A Search for Discriminative Linguistic Markers in ICT Practitioner Discourse, for the Ex Ante Identification of Disruptive Innovation
Disruptive innovations have the potential to disrupt markets, and drive them in new directions. A common problem faced by business organizations is identifying such disruptive innovations. From a managerial perspective, there is real value in being able to accurately identify disruptive innovations early in the product life-cycle, as it affords the organization the opportunity to put in place business strategies that leverage this information, to gain maximal competitive advantage. This investigation was undertaken to determine if linguistic markers could be identified in ICT practitioner discourse that could be used to discriminate between traditional business intelligence (BI) - the legacy or incumbent technology, and software-as-a-service (SaaS) BI - a new technology and candidate disruptive innovation. Quantitative content analysis undertaken using the tool Veneficium WordFrequencyCounter, was used to analyze written practitioner discourse identified from within the Industry Newsgroup file of LexisNexis Academic universe. Analysis was undertaken using attribute sets derived deductively from the academic literature, and inductively from the data itself, which provided both manifest and latent meaning of component words. Individual relative word associations with both the traditional BI and SaaS BI corpora were also analyzed. Analysis of the attribute set usage data provided evidence that manifest and latent word meaning remained consistent for the time period investigated in this study (2000 to 2012), and so could support the purpose of this study, and was suggestive of the fact that SaaS BI could be a disruptive technology. The study also identified that there was a significant difference in vendor and industry attribute set usage between the SaaS BI and traditional BI corpora, consistent with the Abernathy-Utterback model. Analysis of individual word associations with the traditional BI and SaaS BI corpora identified a number of word association patterns that could discriminate between traditional BI and SaaS BI that may be transferable to other technologies. A crossover event pattern was also identified (in which the word association pattern switches between the incumbent and new technology), which may be able to provide an indication that a technology innovation is, or is about to become, disruptive. This study contributes a new approach for investigating disruptive innovation, and highlights the potential of using content analysis of practitioner discourse to identify linguistic markers for disruptive innovation. The key contribution of the study is the observation that discriminative linguistic markers can in fact be identified, and that such markers appear to have predictive capabilities. That is, they may allow organizations to identify disruptive innovations ex ante.