Back to blog
AI Marketing

KakaoTalk Goes Agentic: Search, Book, and Buy in One App

KakaoTalk Goes Agentic: Search, Book, and Buy in One App

Kakao is accelerating an agentic AI platform inside KakaoTalk that can search, book, and buy. As a chat app turns into an app that acts, the touchpoints of commerce and marketing may be redrawn.

KakaoTools launches with 21 services

KakaoTools, part of ChatGPT for Kakao, currently integrates 21 services, including Musinsa, Olive Young, Zigbang, Hana Tour, and OP.GG. Kakao plans to expand third-party integrations further in the second half of the year, letting users complete real tasks — booking, lookups, purchases — within the flow of a KakaoTalk conversation.

Underpinning this expansion is the MCP (Model Context Protocol) ecosystem. Servers registered on PlayMCP have surged to 461. The more servers that connect external APIs and data in a standardized way, the wider the range of tasks an agent can handle.

What marketers should watch

The key question is whether your brand's service is ready to be called by an agent. Going forward, the new conversion point is not just ranking high in search results but being the option an agent selects when it executes a booking or purchase.

Prepare in three ways: (1) expose product, inventory, and booking data consistently via APIs; (2) design structured product information and clear action flows (book/buy) that agents can parse; and (3) plan the brand experience at conversational touchpoints as its own scenario. To discuss channel strategy for AI commerce, explore Best Partner's services or contact our team.

Frequently Asked Questions

What can KakaoTalk's AI agent do?

It performs real tasks — search, booking, purchasing — inside KakaoTalk conversations. KakaoTools, part of ChatGPT for Kakao, integrates 21 services including Musinsa, Olive Young, Zigbang, Hana Tour, and OP.GG.

Why does the growth of MCP servers matter?

MCP is a protocol that connects external APIs and data in a standardized way. With PlayMCP registrations surging to 461 servers, the range of tasks an agent can handle expands — and so do the opportunities for brand services to be integrated.

How should brands prepare for the agentic AI era?

Expose product, inventory, and booking data consistently through APIs, design structured product information and clear book/buy action flows that agents can parse, and plan the brand experience at conversational touchpoints as a dedicated scenario.

Where does your own site stand?

To apply what you just read to your own site, start with a free audit of where things are now.

A strategist replies within 24 hours on business days.

Read next