JapanConsumer goods2026 – ongoingKao
Kao-MIKOMIL AI agent for cosmetics demand planning
Kao put a demand-planning AI agent to work on its KATE cosmetics brand in July 2026, aiming to cut cosmetics inventory about 25% by 2027
Problem
Planning demand for cosmetics took data scientists and brand staff working together to read sales of similar products and past promotions. New products with no sales history were harder still. Kao wanted fewer stock-outs and less excess stock.
Approach
Kao developed its own demand forecasting model and added an AI agent on top. The agent takes the forecast, layers in promotion plans, sales channels and prices, and presents a draft demand plan with its reasoning. Business-unit staff review it, coordinate across departments and make the final call. The agent went live for KATE in July 2026, with other cosmetics brands to follow.
Results
Takeaway
The 25% figure is a stated goal, and Kao has not published results from the first brand. Final planning decisions stay with staff.
Categories
- Size
- Large enterprise
- Industry
- Consumer goods
- Business function
- Logistics and supply chain
- Innovation type
- Process innovation
- AI technology
- AI agentsPredictive analytics and ML
- Adoption
- Built in-house
- Evidence
- Company's own claim
Sources
Source summaries
Kao developed an AI agent that combines its in-house forecast with promotion, channel and price information to propose demand plans with reasons. It says use began in July 2026 for KATE and will spread to other cosmetics brands. Kao aims to reduce cosmetics inventory by about 25% by 2027 against 2024.
Advertimes reports that the agent proposes demand plans while staff keep the final decision. Kao already builds production plans on monthly forecasts about 18 months out and daily forecasts about six months out.
Full details are in the original sources.
