JapanRetail and e-commerce2023 – ongoingSeven-Eleven Japan
Generative AI for new product planning
Seven-Eleven Japan built a generative AI planning system and cut meeting time by 80% in its first department
Problem
Developing a new private-brand product at Japan's largest convenience store chain could take up to about ten months. Planners had to read sales data, supplier information and customer reactions on social media before drafting proposals.
Approach
In 2023 the company built a cloud system that combines generative AI from OpenAI, Stability AI and Google to analyse store sales, manufacturer data and social media and to draft product proposals with text and images. It first ran in one department, and from spring 2024 it was to be used for product planning, with resulting products expected under the Seven Premium brand.
Results
Takeaway
The first department cut meeting time by 80% after using AI to draft product proposals. The company expected planning time to fall by up to 90% at launch; NACS and Retail Optimiser did not report a follow-up result.
Categories
- Size
- Large enterprise
- Industry
- Retail and e-commerce
- Business function
- Marketing and salesInternal work and knowledge management
- Innovation type
- Process innovation
- AI technology
- Generative AI
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Company's own claim
Sources
Source summaries
Citing Nikkei Asia, the US convenience-store association's news service reports that Seven-Eleven Japan built a cloud system using AI from OpenAI, Stability AI and Google to spot trends and draft product proposals. It says one department already using it cut internal meetings by 80%, and that the company expects development time to fall by up to 90%, for example from ten months to one.
Retail Optimiser reports the planned spring 2024 start of AI-drafted product text and images based on sales, manufacturer and social media data. It repeats the 80% cut in meeting time in the first department and the expectation of up to 90% shorter planning. It also says almost 9,000 employees had begun using the technology.
Full details are in the original sources.
