eCommerce Discovery AI Transformation: WCONCEPT’s AI-Driven Launch in Japan

Japan, which is currently taking a stride in the world of fashion, has to keep up with the latest innovations in fashion technology. WCONCEPT has done just that—they launched an AI-powered fashion discovery suite, similar to what European and North American retail giants have done in recent times. 

What is fashion discovery AI, and why do eCommerce store managers like you need it? In this article, we will look into the following:

• The state of the fashion industry in Japan

• Innovations in fashion AI

• Problems that fashion AI can solve and its benefits

 Let us start by understanding the current state of fashion in Japan and then discuss what WCONCEPT did to overcome the challenges.

The State of Fashion Discovery in Japan

The world, including Japan, is facing hard times in the fashion industry. Inflation is to blame, along with cyclical business slowdown.

Experts suggest that for a fashion retailer to thrive, it has to be localized, and brands must be positioned appropriately to the users.

Right now, most Japanese fashion retailers still use outdated eCommerce product pages, search and recommendation systems, and manual tagging and SEO approaches.

AI has changed all this—you now have a tool that can help you: 

• Localize your store through language and currency switchers

• Improve the accuracy of search results and recommendations

• Create a more interactive browsing experience—better than mundane and endless scrolling

• Automate repetitive tasks like product tagging, filtering, categorizing, etc.

 Is this something that your fashion retail business can benefit from? Certainly! In the succeeding sections, you will see the groundbreaking tech that WCONCEPT introduced to their store, which you can also implement so you can keep up with, and even stay ahead of, the changing times.

YesPlz Fashion AI Innovates How Fashion Stores Work in Japan

The fashion AI in WCONCEPT is our response to the emerging trends in fashion. It introduces new technology, with particular emphasis on fashion AI. The screenshot of the WCONCEPT homepage above may look ordinary, but it does have powerful yet inconspicuous features for eCommerce discovery.

What are these features? We’ll show you a quick overview and discuss them in detail later.

1. Virtual Mannequin

The Virtual Mannequin is an interactive filter, which allows the shopper to select specific areas of clothing and particular styles.

2. Hybrid AI Search

A hybrid search may not look like much at first sight, but it’s a sophisticated search engine powered by multi-modal AI, which significantly enhances the search process for both shoppers and eCommerce managers. 

3. Recommendations That Increase Awareness and Average Order Value

Generic recommendations rarely work, and shoppers often ignore them. Amazon and other giant stores have moved away from “typical recommendation” systems. Instead, they use AI to determine what products best suit the shoppers’ interests based on:

• Fashion styles 

• Shopping behavior 

 What works to increase average order value is personalization, and this feature can now be implemented in your fashion store with no labor-intensive processes.

4. Use of Multiple Languages and Currency Support for All Types of Shoppers

Languages and currencies would not be an issue if you only sell to one market. But what if you want to sell to different regions?

As you know by now, adding currency and language switchers can be challenging—you need a robust and accurate application that will translate your store’s content meaningfully.

All these features, and more, can be yours, but how do they work, and more importantly, why do you need them?

How YesPlz Fashion AI Discovery Suite Transforms Your Business

The main areas of priority to boost shopper engagement, and thereby sales, are interaction (or some degree of control), precision and relevance of search results, appropriate recommendations, and elimination of language barriers.

1. Virtual Mannequin Makes Product Browsing More Interactive

The virtual mannequin is a sidebar feature that functions as a filter. Instead of seeing words, the shopper has a real visualization of possible options to narrow her search.

Take a quick look at the clothes lined up on this page:

These clothes are random, but if the highlighted area is clicked (the selected Japanese character means “open”), you will notice how the thumbnails also changed.

All the clothes now have an open neckline. The shopper can also click on the virtual mannequin and select a specific area of the body. Here is a sample:

The shopper first clicks on the “LENGTH,” and when she does that, the options below also change. Now, these secondary options only present sub-filters for length, such as cropped, regular, and long.

She now chooses the first one, “CROPPED,” and you can see that the clothing thumbnails also changed. All the products shown now are cropped tops (see below).

What problems does the virtual mannequin solve, and how does it benefit eCommerce managers?

The two main problems in eCommerce filters are:

• Text blocks—difficult to read and use

• Engagement—shoppers do not have any meaningful or exciting interaction

 The YesPlz virtual mannequin, if implemented, prevents shoppers from getting overwhelmed with texts. In addition, the shopping experience goes through a massive change. The shopper now has more involvement, and that makes shopping all the more meaningful. 

• The virtual mannequin is not just eye candy, as it also reduces repetitious work

• You can customize the mannequin once and it will do the rest

• The fashion AI auto-tags product characteristics on its own; you do not have to “label” each product if it is cropped, long, short, etc.

• The AI implements faceted filtering; will only show products that meet all the filters used by the shopper

For the eCommerce manager, the virtual mannequin’s benefits are:

• Reduced repetitious workload in tagging the clothes one by one; the AI does the tagging on its own after it has been trained

• Accuracy of image tagging, the result of which is relevant results or thumbnails once a shopper clicks on a filter

• No additional code for faceted filtering; it takes only one time to customize the mannequin and sub-filters

• Boost in sales; a feature like this can increase cart size by 1.7x.

• Fast time-to-market; this AI has seamless and quick integration that can be completed in a few weeks

For the shopper, the benefit of the virtual mannequin is an effortless and seamless experience. She no longer has to read a wall of text to choose a filter, much less rely on a drop-down menu to narrow her preferences.

