星巴克正在进一步扩大人工智能技术在门店运营中的应用范围，计划重新设计传统汽车餐厅（Drive-thru）的点餐体验，通过AI识别参与计划的会员，为顾客提供个性化数字菜单，并协助员工更加准确、高效地录入订单。这项计划是星巴克新一轮数字化转型的重要组成部分，预计将在2027财年开始测试。
根据星巴克管理层在最新财报电话会议上披露的信息，公司正在探索一种更加智能化的汽车餐厅服务模式。当参与计划的会员驾车抵达门店时，系统可以识别其身份，并根据过去的消费记录、个人偏好以及常用饮品，自动展示更符合其口味的数字菜单，同时简化后续支付流程。
这意味着，未来顾客可能不再需要面对所有人都相同的固定菜单，而是能够看到根据自身习惯调整过的推荐内容。例如，经常购买燕麦奶拿铁的顾客，可能优先看到相关饮品；习惯购买低糖饮料的用户，则可能获得更加符合其偏好的推荐。
不过，星巴克此次AI计划并不是简单地让人工智能取代传统点餐员。公司计划首先在后台部署AI模型，帮助咖啡师和员工更加准确地记录顾客订单。顾客仍然可以与真人员工进行交流，AI主要负责辅助信息录入，减少因为口音、复杂定制要求或环境噪声造成的错误。
星巴克首席执行官布莱恩·尼科尔表示，公司的目标是让员工能够把更多精力放在与顾客沟通上，同时提高订单处理效率和门店吞吐能力。公司预计从2027财年开始测试这些后台AI模型，但目前尚未公布具体测试门店数量、所使用的AI供应商以及全面推广的时间表。
汽车餐厅对星巴克具有相当重要的商业价值。根据公司管理层披露的数据，大约一半的星巴克顾客使用Drive-thru服务。这意味着，只要能够缩短点餐时间、减少订单错误，就有机会改善高峰时段的运营效率，并提高门店接待能力。
事实上，星巴克已经在其他业务环节逐步引入人工智能。此前，公司推出了Smart Queue智能排队系统，可以对来自店内柜台、汽车餐厅、手机应用和外卖平台的订单进行智能排序，帮助员工合理安排制作顺序，避免不同渠道的订单相互干扰。
在消费者端，星巴克也已经开始探索AI个性化点单。2026年4月，公司推出了ChatGPT平台上的星巴克应用测试版，用户可以直接通过自然语言描述自己的心情、口味偏好，甚至上传服装或周围环境的照片，让AI推荐合适的饮品。
例如，用户可以要求推荐一款不太甜、适合下午饮用的咖啡，或者根据当天的天气和个人喜好寻找饮品。用户还可以进一步调整饮料配方、选择门店，并在星巴克应用程序或官方网站完成最终支付。
星巴克数字业务负责人保罗·里德尔此前表示，消费者并不总是从菜单开始选择饮品，有时他们只是从一种心情或感受出发。AI能够帮助顾客将这种模糊的需求转化为具体的饮品选择。
除了AI点餐，星巴克还计划在2027财年推进一轮大规模技术现代化升级，包括更新员工排班系统、门店销售终端以及移动应用程序，进一步整合订单、库存和人员管理流程。
公司还准备在第二财季开始将外卖服务直接整合进星巴克应用程序，让顾客能够通过同一个平台完成外卖订单，并继续累积或兑换会员积分。
值得注意的是，星巴克并不打算将人工智能作为单纯削减员工数量的工具。管理层强调，技术投资的主要目的是减少员工在重复性操作和订单管理上花费的时间，让他们能够把更多精力投入饮品制作和顾客服务。
当然，个性化菜单也涉及消费者隐私问题。要实现自动识别会员并根据历史消费记录推荐商品，星巴克需要处理用户身份、消费习惯以及相关账户数据。如何取得用户授权、保护个人信息，以及避免系统错误识别顾客，将成为这类服务实际推广时需要解决的问题。
从整个餐饮行业来看，星巴克正在探索一种不同于完全自动化语音点餐的AI应用模式。此前，麦当劳曾与IBM合作测试AI语音点餐系统，但最终于2024年结束相关试点。星巴克此次选择让真人员工继续与顾客沟通，由AI在后台辅助录入和个性化推荐，显然更加注重人工服务与自动化技术之间的配合。
随着AI技术逐渐进入零售、餐饮和日常消费场景，未来的汽车餐厅可能不再只是一个统一菜单和付款窗口，而是能够根据每位顾客的身份、消费习惯和实时需求动态调整服务内容。对于星巴克而言，这不仅是提高点餐效率的技术升级，也是在竞争日益激烈的咖啡市场中增强会员黏性、推动个性化消费的一项长期战略。
本文由 AI图鉴自动抓取并翻译，版权归原作者所有；如需查看原文请点下方链接。
Starbucks is further expanding the application of artificial intelligence technology in door shop operations, planning to redesign the dining experience of the traditional automobile restaurant (Drive-thru), providing personalized digital menus to customers through AI identification of participants, and assisting staff to enter orders more accurately and efficiently. The plan is an important part of the new digital transformation of Starbucks and is expected to be tested in FY2027.
