Naver has published its first proper look at how people use AI Tab, and the headline isn’t about knowledge questions at all. In an AI Tab service brief dated 9 November, produced with Professor Kim Kyung-oe’s team at Yonsei University, Naver says almost one in four AI Tab queries is the kind that can end in a booking, a purchase or a visit. Its own survey says more than half of users have already done exactly that.
For anyone doing SEO in Korea, that changes the question from “will Naver’s AI answer my query?” to “will Naver’s AI send the user to my Place listing, my Smart Store or my competitor’s?”
What the brief says
AI Tab is Naver’s conversational search product. It went into beta in April and opened fully on 25 June. Naver describes it as building answers from experience content in Café and Blog, then linking to Naver Shopping and Map/Place so that a question can turn into a purchase or reservation. The brief is based on real usage data from July and August 2026. The key numbers:
- 23.7% of AI Tab queries were “action-conversion” queries about places or products. Local queries, such as restaurants and places to go, made up 14.4%, and shopping queries that find or compare products made up 9.3%.
- In Naver’s own survey, 53.5% of AI Tab users said an answer had led them to an actual purchase, booking or visit.
- In the usage data, click-through from shopping and local queries to Naver Shopping and Map/Place was above 40% for each.
- Users under 20 tended to keep asking the AI follow-up questions, while users aged 40 and over were more likely to open the cited sources and read the original. Shopping click-through was highest in the 40+ group.
After reading an answer, users opened the Blog, Café and KnowledgeiN posts cited as sources, or went on to Shopping and Place. Naver calls this the starting point for “agentic search”.
Read it as Naver’s pitch, not an audit
Two caveats are worth keeping in mind. The 53.5% is a self-reported survey figure, and the brief doesn’t say how many users were asked. The 40% click-through applies only to shopping and local queries, not to AI Tab as a whole. It is also Naver measuring the success of its own product at a moment when it is selling more ad inventory around AI surfaces, including ADVoost ads inside AI Briefing since July. Still, the direction matches everything else Naver has done this autumn.
Last week Naver launched Experience UGC blocks that group Blog and Café reviews by interest. Before that it began paying Clip creators when its AI cites them. AI Tab is the same model with a checkout attached. First-hand content goes in, and a Naver-owned transaction comes out.
What it means for Korean SEO
The most important number for brands is the 40% click-through into Shopping and Place. When AI Tab handles a commercial or local query, the next click mostly stays inside Naver’s verticals, not on a brand’s own website. So in practice:
- Treat Place and Smart Store as AI landing pages. Opening hours, prices, menus, photos and reviews are what the AI will hand the user to next. Out-of-date Place data now costs you conversions inside the answer as well as in Maps.
- First-hand Blog and Café content is the input. AI Tab builds its answers from experience content. Detailed real reviews give it something to cite, and thin promotional posts don’t.
- Segment by age. If your buyers are over 40, they are the users most likely to open sources and click through to Shopping, so the cited posts and the product page both need to hold up when someone actually reads them.
Naver still doesn’t give site owners any AI Tab or AI Briefing exposure report, so you can’t see these numbers for your own brand. Until it does, track a fixed set of commercial and local queries in AI Tab by hand, and record which sources and verticals it links to.