January 26, 2026
Posted by Aline Puhan-Schulz
SMEs in manufacturing are experiencing a phase in which business models, customer expectations and competitive dynamics are changing at the same time. While products are becoming more technically sophisticated and variants are increasing, decision-making processes on the customer side are accelerating. Sales organizations are therefore confronted with a growing fragmentation of information, which is often spread across different systems, departments and documents. This fragmentation creates friction, makes reconciliation difficult and means that sales representatives spend a large part of their time searching for, processing, or reformulating information. This creates a problem area that is barely visible in many companies, but is massively noticeable: Sales have been busy processing information instead of customer interaction for too long. Since speed is a key competitive factor today, this delay can be decisive.
In order to cope with increasing complexity in sales, many companies rely on additional tools or additional software modules. These solutions make valuable contributions, particularly where processes are stable, repeatable and data-driven. At the same time, there are numerous situations in B2B sales in which classic IT alone is not enough: Customer requirements are heterogeneous, technical issues are complex and many work steps are based on experience, interpretation or contextual knowledge. Traditional software can manage data, but it can't understand or generate content on its own.
This is exactly where artificial intelligence complements the existing system environment. In the Bitkom study: Bitkom (2025) “Generative AI in companies. Legal issues relating to the use of generative artificial intelligence in companies.” (1) AI is described as a technology that simulates human-like cognitive processes, learns from data and recognizes patterns that have not been explicitly programmed. While classic software can update inventory lists, for example, it is unable to predict future demand or generate content flexibly. An AI system, on the other hand, can, in accordance with the definition of the European AI Act, interpret inputs, derive from them which outputs should be generated, and generate this content. It works adaptively, context-sensitively and with a degree of autonomy that traditional IT systems lack. (2)
Within this broad field, generative AI is particularly relevant because it can generate new content, texts, images, audios, videos or program code, based on a simple prompt. It is therefore different from discriminative AI models, which classify patterns or recognize differences. The difference between AI model and AI system is important. An AI model is the trained mathematical foundation that recognizes patterns or generates content. An AI system, on the other hand, comprises all elements that are necessary to use the technology in practice: user interfaces, data management, interfaces, security mechanisms, and application logic. It is only the combination of model and system that makes generative AI a tool that can be used in sales. For companies, this means that they do not have to develop models themselves, but can use systems that build on existing basic models.
These technical principles are relevant because they explain the practical benefits. Generative AI can summarize documents, reformulate technical information, evaluate variants, structure meeting notes, or generate draft offers. It provides support where language, structuring and interrelationships are central, i.e. precisely with the tasks that shape B2B sales. A recent analysis by McKinsey shows that companies that use generative AI in sales not only work more efficiently, but also achieve noticeable economic effects: faster response times, better pipeline quality and significantly higher consistency in customer communication. (3) By automating recurring tasks, intelligently linking data and the ability to provide highly personalized communication on a large scale, generative AI could sustainably increase sales effectiveness.
At the same time, generative AI remains a statistical system that cannot be fully explained. Different prompts can lead to different results, and a generative AI model cannot provide a transparent explanation of how a particular answer came about. The Bitkom guidelines emphasize that generative models are powerful but not deterministic. In order to minimize these risks, technical and organizational guardrails (guardrails) are required. The most important guardrail here is the person: He remains responsible, reviews the drafts and makes the final decisions.
Part 1 of the playbook thus makes it clear that industrial sales need new tools to solve their structural bottlenecks. The consequences of today's stress are too great to ignore. Generative AI bridges the gap between growing complexity and limited resources, not by replacing people, but by enabling them to use their expertise much more effectively. Part 3 of the series will be about how companies can prepare this technological opportunity organisationally and culturally.
.svg)