Based on the information disclosed by Starbuck management at the latest financial teleconference, the company is exploring a more intelligent model of car restaurant service. When the participating members arrive at the door shop, the system can identify them and automatically display a more symmetrical menu based on past consumption records, personal preferences and common drinks, while simplifying the subsequent payment process.
This means that future customers may no longer need to face fixed menus that are the same for all, but are able to see recommended content adapted to their own habits. For example, customers who regularly buy barley latte may give priority to related drinks; users who are used to buying low-sweet drinks may receive more favourable advice.
However, Starbucks' AI project is not simply a substitute for the traditional tampon. The company plans to first deploy the AI model to the backstage to help the cafés and employees record customer orders more accurately. Customers are still able to communicate with real workers, and AI is primarily responsible for supporting the recording of information to reduce errors caused by accents, complex customization requirements or environmental noise.
The CEO of Starbucks, Brian Nicole, stated that the company ' s objective was to enable employees to focus more on communicating with their customers, while improving order processing efficiency and door-to-door capacity. The company expects to test these back-office AI models starting in the financial year 2027, but has not yet published the number of specific testing outlets, the AI suppliers used and the timetable for full roll-out.
Car restaurants are of considerable commercial value to Starbucks. According to data disclosed by company management, about half of Starbuck customers use the services of Drive-thru. This means that there is an opportunity to improve the operational efficiency of peak hours and to improve the capacity of the door shop as long as it is possible to reduce the number of meals and order errors.
In fact, Starbucks has gradually introduced artificial intelligence into other operations. Prior to this, the company introduced the Smart Queue Queuer system, which allows for the intelligent sequencing of orders from shop counters, car restaurants, mobile phone applications and out-of-sale platforms, and helps employees to rationalize their sequencing and avoid interference with orders from different channels.
At the consumer level, Starbucks has also begun to explore AI-personalized lists. In April 2026, the company launched a test version of the Starbuck application on the ChatGPT platform, where users can directly describe their moods, taste preferences and even upload photographs of their clothing or surroundings in natural languages to allow AI to recommend suitable drinks.
For example, a user may request a less sweet coffee suitable for the afternoon or find a drink based on the weather and personal preferences of the day. Users could further adjust the beverage formulation, select the door shop and finalize the payment at the Starbucks application or official website.
Paul Riedel, head of the Starbucks Digital Service, had previously stated that consumers did not always choose drinks from the menu, and that sometimes they started from a mood or feeling. AI can help customers translate this vague demand into specific drinking choices.
In addition to AI snacks, Starbucks plans to move forward with a large-scale technological modernization round in the 2027 fiscal year, including upgrading the staff shift system, door-to-door sales terminals and mobile applications to further integrate order, inventory and personnel management processes.
The company is also preparing to begin the direct integration of out-of-sale services into the Starbucks application in the second financial season, allowing customers to complete out-of-sale orders through the same platform and to continue to accumulate or convert membership credits.
It is worth noting that Starbucks does not intend to use artificial intelligence as a mere tool to reduce the number of employees. Management emphasized that the main purpose of technology investments was to reduce staff time spent on repetitive operations and order management so that they could devote more energy to the production of drinks and customer services.
Of course, the personalized menu also addresses consumer privacy. To automatically identify members and recommend commodities based on historical consumption records, Starbucks needs to process user identification, consumption habits and related account data. Access to user authorization, protection of personal information and avoidance of system errors in customer identification will be issues that need to be addressed in the practical promotion of such services.
In the overall catering industry, Starbucks is exploring an AI application that differs from a fully automated voice order. Previously, McDonald ' s had worked with IBM to test the AI voice order system, but the pilot was eventually completed in 2024. The choice of Starbucks this time to keep real workers in contact with their customers, with AI backstage helper entry and individualized recommendations, clearly places greater emphasis on the interface between manual services and automated technology.
As AI technology moves into retail, catering and daily consumption scenarios, future car restaurants may no longer be a single menu and payment window, but rather be able to adapt the content of the service to the identity, consumption habits and real-time demand dynamics of each customer. For Starbucks, this is not only a technological upgrading to improve the efficiency of snacks, but also a long-term strategy to increase member adhesion and personalized consumption in the increasingly competitive coffee market.
Auto-fetched and translated by AI Atlas. All rights belong to the original author.