Business models in the software industry
Within the Software-Cluster project “SWINNG - Work Package: Business Models In The Software Industry”, CONWEAVER, along with five research partners, studied the modeling, comparison and measurability of business models in the software industry.
Companies can be compared with each other in very different ways. The type of comparison is dependent on the goals to be achieved by the comparison.
Initial question: “How can companies be compared methodically?”
If you want, for example, to compare companies with regard to financial matters, business ratios such as return on investment may do the trick. However, if you want to compare companies with regard to architectural, organizational or strategic aspects, one might draw on the companies' business models (see Burkhart et. al., 2011).
However, business models are not available in a consistent system. In order to compare two or more companies, a solid structure for the comparative description of their business models is first required. But according to which structure should one create descriptions?
For this purpose, our project partners developed a scheme consisting of 25 elements, or differentiators (Schief et. al., 2012). Among other things, one can distinguish between IT companies that offer hardware, software or services. Regardless, you can also distinguish between companies that tend to make their own products (i.e. hardware, software or services), or or rather combine existing products from other companies. All in all, the 25 distinctions create an exemplary, structured description of an IT business model of an IT company.
When a number of companies are modeled according to the fixed formula, a methodical assessment can be done accordingly. In the language of computer science, one would describe such a company as "instances of a class with 25 attributes." In the language of mathematics, one would speak of a vector in a 25-dimensional vector space. Exactly these instances or vectors can be used now to calculate pair-by-pair comparisons between the way companies are represented. The calculated similarities can now be used for various other analyses, for example to summarize groups of related companies.
Conversely, one can ask which differences in business models have a special influence on the performance of a company: are there business models that are more promising than others? This requires additional financial ratios. With these, one can summarize the structuring elements (i.e. the 25 distinctions) into groups according to their impact on company performance (Schief et. al., 2012).
Project Scope
The SWINNG project ran from July 2010 to June 2013 and was funded within the limits of the Software-Cluster by the Federal Ministry of Education and Research (BMBF) under grant number 01IC10S05.
Market overview and visual analytics
The created business models, their similarities as well as the economic indicators of the underlying companies form a complex space that cannot be easily represented.
To make this complex space accessible to a human analyst, a component for the integration and visualization was developed during the project (see Figure 1). The business model dimensions and measures therein are optional and thus allow for an intelligent interaction with the analysis room.
Business models and processes
Adjacent tasks in the joint project studied similar matters in business processes and regarding the transformation of business models into business processes. Similar solution approaches and possible connections were thereby identified that had been little studied, if at all. Figure 2 illustrates the specially developed architecture. We tried to display the different levels and possible correlations along four phases. For more detail, please see Caspar et. al. (2013).
Conclusion
Our research has shown that a structured comparison of companies is indeed possible. This first requires a fixed, domain-specific description formula and a collection of business model instances. The methods of Information Retrieval and Visual Analytics can then be used on instances and additional indicators of the associated companies to perform complex analysis and comparisons.
Project Partner
- CONWEAVER GmbH
- DFKI GmbH
- Fraunhofer IGD
- SAP AG
- Software AG
- TU Darmstadt
Literature
Burkhardt, T.; Krumeich, J.; Werth, D.; & Loos, P. (2011): Analyzing the business model concept: a comprehensive classification of literature. Proc. of the International Conference on Information Systems, Shanghai, 2011
Caspar, J.; Di Valentin, C.; Maier, S.; Mayer, D.; Pussep, A.; Schief, M. (2013): Vom Geschäftsmodell zum Geschäftsprozess und zurück - Konfiguration, Analyse, Transformation, Controlling. Erschienen in: Susanne Strahinger (Hrsg.): HMD - Praxis der Wirtschaftsinformatik, Vol. 50, No. 292, Pages 13-22, dpunkt, Heidelberg, 8/2013
Schief, M.; Pussep, A.; Buxmann, P. (2012): Performance of business models: empirical insights from the software industry. Proc. of the 20th European Conference on Information Systems, Barcelona, 2012
