresses research from hours into minutes for routine questions. For Gardner, “speed matters, but controllable context matters more, because it lets an investor interrogate a thesis without losing the thread”.

Agentic monitoring extends the idea beyond search. Market forecasts over the next five years project the AI agents sector growing from about $7.8 billion to about $52.6 billion, implying a 46.3% compound annual growth rate across that window. In executive surveys fielded over the most recent quarter, 53% report production deployments of AI agents and 40% report more than ten distinct agents, with early adopters citing workload reductions above 60% across reconciliation and data-entry tasks during a standard operating cycle. “The first order benefit is time, but the second order benefit is coverage, because agents do not sleep and investors do not have infinite attention” is Gardner’s view.

The trade-off shows up in referral economics. In a multi-week measurement of United States desktop searches, AI Overviews appears in roughly 20% of queries, and the presence of an overview correlates with a click occurring about 8% of the time, versus about 15% when an overview does not surface in the same window. Source links embedded in the overview attract clicks at roughly 1% frequency in that dataset, while top-position organic results register a 34.5% drop in click-through rate when the AI layer appears. Google is also introducing a universal cart across Search, Gemini, YouTube and Gmail, automatically tracking prices, inventory and loyalty benefits, with a phased United States rollout expected over the coming months. “If the answer and the checkout both stay on one surface, the negotiating leverage shifts from the open web to the gatekeeper” is Gardner’s warning.

Sunnov Investment sees the re-engineering of search as a near-term catalyst for faster diligence and tighter feedback loops, and as a longer-term reset for media and commerce business models that still depend on outbound clicks. “Capital markets reward the firms that adapt their distribution and data strategy to the new interface first” is how Gardner frames the competitive implication.

About Sunnov Investment

• Singapore-based investment manager founded in 2012, serving accredited investors, foundations and endowments worldwide. 
• Manages long-only equity strategies with complementary long/short equity, global macro, event-driven and systematic mandates, while developing structured routes for eligible retail participation. 
• Website: https://sunnov.com 
• Media enquiries: Deng Hui at d.hui@sunnov.com 
• Registered entity: Sunnov Investment Pte. Ltd., UEN 201225494E.