Cainz Introduces AI Post Assistance | The Technology That Boosts Reviews While Hiding the 'AI Feel'
機械翻訳 / Machine-translated
A
AI Friends
6時間前・13分で読める
機械翻訳 / Machine-translated
@aifriends
AI Friends（https://aifriends.jp）のクロスポスト公式アカウント。AIツールの紹介・使い方・できることを、中学生でもわかるやさしい日本語で届けます。
What you'll learn from this article
Major home improvement retailer Cainz has implemented an AI post assistance feature on its community site "KuraShere" (暮らシェア). The site is a platform where users share product usage ideas and everyday life tips with one another.
The development was handled by System Support Co., based in Gifu Prefecture. Leveraging Google Cloud's Gemini API (an AI service provided by Google), they built a system that makes it easy for users to submit posts.
Users can start a post with as little as one sentence, and the AI suggests purposes such as "a family memory" or "an unexpected use." Users can also adjust the tone of the writing style, choosing from options like "formal" or "friendly."
What makes Cainz's post assistance AI distinctive is its design philosophy of "suppressing the feeling that AI wrote it." In other words, it avoids producing unnatural text by leaving everything entirely up to AI.
As of 2026, AI-generated text tends to fall into generic and vague expressions, lacking any sense of real personal experience or individual opinion. For example, product reviews on e-commerce sites can end up sounding like something anyone could have written — such as "This product is wonderful. I recommend it to many people."
When posts that are obviously "pure AI" increase, readers begin to wonder, "Did this person actually use it?" As a result, the reviews fail to earn trust and don't lead to purchases.
Cainz's AI post assistance uses a style where the user writes the first sentence in their own words, and the AI expands or refines it from there. In other words, the positioning is not "AI ghostwriting" but rather "AI supporting the user's own words."
In the e-commerce and retail industry, a major challenge has been users feeling "I want to post, but I don't know how to write it." Products purchased at home improvement stores in particular — think DIY projects or everyday life hacks — tend to require complex explanations.
Cainz's AI post assistance lowers this "posting barrier" in three ways:
Through this, the development team explains that posting has transformed from "a difficult task" into "something easy and enjoyable."
UGC stands for "User Generated Content" — reviews and posts created by users themselves, not by companies.
As of 2026, it is said that advances in AI have weeded out "false information," ushering in an era where only "genuine experiences" hold value. Purchasing high-ticket items or large pieces of furniture in particular carries a high barrier, and reviews from other users serve as an aid in making buying decisions.
Users tend to feel an affinity for comments from existing buyers who seem to share the same needs, making it easier to think, "If other people are satisfied, maybe I'll buy it too." In other words, UGC is a crucial factor that directly connects to sales.
However, increasing UGC comes with challenges. Without designing mechanisms that motivate users to post, the number of posts simply won't grow. Cainz's AI post assistance can be seen as one solution to this challenge.
Let's also look at the technical workings. Cainz's AI post assistance leverages Google Cloud's Gemini API.
The Gemini API is a "multimodal AI" (an AI capable of processing data in multiple formats) that handles not just text, but also images, audio, and video. As of 2026, its strength lies in providing a single API as the foundation for "accepting any input and producing any output."
System Support Co. implemented the following:
It is also important to use different models depending on the difficulty of the task. In practice, using the fast, low-cost "gemini-1.5-flash" for simple summarization or classification, and "gemini-1.5-pro" for complex reasoning or long-form text generation, is considered effective.
Through "KuraShere," Cainz also aims to narrow the gap between online and offline. By linking the in-store experience with the online community, the goal is to improve the overall customer purchasing experience.
For example, one could imagine users posting about a product they saw in-store right on the spot, or using other users' posts as a reference while searching for products in a store. This kind of "fusion of online and offline" is also a trend across the retail industry as a whole.
Cainz also operates a related DIY community, "Cainz DIY Square," which has surpassed 100,000 members. Through these multiple communities, the company is advancing a strategy to increase customer touchpoints and boost loyalty (attachment to the company and brand).
Cainz's AI post assistance is an initiative that other e-commerce operators and retailers can also learn from. Two points in particular are significant.
The first is the concept of "positioning AI not as a ghostwriting tool, but as a support tool." Rather than leaving everything to AI, by committing to the role of drawing out and refining the user's own words, natural and trustworthy posts are produced.
The second is "using data analysis to improve the UX." By accumulating post data in BigQuery and analyzing which post purposes tend to be chosen and which tones are preferred, the system can be continuously improved.
Going forward, "writing assistance that suppresses the AI feel" is likely to spread to other industries. For example, similar technology could be applied to travel reviews, restaurant reviews, social media posts, and more.
This article is a cross-post from AI Friends.
