How is AI changing sales and customer service? What role do use cases play? And why do implementations often fail in day-to-day business despite technological progress?
These were the questions discussed in the Expert Talk “No Use Case, No AI Success” by three experts from consulting and industry: Roman Erlacher (NovaTaste), Ron Boes (SYBIT) and Jonas Degener (SYBIT). Nearly 200 participants heard one key takeaway: To use AI profitably, companies must move away from the tool hype – and toward clearly defined business goals.

Below are the most important questions, answers, and application examples from the talk.

 

Between Hype and Reality: How Established Is AI in Business?

In many companies, AI is still heavily technology-driven. New tools are released almost weekly, and headlines focus more on features than on impact. This is beginning to change: The focus is shifting towards solving specific business problems – particularly in the context of Customer Experience (CX). AI projects are no longer implemented “just for show” but to address issues such as long response times, overloaded call centers, or inefficient document handling.

 

From Problem to Solution: How Does a Meaningful AI Use Case Emerge?

The key is to reverse the perspective: Instead of starting with “Which tool should we use?”, companies should ask “Which problem do we want to solve?” Only then should they decide on the right tool, platform, and architecture. Platforms like SAP BTP or Microsoft Azure can offer strategic advantages, but they must be evaluated within the overall architecture – including data models, governance, and interfaces.

 

Developing Use Cases: From Process to Real Value

An effective AI use case is neither accidental nor overcomplicated. The Expert Talk made it clear: Analyzing processes, systematically identifying pain points, and involving the right people are the foundation for practical applications. The solution must fit seamlessly into day-to-day work – “Bring AI to people, not the other way around.”

One particularly useful method: shadow working. Here, an expert accompanies employees in their daily work, observes workflows, and identifies improvement opportunities that are often invisible in process documentation. Combined with targeted sparring sessions between business units and consultants, this approach produces use cases that are practical, feasible, and economically relevant.

 

Measuring Business Value: How to Evaluate Use Cases

A recurring challenge is measuring the business impact of AI use cases. ROI is not always easy to calculate. In practice, a relative assessment along two axes – expected benefit vs. feasibility – works well. The aim is not perfection but prioritizing the most promising ideas and achieving quick wins. Typical measurable effects include:

  • Fewer service requests
  • Shorter processing times
  • Increased self-service adoption
  • Improved data quality

Application Examples: AI in Everyday CX Operations

The talk showcased numerous real-life AI use cases. Here are five examples:

 

Use Case: Service Triage Agent

  • Description: Automatically detects, categorizes, and routes customer inquiries
  • Business Value: Reduces manual effort in first-level support, shortens response times, and significantly relieves the service team

 

Use Case: Ticket Enrichment

  • Description: Automatically fills service tickets with structured information
  • Business Value: Improves data quality and completeness, provides a better basis for analysis, saves time per ticket, enhances quality management, and creates value for product development

 

Use Case: Translation

  • Description: AI-supported translation of content for international collaboration
  • Business Value: Speeds up and improves cross-border communication without relying on external services

 

Use Case: Quote Generation

  • Description: Automated creation of sales quotes based on customer data
  • Business Value: Shortens time-to-quote, improves consistency in communication, and frees up time for consultation instead of administration

 

Use Case: Analysis

  • Description: Intelligent evaluation of customer, ticket, and feedback data
  • Business Value: Identifies recurring issues, provides input for quality management, and supports continuous process improvement

These examples show: AI doesn’t have to be a large-scale project. Many applications can be integrated into existing processes with manageable effort – as long as there is a clearly defined use case.

 

Challenges and Solutions: What Companies Should Consider

The discussion also covered common challenges companies face – and practical solutions.

 

Typical challenges:

  • Actionism without clear goals (“We need AI, too!”)
  • Uncertainty and fear within teams
  • Poor data quality and structure
  • Shadow projects without governance
  • Unclear responsibilities
  • Technology overkill
  • Concerns about supposedly long-term technology commitments

 

Proven solutions from practice:

  • Start with the business problem – not the tool
  • Use shadow working and interviews for process analysis
  • Hold joint use case workshops with business and IT
  • Build central governance structures
  • Establish standards
  • Involve and train employees early on
  • Actively integrate existing internal initiatives instead of blocking them

 

Standardization as an Enabler – Not a Limitation

Standardization plays a crucial, yet often underestimated role in AI. It is essential for scalability, reusability, and security. Data protection and compliance are particularly critical: Many AI experiments at the departmental level unintentionally violate regulations. Building a central architecture with clear roles, APIs, and data models creates security – and speeds up the implementation of new use cases, especially when each one is designed from the start to serve as a foundation for future applications.

 

Success Factors from Practice: The Experts’ Advice

At the end of the talk, all three experts shared their most important advice:

  • Jonas Degener: “Use as much AI as necessary, but as little as possible. Every model call has a cost – financially and technologically. Often a traditional algorithm is enough.”
  • Roman Erlacher: “There needs to be real collaboration between C-level and business units. Everyone has a different level of knowledge – that must be taken seriously and supported with change management.”
  • Ron Boes: “Don’t act out of fear. If you try to secure everything, you’ll never get moving. You need the courage to take the first steps – but with clear priorities and focus.”

 

Conclusion: AI Creates Business Value Only with the Right Use Case

The talk made one thing clear: The question is not if you use AI – but what for. To successfully leverage AI in CX processes, companies need three things:

  1. A specific problem to solve
  2. A strategy for platform, data, and governance
  3. An organization ready to learn step by step

Technology alone does not provide an advantage. Only when it is connected to real business needs does it generate measurable value – and the potential to truly rethink CX.

Ready to Turn AI into a Business Case?

Our specialized AI workshop series takes you from idea to a solid business case in three steps:
We identify economically relevant use cases, evaluate them systematically, and translate them directly into an actionable roadmap.

In a free discovery call, we’ll show you how our workshop series works, what results you can expect – and how we can quickly create a solid decision-making basis for your organization as well.

 

Book your discovery call now