AI Methods in Innovation Management: Guidance for Small and Medium-Sized Businesses

October 30, 2026

Artificial intelligence opens up new possibilities throughout the entire innovation process. But which AI-based methods are suitable for which tasks, and which of them can already be effectively implemented in small and medium-sized enterprises today? A new methodology radar provides guidance to innovation leaders and supports them in selecting appropriate approaches.

Objective: Turning a fragmented landscape of methods into a decision-making tool

Artificial intelligence can support innovation processes in a variety of ways: It helps evaluate large amounts of information, identify customer needs, develop and structure ideas, digitally test concepts, and assess market developments at an early stage. In particular, it offers SMEs the opportunity to make innovation activities more efficient and to use limited resources in a more targeted manner. In practice, however, a key question often arises:

Which AI-based methods are suitable for which tasks in the innovation process?

Until now, there has been a lack of a structured overview that offers practical guidance to innovation managers in small and medium-sized enterprises. This is where the new Method Radar comes in. It originated as a master’s thesis by Jona Negwer at the Chair of Industrial Management (Prof. Voigt), in collaboration with the Innovation Management team at Bayern Innovativ.
The master’s thesis was awarded the Norbert Thom Innovation Prize by the innovation association Querkraft.

The Method Radar compiles scientifically described AI-based methods of innovation management, assigns them to specific tasks along the innovation process, and supplements them with assessments from experienced professionals in the business world. This provides an initial guide for companies that wish to use AI specifically for concrete tasks in innovation management.

Procedure: From Scientific Research to an Interactive Methodology Radar

The development of the Methodology Radar took place in three sequential steps: identification, evaluation, and consolidation.

1. Identifying and Structuring Methods

In the first step, scientific publications were systematically reviewed to identify which AI-based methods are described for tasks in innovation management.

To ensure that only truly applicable methods were considered, each identified approach was evaluated based on four criteria (well-defined procedure / specific task in innovation management / clear sequence of steps / AI as a central component).

In this way, a total of 23 AI-based innovation management methods were identified. They were then assigned to four phases of the innovation process and eight more specific areas of application. These range from identifying opportunities and customer needs, through idea generation and evaluation, to concept development, prototyping, and support for market entry.

2. Evaluating Methods by Experts

In the second step, 13 representative methods were selected and evaluated by 33 experts in innovation management from German small and medium-sized enterprises as part of a standardized online survey.
Three action-oriented categories were used for the evaluation:

Within these four phases, the methods are assigned to eight more specific areas of application. This allows innovation leaders to start with their specific task and systematically assess which AI-based methods are generally suitable for it.

An overview of the eight application areas:

1. Analyzing innovation capability
2. Utilizing strategic foresight and market information
3. Better understanding customer needs and target groups
4. Developing ideas
5. Structuring and selecting ideas
6. Evaluate concepts and investment projects
7. Developing prototypes and product solutions
8. Support market entry decisions

Key Insights: What the Methodology Radar Reveals

The Methodology Radar makes it clear that AI can support innovation management throughout the entire innovation process.

However, the methods differ in terms of how easy they are to implement and what prerequisites companies need to use them.

→ A simple start is already possible today
Especially in the early phases of the innovation process, several relatively low-threshold methods are available. For example, they help identify customer needs, prepare information for strategic decisions, develop new ideas, or structure larger collections of ideas.

→ More sophisticated applications require preparation
The Radar also highlights relevant use cases for the development, testing, and market launch of solutions. However, methods such as virtual product development or AI-supported demand forecasting are more often dependent on existing data, digital development environments, and specific expertise.

→ Building expertise becomes a key success factor
The study shows that the availability of AI or technical infrastructure alone does not determine successful implementation. A key challenge lies in selecting suitable methods, critically evaluating AI results, and integrating the insights gained into existing innovation processes. Only in this way can companies truly benefit from the opportunities offered by AI-based innovation management methods.

How can SMEs make use of the Methods Radar?

Companies should not start by asking where an available AI tool could be used. It is more effective to first clearly define the innovation task.

Selecting a suitable method in five steps:

1. Define the innovation task: Clarify which specific problem needs to be solved or which operational goal needs to be achieved.
2. Select the appropriate application area: Use the segments of the radar to narrow down the selection to methods suitable for this specific task.
3. Use the evaluation as an initial guide: Consider “Act” methods first, without ruling out “Prepare” methods outright.
4. Examine the process and sources in more detail: Review the workflow, the expected outcome, and the underlying information for the selected method.
5. Assess the fit for the company: Align the method with the available data, existing expertise, technical infrastructure, available resources, and organizational framework.

Classification as “Prepare” or “Watch” does not equate to a negative evaluation. Rather, it highlights that a method can offer benefits, but these are contingent on certain prerequisites. In this way, the Radar becomes more than just a selection tool for methods that can be implemented immediately. It can also highlight the capabilities a company should develop in the future for more sophisticated AI applications.