Co-developed with Amazon Web Services to overhaul merchandiser workflows

System cuts product selection and event preparation time by about 70%

Promotion prep, which consumed 40% of MD working hours, now largely automated

GS Shop merchandisers discuss promotional product selection and store layout using the AI-powered MD curation system co-developed with Amazon Web Services. [GS Retail]
GS Shop merchandisers discuss promotional product selection and store layout using the AI-powered MD curation system co-developed with Amazon Web Services. [GS Retail]

GS Shop has officially launched an AI-powered "MD curation system" co-developed with Amazon Web Services, the company said Tuesday, marking a push to transform how its merchandisers plan and manage promotions.

The system allows merchandisers to enter a promotion's intent and conditions, after which AI analyzes sales performance and customer behavior data to recommend suitable products and support the entire process from promotion planning to registration. The tool is designed to reduce repetitive operational tasks so merchandisers can focus on higher-value work such as product sourcing and event planning.

According to GS Shop, merchandisers responsible for product operations within its mobile app had been spending about 40 percent of their working hours on promotion preparation — a burden driven by the need to pull product information and sales data from multiple systems before manually selecting and arranging items.

The system runs on four AI agents — product curation, store auto-mapping, trend sensing and copywriting — that together cover everything from product selection to promotion setup and registration. When a mobile merchandiser sets a promotion period, planning intent and product scope, the product curation agent draws on sales and customer behavior data to extract up to 1,000 candidate products and recommend those that best fit the brief. Merchandisers can then refine the list by conversing with the AI in natural language — for example, asking it to "pick only products suited for moisturizing, hydration and skin-barrier care."

Once products are selected, the store auto-mapping agent assists with promotion layout and product placement. The trend sensing agent then analyzes internal and external data to suggest event themes, while the copywriting agent proposes event names and marketing copy, rounding out support for the full promotion planning workflow.

During a month-long pilot in October, about 40 percent of all promotions were built using the new system, and the time spent on product selection and event preparation fell by roughly 70 percent compared with before. Some 72 percent of mobile merchandisers used the system during the trial period. GS Shop plans to push the system's adoption rate above 90 percent by year-end.

Kim Yo-han, head of GS Retail's home shopping AI transformation division, said the project had strengthened the company's on-the-ground AI capabilities. "We will continue to spread an AI-driven work culture and translate it into tangible results," Kim said.


spa@heraldcorp.com