Objective: Exploring consumer topics and trends

Knowing consumer perception regarding common and trending topics is an important prerequisite for innovation, product development and other strategic decisions. Topic Network Analysis delivers this information by providing quantitative and qualitative figures regarding which attributes consumers state in the topic of interest, which ones they criticize, appreciate, or wish for. Moreover, pairs of subjects can be compared and their differences visualized, while the subjects may either be different topics, relating topics, or the same topic at a different point in time in order to detect shifts in perception. Another finding is consumer associations, showing how attributes are linked to each other in consumers' minds. In the Topic Network Analysis method an extensive amount of data from Web and social media sources is collected, filtered for consumer statements and developed using the most advanced technologies from the field of machine learning, social network analysis and natural language processing.

Die Meinung des Kunden zählt

The method's origin and results

Topic Network Analysis is derived from the Brand Network Analysis method originally invented at the University of Cologne. Its purpose is to visualize consumer's associations regarding a certain topic by means of a network graph. In these graphs, one may easily read the characteristics of the topic to be explored:

  • Consumer associations,
  • Emotional attributes (including their sentiment),
  • Justification links and patterns of thought.


Process: Highly customized analysis

Before starting a Topic Network Analysis study, our first step is always a free pre-test in which we evaluate the feasibility of the Topic Network Analysis study. When confirmed, our analysts draft a custom research design with respect to your objectives.

If necessary, available algorithms will be adjusted or new ones implemented in order to perfectly match the research design. Finally, relevant Web and social media data will be collected using our own crawler infrastructure.

When the preliminary results reveal a first glance at the consumers' perception, a feedback meeting is held to agree on the key aspects to further focus before final data filtering, analysis, and results presentation can take place.

By request you may subsequently get access to all raw and processed analysis data, thus making Topic Network Analysis thoroughly comprehensible.