2. Hybrid AI Search Delivers Precise Product Showcase for Faster Shopping Experience

The screenshot above shows the search term “long sleeves blouse.” You will notice that all the thumbnails show blouses that have long sleeves. 

What you may not have seen is that no product title has the complete keyword “long sleeves blouse.” So, how did the store show these results?

Something like this is made possible with a hybrid AI fashion search engine. It improves search accuracy, making search results effective and product management efficient. 

The two main problems when it comes to search are:

• Incorrect search results – most internal search engines in online stores show unrelated products because they have no contextual understanding. This can also result in “no results” at all, which ends in shoppers leaving.

• Manual work—conventional search engines rely on text tags. If specific tags are not in the product pages, they will never appear for those search terms.

So why use AI? How does it solve the problem? Here are the main features of a hybrid AI search for eCommerce discovery:

• Image embedding—the AI knows the characteristics of clothing based on an image. So, if your product is long-sleeved but you never used that in the tags, the AI will still show that product to the shopper.

• Recognition of Trends—the AI can be trained to understand emerging fashion trends like Y2K, Barbiecore, etc. You do not have to go back to your product pages and tag each manually. The AI will know and show “Y2K” or “Barbiecore” clothing if the shopper used these terms.

• Auto-Tagging—the AI automatically tags products based on their properties or characteristics, such as the occasion, type, color, gender, vibe, etc. The AI will also update these tags.

• Multi-language and multi-currency—many consumers speak one language only and are only comfortable buying in their local currency. YesPlz supports both features for several languages like English, Korean, Japanese, with more language and currency options to come. 

The most important feature of a hybrid AI search is contextual fashion understanding. Each AI is built differently, so not all of them “understand” fashion terminologies, trends, and styles.

Just to give you a glimpse, here is a sample of how the AI understands an image and tags it accordingly:

YesPlz hybrid AI search is built specifically to cater to fashion retailers. Once it reads a phrase, like “long sleeves blouse,” it knows that the shopper is looking for a blouse, and not only a blouse—it must have long sleeves. 

Because of this, the AI will only show products that meet these conditions, thus eliminating shopper frustration from irrelevant search results.

3. Recommendations in Shopping Cart and Product Detail Page Are Targeted and Personalized

The image above is found at the bottom of a product page. The specific product the shopper is looking at is the “String Point Knit Tank Top.” 

The recommendations, as you may have observed, are all clothing items that are knitted. Why is this feature a necessity?

The two most common problems in any recommendation system are:

• Irrelevant recommendation—the products recommended are not suited for the shopper; they are mostly random and impersonal.

• No collaborative filtering—the recommendation system does not analyze user data and crowd behavior. Because of this, the recommendations do not sync with the shopper’s vibe or interests.

An AI-powered fashion recommendation system is essential for eCommerce discovery. At the outset, one may say that it just recommends clothing tagged the same way, but it actually works more than that. A fashion AI does these things:

• Understand crowd behavior—the AI gathers data based on shopper behavior and analyzes the data. It knows what other shoppers bought or added to their carts. A shopper who displays the same behavior as other shoppers in the past will get recommended products that previous customers bought.

• “You Might Also Like”—this recommendation is not random, and the eCommerce manager does not have to choose which product to show as a recommendation. The AI analyzes what the shopper is looking for, like style attributes, and will make recommendations based on this behavior.

• Brand Recommendations – if you want to showcase a particular brand, you can train the AI to prioritize specific items for that brand. This is useful if you see a fashion trend where the brand name is the primary buying factor of the shopper.

How does this help your fashion store? For one, shoppers do not get irrelevant recommendations. If a shopper wants to buy a knitted shirt, the AI will show other knitted shirts, not products like brown shoes.

Since the AI also does collaborative filtering, it can suggest products that people of a similar buying behavior already bought in the past, which makes conversion more likely.

4. Use of Multiple Languages and Currency Support for All Types of Shoppers

You can scale your market by making your store accessible at a global level. The problem is the language barrier, along with currency standards.

One big challenge is customizing the search engine to work for a different language. Each language has its own system, and what we did for WCONCEPT was to tailor the search engine to understand Japanese and therefore recognize Japanese characters, context, and find the matching items. 

For more clarity, one should know that Japanese and Korean are not based on the same alphabet or characters. Because of that, extra work is required for a website to support it. 

YesPlz fashion AI can power international fashion brands—it can support multiple languages and currencies. 

As an eCommerce manager, you probably know how challenging it is to find applications that can support multiple languages and currencies. YesPlz Hybrid AI search is the solution. It is built to support not just direct translation but to really understand linguistic nuances and contexts.  

With currency and language optimized locally, you can remove barriers that prevent shoppers from browsing your site. The currency switcher also makes the shopper feel more confident in making a buying decision.  

Summary

Fashion, if anything, is one of those industries that will always keep up with emerging trends and technologies. eCommerce managers have much to benefit from a fashion AI, such as:

• Enhancing a shopper’s browsing experience 

• Improving search results

• Personalizing product recommendations  

• Localizing an international store through multi-language and currency switchers

If you like what you see in WCONCEPT, be assured that you can implement the same in your online retail store. Contact us for a FREE demo to get started! We’ll listen to your concerns and train fashion AI that specifically solves your business challenges!