Styling Recommendtions
In the e-commerce industry, recommendation using AI has already become commonplace. After the COVID-19 pandemic, many companies have pursued digitalization, but there have been many cases where they have not seen tangible results directly linked to sales. In such situations, AI-based product recommendations are being evaluated as the most effective AI product that leads to sales.
In particular, recommending styling in the fashion industry can be seen as a core AI with a high contribution to sales in terms of providing new experiences and insights.
OMNICOMMERCE Styling recommendations recommend magical style combinations, learned from insights of thousands of influencers and fashion experts. OMNICOMMERCE Styling recommendations provide customers with optimal combinations using AI-based Recommendations from fashion items sold on the site. Through this, customers can stay on the site longer while viewing recommended style sets, and have a more satisfying shopping experience.
What Problems It Solves
- Styling recommendations vary depending on the product characteristics of the e-commerce platform that provides the service.
- Can we recommend optimal styling using the products sold on our e-commerce platform?
- Is it possible to create styling that combines core items such as shoes and bags in addition to clothing?
- Is the recommended styling attractive enough for customers to want to see more?
- Is it possible to provide various types of styling recommendations?
- Various seasonal styling recommendations
- Styling recommendations based on personal preferences
- Styling recommendations based on age
- Styling recommendations considering TPO (Time, Place, Occasion)
- Difficulty in providing styling that matches rapidly changing seasons and trends
- As the standard for styling changes according to seasons and trends, there are difficulties in providing immediate styling recommendations.
- For example, there are color changes in fashion items based on representative colors every year, and we want to reflect product recommendations based on this and provide an improved recommendation service that is connected to styling recommendations that match it.
- There are limitations to guiding and informing buyers of a diverse range of products.
- In the case of clothing, most of the recommendations are within the same outfit.
- There are not many sets that recommend multiple outfits together, such as coordinating recommendations or proposing coordinated outfits.
Value to Customers
- Customers can receive tailored recommendations based on the style and characteristics of the e-commerce platform.
- Styling recommendations can reflect seasonal and trend changes, allowing us to provide services to users that keep up with the constantly changing fashion environment.
- The recommended products can be selected and organized according to brand, price range, and image type, making it possible to create characteristic recommendation sets centered around the e-commerce platform and provide services.
- By recommending entire styling sets, including clothing, bags, shoes, jewelry, and accessories, we can expand the range of products that users can select on a single service screen.
Key Features
- Ensuring diversity in styling recommendation sets
- We provide recommendation sets for clothing, shoes, bags, hats, jewelry, and other items necessary for fashion styling.
We provide up to 5 styling sets for each product.
We provide AI recommendation solutions tailored to the product configuration and characteristics of the site.
- Customized recommendations are provided based on image composition and style.
How to Use Styling Recommendations
To use the Styling Recommendations solution, you need a Workspace that has the Styling Recommendations solution enabled. For information on how to enable the Styling Recommendations solution in a Workspace, please refer to the ↗️ Connecting Solutions to Workspace page.
Styling Recommendation API
You can automate the process of registering products in a Workspace and receiving recommendation results for registered products via API. For detailed information on each step, please refer to the following API documentation:
- Register products through the Management API: ↗️ Management API documentation
- Obtaining recommendation results using the Recommendation API: ↗️ Styling Recommendation API documentation
Before using the Styling Recommendations, the following prerequisites must be set up.
Gender Setting
- Gender information for products is required to use the Styling Recommendation solution.
- If it is difficult to send gender information, we recommend using the Omnious Image Moderation solution. The Image Moderation solution recognizes the gender of products by recognizing the product image.
- When requesting management API information for products, the format of the gender information in the metadata is as follows:
- Gender information format
Gender setting Recommendation pool male Recommend within men's products female Recommend within women's products male, female Recommend within all products unisex Recommend within unisex products
- Gender information format
- Example) Styling recommendation for men's products
Recommendation Category Setting
When using Styling Recommendation, you need to set the range of the category for recommendations.
Currently, Styling Recommendation provides recommendations for clothing, shoes, bags, and accessories.
- Example) Recommendation Category Configuration
Styling Recommendations via OMNICOMMERCE Web Service
You can only view Styling Recommendations if you use the Styling Recommendations solution in the Workspace.
You can view up to 5 outfit sets that match the current product the best out of all the products connected to the Workspace.
(↗️ Styling Recommendations Tab)