Over Half of Japanese Consumers Make Purchases Based on AI Recommendations, Generative AI Penetrates User Buying Decisions

[Ebrun Original] On June 17th, Japanese digital marketing service provider PLAN-B released a research report on the impact of generative AI on consumer decision-making. The report focuses on users who use generative AI in their daily lives, aiming to reveal the extent to which AI intervenes in personal decisions such as purchases and destination choices, as well as the latest behavioral changes.

The survey points out that high-frequency users who use generative AI for queries several times a week or more account for 26.8% of all respondents, an increase of 9.7 percentage points compared to last year. Despite the clear growth trend, this proportion still falls short of 30%, indicating that the high-frequency user group is still gradually expanding.

Meanwhile, purchases and destination decisions initiated by generative AI conversations have already covered more than half of the respondents, with 54.1% giving an affirmative answer. This shows that purchasing behaviors originating from generative AI are further penetrating.

Looking at the categories of goods or services that respondents have actually purchased through generative AI conversations, computers, smartphones, and related peripherals rank first at 40.9%, followed by clothing and apparel (including fashion accessories) at 36.0%, and food and beverages (including health supplements) at 34.9%.

Cosmetics, skincare, and haircare products also reached 30.7%, consistent with last year, maintaining a relatively high level. Additionally, home appliances accounted for 25.4%, and financial products such as credit cards and securities accounted for 20.8%. For the latter category, which requires detailed comparison and analysis among multiple conditions, over 20% of users have completed purchases with the help of generative AI conversations.

It is worth noting that most users conduct multi-source information verification before making a final decision. 87.4% of respondents use Google search to additionally verify AI's responses, consistent with last year's behavior pattern.

Another 34.4% use e-commerce platforms like Amazon for confirmation, and 34.1% choose social media like Instagram for verification. The proportion answering 'did not verify information outside of generative AI' was only 2.7%, an extremely low figure that has hardly changed compared to last year's survey results.

Based on this, PLAN-B's research concludes that the current Japanese market is in a new stage of information gathering and purchasing behavior formation. The typical path is summarized as: consumers discover candidate options through generative AI, then repeatedly verify and compare them across search engines, social media, and e-commerce websites, before finally making a decision. This back-and-forth, multi-layered confirmation process is becoming a distinctive consumer decision-making model.

The report also conducted a questionnaire survey on the generative AI conversational methods that lead to purchases or destination decisions.

The results show that when users are exploring products, pain points are often the primary driver.

Vague 'needs' or 'troubles' often become the starting point of the conversation. Users tend to describe their concerns in natural language and gradually clarify their choice direction through AI's responses and follow-up questions. The report also excerpts some dialogue examples between users and generative AI from the responses, further illustrating this process from vague intention to specific decision.

Ebrun will continue to track and report on this information. To learn more related information, please scan the QR code to follow the author's WeChat.