The Evolution of Corporate Prediction Aggregation Mechanisms: Towards Leveraging the Frontline for Strategic Issue Identification under Uncertainty

Carina Antonia Hallin, Torben Juul Andersen, Sheen S. Levine, Sigbjørn Tveterås

    Research output: Contribution to conferencePaperResearchpeer-review

    5 Downloads (Pure)

    Abstract

    This paper presents different corporate prediction aggregation techniques and introduces a new type of prediction mechanism linking the sensing of operational capabilities by frontline employees to the identification of fuzzy events and emerging strategic issues as ‘early warning signals’. Based on the literatures on prediction markets and fuzzy logic the methodology collects information from many diverse frontline employees to develop valid signaling predictors. Individuals in the frontline gain deep insights as they perform operational activities in direct interactions with many internal and external stakeholders and we tap into this unique knowledge source to identify new issues and opportunities for ongoing strategic decision-making. Aggregating dispersed information from crowds is not a new phenomenon. The capacity to aggregate heterogeneous and dispersed information from the environment is seen as a critical input for strategic decision making [Arrow, 1974; Hayek, 1945; Stinchcombe, 1990]. Hayek’s notion of information aggregation and dispersed knowledge, has established the foundations for prediction markets where the main objective of prediction markets is to create accurate predictions of given issues of interest, and such markets have demonstrated that crowds have the ability to predict outcomes [Berg, Forsythe and Rietz, 1996; 1997; Thompson, 2012: Wolfers and Zitzewitz, 2004]. Corporate prediction markets take various forms. Borrowing from the concepts used by Spann and Skiera (2003), they refer to the evolution in prediction aggregation as first-generation (G1) and second-generation (G2) prediction markets. In G1 markets participating employees invest in the outcome of already defined problems, such as, forecasts on next quarter’s sales volume, market entries by new competitors or performance of certain markets. In recent years, G2 markets, preference markets, aggregate predictions from the firm’s stakeholders about the probable success rates of various product concepts and ideas [Slamka, Jank and Skiera, 2012]. Hence, the participants in G1 and G2 prediction markets typically invest in the outcome of predefined time constrained issues. Here we propose a spring-off mechanism to G1 and G2 markets based on predictions without markets of fuzzy events or emerging issues not yet clearly defined, but nonetheless evolving phenomena to consider for responsive strategies. The notion of an event and its related probability constitute the most basic concepts of probability theory. An event is an accurately specified collection of points in a sample range. In contrast, in everyday life individuals often encounter situations in which an “event” is fuzzy and rather ill-defined than being a sharply defined collection of points [Zadeh, 1965]. We draw on fuzzy sets theory that offers mathematical models to deal with information that is uncertain and vague. That is, our contribution proposes formalized tools to deal with the intrinsic fuzziness in decision making problems [Fisher, 2003].
    Original languageEnglish
    Publication date2015
    Number of pages4
    Publication statusPublished - 2015
    EventCollective Intelligence 2015: Complex Systems - Santa Clara, CA, United States
    Duration: 31 May 20152 Jun 2015
    https://sites.lsa.umich.edu/collectiveintelligence/

    Conference

    ConferenceCollective Intelligence 2015
    CountryUnited States
    CitySanta Clara, CA
    Period31/05/201502/06/2015
    Internet address

    Cite this

    Hallin, C. A., Juul Andersen, T., Levine, S. S., & Tveterås, S. (2015). The Evolution of Corporate Prediction Aggregation Mechanisms: Towards Leveraging the Frontline for Strategic Issue Identification under Uncertainty. Paper presented at Collective Intelligence 2015, Santa Clara, CA, United States.