Business intelligence (BI) is a strategic approach for systematically targeting, tracking, communicating and transforming relevant weak signs1 into actionable information on which strategic decision-making is based. Despite the increasing importance of BI, there is little underlying theoretical work, which directly can guide the interpretation of ambiguous weak signs. This paper gives an insight into the issue through a new strategic business intelligence system called PUZZLE. We describe this system and validate it by designing a prototype, test the system using in-depth interviews, and hold learning sessions in order to further knowledge about BI. The main results from tests show that: interpreting weak signs is potentially important for senior managers, consultants, and researchers; interpretation can be achieved gradually by bringing the weak signs together using a tracking form based upon the concept of actor/theme/weak signs/enrichment /links; interpreting weak signs is a complex process of establishing links between the weak signs. Final results show that the individual cognitive process appears heuristic when interpreting weak signs. Implications for strategic management practice and research are addressed.
This paper has presented a SBIS, based on exploratory research, to support managers in practice to better handling weak signs and to ensure success of the business strategy. This is an elusive issue related to knowledge management that has received little academic attention.
This paper brings three main contributions: conceptual contribution in the form of a creative method to support interpreting weak signs, design contribution in the form of computer tool to support this method, and empirical contribution for understanding the process of interpreting weak signs.
These findings have implications for both research and practice. From a practical perspective, our findings encourage the use of PUZZLE by managers to perform the process of BI. The use of a tracking form, based on ‘actor/theme/weak sign/ personnel enrichment/link, transformations’, continuous discussion and creation of puzzles, are all suitable for company's use. That form constitutes a shared language for understanding BI and that will facilitate its use in companies. PUZZLE can also be a viable training method for managers who want to learn BI process quickly and effectively.
From a research perspective this paper supports previous researches and adds new findings. This paper suggests interpreting weak signs as both a process and a product. The process requires the transformation of seemingly unrelated information elements into useful puzzles (product) using links in order to produce hypotheses about potential opportunities and threats. We found that collective interpretation produces more value than when individuals work alone. Convergence of individual interpretations following discussions seems to be fast and easy. These findings support the previous results of El Sawy and Pauchant (1988). Moreover, this study suggests interpreting weak signs and BI as complex processes in which individuals follow a heuristic cognitive process. The discovery of managers' difficulties in connecting weak signs was totally unexpected because it had never been highlighted before and brings new findings to strategic management literature. This finding also contradicts most statements of previous researches (Daft and Weick, 1984, Martinsons, 1994, Attaway, 1998, Subramanian and IsHak, 1998 and Freeman, 1999).
In addition, this research encourages the use of the engineering research methodology proposed by Channel et al. (1997). This could be very helpful for researchers who want to understand the application of SIS in an area where little underlying theory exists, or for ill-structured problems. This methodology is very useful because of its ability to stimulate interest, create new information, open new research directions, and to propose refinement of concepts.
Results of this study should be examined in light of its limitations. First, we do not claim that this research is other than exploratory since the sample size was small, which limits the ability to generalise the result beyond this sample. Second, we obtained data from a single informant and that may introduce respondent bias and limit the generalisibility of the results. However, as companies compete in a turbulent environment, the results and their implications are not industry-specific. Third, limits of this research should also be examined with the reliability of its results. This is concerned with the extent to which our results are reproducible and consistent with those produced by other approaches. This is a difficult question to answer in any research study, particularly one as exploratory as ours. Our intent was to explore whether the PUZZLE approach could provide insights into the important research question posed above, and not to engage in a rigorous comparison of alternative BI processes and software.
Future research can take several directions. First this research suggests to continue testing PUZZLE, through replication under different conditions and in other organisations, in order to learn more about weak signs interpretation including: what are the frequency of using and updating puzzles? For which enterprises the PUZZLE approach is more suitable? What are the characteristics of those enterprises (size, activities, etc.)? Does the use of the prototype result in more creative ideas and better interpretation of weak signs? Also there is a need to apply PUZZLE to a specific case study such as an emerging technology, like the web telephone.
Second, findings about the difficulty to connect weak signs to each other suggests new research directions toward understanding the adequacy of organisational memory models with PUZZLE. One model that is well-known and respected in this field is that of Horst Rittel's Issue Based Information Systems (IBIS) ( Rittel and Kunz, 1970). The IBIS model focuses on the articulation of key issues in the design problem. Each issue can have many positions. A position is a statement or assertion that resolves the issue. Each of an issue's positions, in turn, may involve one or more arguments that either support that position or object to it. Thus each separate issue is the root of a (possibly empty) tree, with the children of the issues being positions, and the children of the positions being arguments. In addition, new issues raised during discussion may be posted at any time and linked into the most appropriate nodes. Those postings serve as an organisational memory that not only captures the final decisions when trees are ‘closed,’ but also shows the history of the alternatives explored ( Conklin and Begeman, 1988). This paper encourages future research to apply this model with puzzle by creating a graphical web based IBIS that allows all members of a company to read and post an issue (weak sign), positions (enrichments and questions), and arguments (links) at any time using their web connections. Those suggestions will help to build a complete and coherent theory around PUZZLE for BI.
