“Data-based decisions allow us to tap sales potential and increase customer satisfaction at the same time.” The importance of customer feedback can be expressed in two figures: 93 percent of consumers read online reviews before making a purchase, and 94 percent of consumers won’t buy a poorly rated product1. AI adds a great amount of value, particularly in the areas of customer service and sales and marketing2. At the Bosch booth in hall 1.1 at this year’s IFA, BSH customer service will literally be in the spotlight with its own area.
Hybrid model in customer service: humans and AI work hand in hand
Dealing with “ones and zeros” is something that needs to be learned, and it takes time for people to get used to it. According to independent market observers, some two-thirds are skeptical about AI in customer service and prefer to interact with a human3. That’s why Bosch prefers to use a hybrid model, in which humans and AI work hand in hand. Bosch’s consumer-facing areas receive over 30 million inquiries every year. They range from product questions and job offers to general information, and Bosch associates provide answers in up to 40 languages. In customer service, AI is primarily used for recurring processes that are tedious and time-consuming for people, such as translating, sorting, classifying, and forwarding service tickets. This saves service teams valuable time and lets them deal with customer concerns in a more focused and personal way.
AI is almost always right: 2,500 hours of manual rework saved
AI technologies such as chatbots, virtual assistants, and automated analyses are now ushering in a new era of customer communication – with astonishing success. At Bosch Power Tools, AI agents are helping to categorize the roughly 120,000 after-sales service tickets that the location in Willershausen, Germany, receives each year – and are replacing a rule-based system built on machine learning. Less than a year later, tickets are now categorized with over 90 percent accuracy. This has enabled Power Tools to save an average of around 2,500 hours of laborious manual rework per year. Due to that outstanding result, the service is now to be rolled out to the 23 Power Tools service locations worldwide, covering 40 countries and around 1.5 million service requests. In Willershausen, the service team was also trained to configure prompts – the commands or instructions for the AI. After all, AI is only as smart as humans train it and teach it to be.
A new era: agentic AI analyzes written customer inquiries in real time
In another global project at Power Tools, agentic AI interacts with several connected systems to take customer service to a new level. Since June, an intelligent platform has been supporting service teams in analyzing the more than 1.2 million written inquiries received annually in real time, searching through relevant information from CRM and knowledge systems, and creating context-based response suggestions – quickly, reliably, and in a way that can be tailored to the individual. In addition, an intelligent translation function greatly improves customer contact for low-volume languages. Plans are also in place to extend the service to telephone calls, whereby the AI will track conversations with the caller’s consent and supporting the service associate with suggestions for answers and actions. The project has huge scaling potential: the service is to be rolled out to other Bosch units at the beginning of next year. Standard tools and IT solutions – provided by Bosch’s internal IT and service areas – also help here.
High customer satisfaction: Bosch Smart Home’s chatbot can almost always help
The Bosch Smart Home division has also benefited from standardized solutions from Bosch IT. Only six months passed from the conception of the idea to the first use of a chatbot with generative AI on the division’s website. It’s been supporting customer service at Bosch Smart Home for a good year now, and with great success: more than 97 percent of chatbot users who submit a rating are satisfied with the answer they received. In addition, over 95 percent of customer queries are resolved with the chatbot alone and not forwarded to a service team member. Around 20 percent of the incoming volume is now processed by the chatbot, achieving one of the division’s strategic goals: give customers the option of self-service so that they can resolve questions themselves more quickly. In addition, an internal AI copilot supports the service teams by suggesting answers. The copilot is already used for around 12 percent of queries and translates international tickets for over 6 percent of inquiries. This combination of self-service and support for service teams allows them to focus on complex cases and further increase service quality.
In over 80 percent of cases, a single service call is sufficient
Bosch home appliances are built to last. One basis for this is the company’s voluntary stockpiling of spare parts for up to 15 years. To make these logistics efficient, predictive AI optimizes warehousing and helps prepare service calls, considering everything from route planning to which spare parts to carry in the vehicle. Coupled with the expertise of the service technicians, this means that over 80 percent of repairs are completed on the first visit and parts are on hand without any waiting time. In addition, generative AI helps associates at the BSH Home Appliances service center answer customer queries faster and more precisely. It acts as an intelligent tool that supports associates, but doesn’t replace them. The result is a high level of customer satisfaction thanks to fast service and competent advice with optimized processes.
AI listens to the “voice of the customer” and evaluates millions of reviews
Customer-centric product development has been firmly anchored in the corporate strategy of Bosch’s two largest consumer divisions – BSH Home Appliances and Power Tools – for many years. Successful pilots were already underway in 2018, using AI methods to systematically record publicly available product reviews and consumer feedback and derive potential improvements for product development, marketing, sales, and customer service. Without AI, this would be an almost impossible undertaking due to the huge masses of unstructured data. At Bosch Power Tools, for example, 3.4 million online customer feedback messages were evaluated in 2024. Generative AI has been in use since the beginning of 2024. It provides BSH with important information about all BSH home appliance brands, all product categories, and data sources for product evaluations and NPS feedback. It can now also recognize thousands of previously undefined topics. This enables highly granular searches for special topics related to individual products, which increases the added value for the divisions. AI-based consumer feedback analysis is currently being used in 45 countries – and in potentially all languages, since these can be translated if required. In the case of the web store of a major German retail partner, this resulted in improved product descriptions and additional information on accessories, which led to more sales and better product ratings. In another example, customers complained about the lack of contrast in the symbol lettering on the control panel of an oven, and so the design team made improvements for the launch of a new oven series. And for cooktops sold via a large online retailer, the packaging, logistics, and transportation process were improved – without data-driven insight, this potential might have gone undetected.
The Bosch press conference at IFA will take place on September 3 at 12:00 noon at the Bosch booth in hall 1.1.