One of the biggest issues that vertical scrolling comics portals have faced worldwide is how to effectively curate their services for every individual user. With massive catalogs and thousands of titles across a multitude of genres, no one service can effectively serve every user with new titles based on their previous reading history without some sort of bottleneck or heavy bias.
Now, it looks like Kakao Page and Naver WEBTOON are separately addressing the curation problem in Korea by actively rolling out new curation engines powered by artificial intelligence on their webnovel and comics portals, after months of internal testing and development. Regardless of personal feelings on AI regarding comics and the valid issues they do raise, it looks like AI is actually providing a huge help for both services in terms of surfacing the right titles for each user.
Kakao Page’s introduction of ‘Helix Curation’ marks a significant milestone in this journey. This innovative approach, powered by the AI brand Helix, was first unveiled on April 29 and has been progressively rolled out to all users as of August 2nd.
Helix Curation stands out as a sophisticated method that leverages artificial intelligence to analyze individual users’ purchase histories and interests. By doing so, it curates a personalized selection of works and prominently displays them on the first screen of Kakao Page. This means that each user is greeted with a unique home screen, tailored to their preferences, a move that signifies a departure from the one-size-fits-all model.
The impact of this change extends beyond mere aesthetics. The effectiveness of AI curation as a tool for engagement has been substantiated through a three-month pilot operation. The data speaks volumes: post-implementation of ‘Helix Curation’, the click-through rate (CTR) for the recommended tab on Kakao Page’s first page soared by an impressive 96%. This spike in engagement didn’t just stop there; the webtoon and web novel tabs witnessed a 42% and 138% increase in CTR, respectively.
These statistics underscore the potential of AI-driven curation in revolutionizing how content is discovered and consumed. It’s not just about changing the home screen; it’s about creating a more intuitive and satisfying user experience that can lead to increased content consumption and, potentially, revenue growth for content providers.
Naver Webtoon’s ‘Know!’ feature is another prime example of this evolution. Since its official implementation on June 26th, ‘Know!’ has revolutionized the way readers discover and engage with scrolling comics.
Unlike traditional recommendation systems that relied on popularity metrics or recent updates, ‘Know!’ employs a sophisticated AI curation model developed by Webtoon AI. This model personalizes the user experience by sorting content based on the individual’s predicted preferences. The most appealing webtoons are displayed at the forefront, while less relevant ones are relegated to the bottom, ensuring a tailored browsing experience for each user.
This shift from generic sorting methods to a personalized curation approach has yielded impressive results. Users are not only visiting the Naver Webtoon app more frequently but are also exploring and reading a broader range of content. The ‘Know!’ feature has led to a substantial increase in the discovery of previously unread works, as reported by Naver Webtoon. Moreover, new users are finding themselves immersed in a diverse array of webtoons, including fresh releases, thanks to the AI’s nuanced understanding of their tastes.
The success of ‘Know!’ highlights the potential of AI in enhancing content discoverability and user engagement. By moving beyond mere popularity rankings, Naver Webtoon is paving the way for a more personalized and satisfying reading experience for its millions of users. Whether these same tools will be implemented overseas remains to be seen, but the fact that they’ve been so successful in their home market means its a matter of time before some sort of implementation is rolled out on the international portals.
