[Edaily Reporter Lee So-Hyun ] More than half of users of Naver’s (NAVER(035420)) artificial intelligence (AI) search service, “AI Tab,” reported that they have taken action—such as purchasing actual products or making restaurant reservations—based on AI responses. In particular, users aged 40 and older showed a high shopping click-through rate, confirming that AI search goes beyond simply providing information to drive actual consumer activity.

On the 9th, Naver published the “Naver AI Tab Service Brief” on its website, which analyzed actual usage data from AI Tab in collaboration with a research team led by Professor Kim Kyung-oe of Yonsei University. The report is based on an analysis of what users asked and which services they accessed after viewing answers since AI Tab’s official launch on June 25.

An analysis of questions entered into the AI Tab from July to August showed that knowledge-based inquiries accounted for the largest share at 39.0%, followed by entertainment (17.7%), local (14.4%), and shopping (9.3%). Among these, local queries—such as searching for restaurants or places to visit—and shopping queries—such as comparing products—accounted for 23.7% of the total. This means that roughly one in four questions is likely to lead to an actual action, such as a purchase, reservation, or visit.

This trend was also evident in a survey. In a survey conducted by Naver on August 20–21 among 415 users with experience using the AI tab, 53.5% reported that AI responses had led them to take actual actions such as making a purchase, booking, or visiting a location. Among those who took action, the most common activities were purchasing products (25.3%), visiting stores or locations (18.3%), making reservations at restaurants or cafes (17.6%), and booking accommodations (15.9%). Actual service data also showed that following shopping- or local-related questions, the click-through rates to Naver Shopping and Naver Maps/Places each exceeded 40%.

There were also differences in usage patterns by age group. Among those under 20, the proportion of “multi-turn” interactions—where users continued the conversation by asking the AI follow-up questions—was the highest. In contrast, those aged 40 and older were more likely to click on sources provided by the AI to check the original text, while multi-turn interactions were relatively less common. The shopping click-through rate was also highest among users aged 40 and older. While younger users narrowed down their search results through conversation with the AI, middle-aged and older users showed a pronounced tendency to verify sources and proceed to shopping, much like in traditional search.

Evidence also emerged of the AI integrating with Naver’s existing search ecosystem. Following responses from the AI tab, 68.7% of users navigated to information-based services such as blogs, cafes, news, and Knowledge iN, while 16.7% moved to action-oriented services like shopping, maps, and Places. This indicates that even when the AI summarizes answers, users continue to verify sources directly or utilize related services.

Naver assessed that this user behavior demonstrates the competitiveness of the AI Tab in connecting its own content with commerce and local services. The structure works by having AI synthesize actual user reviews and experience data accumulated on platforms like blogs and Cafes, then combining this with specific information—such as product prices or real-time booking availability—to facilitate purchases and reservations.

Professor Kim Kyung-woe of Yonsei University explained, “With typical generative AI services, the answer itself is often the final output, and users must navigate to other services to make a purchase or reservation.” He added, “The AI Tab has the advantage of providing both the qualitative information needed for decision-making and the specific information required for actual action.”

He went on to emphasize, “In the future, it will become increasingly difficult to differentiate a service simply by providing accurate and useful search results or answers,” adding, “Competitiveness in the search market will depend on how effectively a service supports users in translating information into actual decision-making and action.”