Starbucks Uses ChatGPT to Turn Emotions and Images into Coffee Orders
Starbucks has launched a Beta application on the ChatGPT platform that allows users to receive AI-generated drink recommendations by describing their emotions or uploading photos that reflect the current "atmosphere." Users can then customize and place orders within the Starbucks App. The official statement indicates that the system is designed to align with Gen Z and younger users' decision-making process of "starting from feelings rather than from the menu," integrating AI as an "ordering inspiration layer" into the consumption process.
After enabling the Starbucks mini-program in ChatGPT, users can input open-ended prompts like "I need a not-so-sweet afternoon pick-me-up drink" or upload images related to the weather or their outfits. The AI combines the available menu options with personalized preferences to suggest drink combinations and guides users to complete payment through their own channels. Starbucks' digital and loyalty business team positions this feature as a retail AI interface for "instant personalization," rather than a standalone tool experiment.
Source: Public Information
ABAB AI Insight
Starbucks此举实质是把“人类状态”与“消费行为”的接口从菜单栏上移至自然语言与图像层。在传统零售中，用户必须先将“情绪—场景”翻译成“产品—规格”；在AI模态下，系统承担这一翻译过程，零配比、糖量、奶基、温度等参数由AI在后端菜单与库存范围内动态组合，极大压缩了“发现—决策—执行”链路。
这一路径也标志着实体零售正成为“AI感知终端”：消费者的穿搭、天气、工作节奏与情绪被纳入推荐系统，零售空间与AI模型的边界进一步模糊。随着时间推移，这类系统在行为数据上的积累，将不仅用于优化SKU，更可能重构门店陈列、人力排班与库存分配的底层逻辑。
从资本结构看，将AI接入点单，本质上是在“人—商品—平台”三者之间插入新的定价层。AI推荐的“喝你今天的状态”式产品，既是营销工具，也渐渐演变为“情绪—消费—数据”的制度化反馈循环，让星巴克在传统利润之上，额外获取基于个性与习惯的“注意力结构”租金。在AI时代，一杯咖啡的价值，越来越取决于模型在多大程度上理解“你此刻是谁”。