[DigitalToday reporter Hojeong Lee (이호정)] Naver and Kakao are moving competitively to secure everyday data to raise the level of personalisation in artificial intelligence (AI) services. Naver’s strategy is to collect users’ visit, order and payment data through offline payment terminals and link it with existing search and booking information. Kakao, by contrast, appears focused on long-term memory of schedules, interests and user preferences scattered across KakaoTalk conversations.

As AI develops into an agent that goes beyond information searches or answering questions to make bookings, orders and payments on behalf of users, the importance of data is also growing. To propose appropriate services and actions, it is necessary to understand not only what a user searched for, but also what choices followed, where they actually paid, and what intentions were revealed in conversations.

Naver captures offline behaviour, linking from search to repeat visits

Naver is using its offline integrated terminal, 'N Pay Connect', as a key touchpoint for securing everyday data. It judged that the offline commerce payments market, which is more than 3 times larger than online, has not yet been sufficiently digitised and its data is not structured and remains separated from online. It decided it needed to preempt the market and brought forward a rollout schedule it had initially planned to 추진 in stages through next year.

The company judges that as AI evolves toward directly carrying out bookings, orders and payments beyond recommendations, offline behaviour data such as store visits and order histories becomes an essential asset. To that end, it plans to combine Place search and booking information, Naver IDs, and payment and membership data with store data from N Pay Connect. It aims to link the process from search to booking, ordering, payment and repeat visits under a single ID.

N Pay Connect will expand beyond payment terminals into an AI platform for offline merchants. It plans to broaden monetisation areas, including merchant solutions such as trade-area analysis and coupon and promotion management, as well as loan and insurance brokerage and ad products based on visit and purchase conversion data. Naver presented a goal of becoming South Korea's No. 1 operator next year by combined online and offline B2C payment volume, and of expanding ID-based offline payment volume to 130 trillion won within 5 years.

Second-quarter N Pay payment volume was 25.2 trillion won, up 21 percent from a year earlier.

As follow-on monetisation in areas such as advertising and finance can expand only after merchant and transaction data accumulate, the pace of N Pay Connect adoption is likely to be a key variable for future monetisation.

Kakao reads intent in conversations, personalising with long-term memory

Kakao is making KakaoTalk conversations themselves the core foundation for AI personalisation. Its aim is to link schedule and interest information scattered across multiple chats through long-term memory of on-device AI, comprehensively understand users' situations and provide more personalised services as usage lengthens. The lightweight model it recently disclosed, 'Kanana2 1.3B', showed better performance than global open-source models in areas such as conversation, knowledge and instruction-following despite a sharp reduction in parameter size, the company said.

Kakao sees the internet and mobile era as an advertising-centred 'attention economy', and expects an agent AI era to shift to an 'intention economy' that understands user intent and connects it to action. It chose food delivery as its first external integration area and announced a partnership with Coupang Eats. Once the integration is completed, when delivery intent is detected in KakaoTalk conversations it will suggest menus based on accumulated preferences and complete ordering and payment within the chatroom.

It plans to expand agents to commerce, bookings, travel and payments, and also to use the open platform 'PlayMCP', where more than 3,400 tools are called more than 100,000 times a day. The share of positive evaluations for 편톡 and answers in Kanana in KakaoTalk each topped 80 percent and 90 percent, respectively, and retention rose from about 70 percent early after launch to near 80 percent. Kakao expects monthly active users (MAU) of AI services within KakaoTalk to exceed 10 million by year-end.

Even its first vertical, food delivery, is still at the stage of a partnership announcement and the actual integrated service has not started. As improving user metrics and converting into actual transactions are separate issues, when delivery integration is realised and how quickly follow-on partnerships are added are expected to determine the timing of monetisation.

Naver focuses on behaviour, Kakao on intent; data linkage will decide monetisation

Naver has search, shopping, Place and payments, making it advantageous to connect a user's exploration to post-purchase behaviour. N Pay Connect adds offline in-store behaviour that was previously difficult to confirm. Kakao has strength in first reading intent before purchase or booking through KakaoTalk, an everyday conversation space, and can convert that intent into orders and payments once external operators' transaction functions are connected.

Their monetisation methods also differ. Naver plans to secure offline merchant touchpoints directly and expand its business through payment fees, advertising, solutions and financial brokerage. Kakao plans to bring external partners' transaction functions into KakaoTalk and expand revenue sources through transaction fees, subscriptions and AI advertising.

Both companies share the same point that connectivity, not total data volume, is the key. Naver must tie searches and bookings with store visits and payments under a single ID, and Kakao must link intent detected in conversations with ordering and payment processes on external services. If data is disconnected, it is difficult for AI recommendations to lead to actual transactions, making it likely that future gaps will be determined not by total data volume but by the completeness of linkage.

The key will be metrics. For Naver, the number of N Pay Connect merchants and the pace of growth in offline payment volume, and for Kakao, the pace of expanding partnerships into verticals beyond delivery, are expected to be the next indicators of whether each company's data strategy converts into actual revenue.