Search | AI HUB

The COVID-19 pandemic has accelerated the digital transformation of small and medium-sized enterprises (SMEs). However, the first step in this transformation is to create beautiful product designs and quickly list products online. Using AI image recognition technology, Choice Technology has identified 'explosive' e-commerce designs, creating an AI workstation for retail e-commerce that features automatic object detection, background removal, and beautification. For small e-commerce businesses with about 500-800 product items, this can save approximately 90% of the time. It is an excellent tool for small and medium-sized retailers making the transition from offline (physical channels) to online (virtual channels) sales.

The meaning behind 'Choice' is 'Choosing the best tech experience for our clients.' Liu Yi-Han, the founder and CEO of Choice Technology, hopes to leverage his expertise in image engineering to help SMEs with technology barriers achieve their dream of easily listing products on e-commerce platforms without needing to learn a multitude of tech tools.

▲Using AI image recognition technology, photos can be automatically background removed, saving time in photo enhancement.

▲E-commerce AI Workstation, from material design to e-commerce listing, get it done with one 'click,' fast and convenient.

Machine learning automatically generates 50 types of product recommendations for direct listing on e-commerce platforms with one click.

Since its foundation in May 2018, Choice Technology has used machine learning algorithms to collect millions of product designs on social media platforms like Instagram and the largest handmade marketplace Etsy, automatically generating 500,000 design recommendations. Clients simply need to upload their product photos, and the AI tools at the e-commerce workstation perform background removal, background addition, and other image editing tasks. Unlike the traditional method where retailers had to hire professional photographers and designers to prepare products for e-commerce, this approach saves time, money, and effort.

According to statistics, the average cost for photographing a product, arranging its layout, and designing it varies from 2,000 to 3,000 RMB per item. For a product range of 1,000 items, the cost in money and time can be overwhelming for small and medium-sized e-commerce businesses. Choice Technology draws product types from the sales rankings of major e-commerce platforms like Amazon and Shopify, selecting categories such as fashion apparel, catering food, home accessories, fresh fruits and vegetables, sports equipment, and nutrition & health fashion, among others, with the largest category being fashion apparel, which accounts for up to 56% of sales on the Choice platform.

Furthermore, during the pandemic, consumers mostly opted for takeout or delivery services such as UberEats and Foodpanda, which has led to a surge in catering food, also becoming an important design recommendation on the platform. With over 500,000 photos used to train the AI model, the strongest recognition capabilities are in home and apparel categories.

▲Choice Technology team, photo (second from the right) shows founder and CEO Liu Yi-Han.

Liu Yi-Han pointed out that after the pandemic, WFH (Work From Home) has triggered another wave of sales for delivery meals and home accessories. In the future, the database will be adjusted dynamically based on the sales ranking movement of e-commerce platforms in terms of product images and attractive background scenarios.

Helping SMEs transform into e-commerce by offering the first month of rapid listing services for free

The operating model of Choice Technology is based on a SaaS B2B model, charging based on the number of photo uploads, billed monthly, and catering to individual retailers, small to medium-sized customers, and corporate clients. Since the pandemic, offline retail opportunities have been keen to transition to online sales. The Ministry of Economic Affairs Industrial Bureau offers digital transformation schemes for SMEs by providing subsidies to help retailers transform. Currently, Choice Technology provides the first month free of rapid listing services to small and medium-sized retailers using online tools like Google Forms for group orders, eventually integrating with e-commerce platforms such as Shopee and Shopify to save listing time, allowing retailers to focus on product quality, beautification, and marketing channel work.

Regarding the distribution of customers, Liu Yi-Han analyzed that Choice's clientele is divided into two types. The first type are physical retail businesses new to e-commerce, who primarily need fast background removal and automatic lighting adjustments for their product photos, often preferring pure white or simple backgrounds. The second type is medium to large enterprise brands, who require high-quality product photos with personalized designs suited for different festivals, such as pink items for Valentine's Day, which realistically enhance the shelf presentation and indeed have the potential to become 'explosive products.'

'The technology for optimizing product photos is not the issue; the challenge is creating suitable and high-conversion scenario photos.' Liu Yi-Han stated that data collection and training are automated processes, and AI technology is quite mature. The difficulty lies in efficiently collecting data with attractive, high-conversion scenarios for machine learning, requiring continuous dynamic extraction from the sales rankings of major e-commerce platforms.

Regarding the future business layout of Choice, Liu Yi-Han explained that the short-term goal is to assist physical stores with digital transformation, since many SMEs were severely affected by the COVID-19 pandemic, and moving sales online can help reduce the impact of the pandemic. As for the mid to long-term goals, affected by the severe pandemic situation in overseas markets, Choice will first serve Taiwan customers well. After the pandemic eases further, they will expand into the US and Southeast Asia markets.

▲Liu Yi-Han attended the 2020 MarTech Marketing Forum for a panel discussion.

「Translated content is generated by ChatGPT and is for reference only. Translation date：2024-05-19」

Internet Explorer

Mozilla Firefox

Google Chrome

Apple Safari (5.0)