The end of search, the zero click era…
The era is shifting to one where artificial intelligence (AI) shops instead of consumers. In the past, if consumers searched for "a dress that's good to wear on a summer vacation" and purchased by picking through countless search results one by one, now they ask in conversational form, "Recommend a light, airy dress that's good to wear on a summer vacation. The color is white. The length should fall below the knee and it would be good if it has sleeves. The price is 40,000 won or less," and get introduced to just a few selected items.
At first, retailers felt confusion and fear at the fact that AI, not people, would be more actively involved in shopping, but they are now adapting proactively. They are rolling out conversational chatbots and actively investing in generative engine optimization (GEO).
According to the retail industry on the 12th, retailers are moving quickly to align with the era of Agentic Commerce. Agentic Commerce is a next-generation shopping model in which an AI agent autonomously handles the entire process on behalf of consumers, from product search to comparison and recommendation, and further to payment and order execution.
In the Agentic Commerce era, generative engine optimization is essential. This means tuning the environment so a generative engine can select the right products for consumers. According to Deloitte Touche Tohmatsu Limited (DTTL)'s Global Retail Industry Trends and Outlook, recommendation traffic coming from AI-based conversational services, including ChatGPT, already accounts for 15% to 20% of total inflows at some retail corporations.
To adapt to this change, retailers are first introducing conversational chatbots. A representative case is Amorepacific. Amorepacific rolled out Amore Chat in May last year as an AI chatbot for consumer counseling. CJ Olive Young also plans to introduce an AI-based conversational search feature in its app at the end of this year.
This is not a move unique to the beauty sector. Lotte Himart also began an open beta test of its AI shopping agent HAVI on the 2nd. When consumers ask questions as if talking to a person, it finds the product they want and supports comparison and recommendation.
In the retail industry, building optimization consulting for GEO work and the corresponding work environment is seen as important. Some say this takes three to four times more effort than building an AI chatbot.
According to multiple GEO optimization consultancies, to raise an AI chatbot's curation ability, product information tailored to requirements is essential so a generative engine can find it well. Key elements include clear, top-down descriptions, structuring information in tables or lists, and highly reliable sources. The most important factor is that concrete figures and data, rather than abstract language, must be included.
A fashion industry official said, "Rather than the description of a light, airy dress, additional modifiers such as chiffon fabric, linen fabric, or poly-nylon blend should be attached as associations or sub-concepts of 'light, airy,'" adding, "Beyond that, you need to compare pros and cons clearly and explain the recommendation context and grounds together so AI can introduce products well to consumers."
The results of GEO optimization consulting are good. Amorepacific carried out consulting last year and changed its system, resulting in Osulloc Mall's generative AI-based inbound traffic in the fourth quarter of last year increasing 602% from a year earlier. During the same period, the purchase conversion rate rose 725%. Innisfree Mall's inbound traffic also increased by about 366%.
Retailers expect the nature of hired roles to change during the GEO optimization process that will take place for the time being. To recommend products optimized for consumers, they must change the existing method of entering product information, and the interface for this also needs to change. Retraining personnel is also important.
A beauty industry official said, "There is a mood of making large-scale investments out of concern that, if we respond passively and AI transition is delayed, we could lose competitiveness going forward."