Perceptions of barriers to Business Intelligence and weak signals management in Kuwait

Perceptions of barriers to Business Intelligence and weak signals management in Kuwait Dr Kamel Rouibah Department of Quantitative & Information Systems College of Business Administration Kuwait University Abstract Strategic information is currently considered by proactive companies as a real strategic resource, the same as corporate assets and human resources. Business intelligence systems created for this purpose have become a real decision support system to cope with information that informs about unexpected events and turbulent changes, also referred to as "weak signals", and transform it to actionable knowledge. Business intelligence is the informational process through which a company keeps itself aware of opportunities and anticipates threats which occur in its socio-economic environment. This paper aims to describe: (i) the concepts of weak signal and business intelligence, (ii) the process that eases weak signal management, and (iii) a number of issues (24) that may inhibit the appropriation of the concept by practitioners. An instrument was developed and validated in terms of reliability and validity in a pilot phase. Data collected from 194 Kuwaiti executives revealed the ten most important issues facing the respondents. These problems are related to six phases of business intelligence: assessing the need for business intelligence and weak signals, targeting efforts, tracking weak signals, selection of weak signals, rooting and communicating weak signals, and processing of collected pieces of information. Results also reveal a significant relationship between three problems related to the selection phase and availability of R&D activities in the sampled organizations. This paper discusses these issues as well as their implications for practitioners Key-words Strategic management, business intelligence, environmental scanning, weak signals, anticipatory information, problems of scanning activities, training, information quality. ١

1 Introduction Studies in the well developed countries have theorized that Early Warning Systems can be quite successful in helping managers to identify organization's environment strengths and weaknesses. These systems intend to identify the so-called weak signals at an early stage and try to interpret the information to ensure a preventive quality assurance. These systems are also able to provide support to a manager s need to be aware of external changes which leads to threats or opportunities. When organizations in less developed countries (LDC) try to adopt such systems, they may face several issues. We argue in this paper that several managerial and organizational issues must be first identified and solved before the introduction of these systems into organizations. This paper is a first step toward understanding the executives' requirements for an early warning system in LDC before the development of such systems. By learning from the best practices in Western, developed countries, organizations in LDC are able to imitate them with positive results. Learning is therefore becoming the most indispensable activity in the current knowledge-based economy. In the current turbulent environment, in order to compete and survive, firms must be constantly alert, capable of adapting to fast changes constantly learn, evolve, and transform themselves rapidly. An effective strategic response to environmental changes requires a clear perception of events and trends in organization's external environment. Literature in the West labels these perceptions through a process named environmental scanning and business intelligence (Rouibah and Ould-Ali 2002) While several studies investigated business intelligence, environmental scanning and competitive intelligence in well developed countries, little research has been conducted in LDC. For example, in the West, the topic was studied in several countries including France (Lesca 1994), the USA (El Sawy 1995; Beal 2000), Canada (Auster and Choo 1994) and the UK (Xu et al., 2003), Israel (Fiegenbaum and Latvia (2000), Australia (Hall 2001), Korea (Ghoshal 1988), Greece (Kourteli 2000), and Germany (Voigt and Czaja 2007). However, very few studies were done in LDC including Nigeria (Sawyerr et al. 2000), Algeria (Rouibah and Bessam 2001), Bulgaria (Elenkov 1997), Tunisia (Chouk-Kamoun and Salles 1998), Thailand (Ngamkroeckjoti and Johri 2000), and South Africa (Du Toit 2003). While these studies focused on different issues (see Table 1), the current study aims to investigate the following research question: (1) What are the major perceived problems encountered by executives when translating theory of business intelligence into practice? (2) What is the level of business intelligence awareness of this concept in Kuwait? This paper focused on the business intelligence perception in Kuwait. This study was conducted with the potential differences may exist between perceptions in the West and those in LDC in mind. The remainder of this paper is structured as follows: 1) it discusses the theory of business intelligence and the problem that may be raised when translating theory into practices; 2) it describes the used research methodology, following by the description of the study's results, and 3) it summarizes the results and points to potential implications for practice. 2 Theory and background 2.1 From environmental scanning to business intelligence Strategic management is defined as the art, science and craft of formulating, implementing and evaluating cross-functional decisions that will enable an organization to achieve its objectives (Davis 1989). Strategic management process can be divided into two phases (Wang & Turban 1991): 1) scanning information from the external and internal environment and the interpretation of such information the outcome of which feeds the strategic decision-making and implementation phase; and 2) utilizing four basic activities within the strategic decision-making and implementation phase: strategic formulation, corporate capability planning which attempts to support new ٢

strategies, real-time strategic response to various surprises in the environment, and implementation of strategies. Environmental scanning has its origins in publications on strategic management. It is explicitly recognized as a starting point and vital phase in the strategic management process (Aguilar 1967). It refers to the acquisition of information about events, trends and relationships in an organization's environment, the knowledge which will be of assistance to decision makers (Aguilar 1967). This concept calls for internal and external diagnosis in order to evaluate strengths and weakness as well as opportunities and threats in the organization's environment. External environment Internal environment Phase I: scanning & Interpretation of information Intelligence Phase II: strategic Decision making Strategic decision making Strategy Formulation (planning) Corporate capability Planning Real time response Implementation Figure 1. The relationship between the strategic management process and environmental scanning adapted from Wang and Turban (1991) While environmental scanning focuses on all types of information useful to assessment of SWOT (strengths, weaknesses, opportunities and threats), this paper focuses only one type of information named as "weak signals", which constitute the core of business intelligence with the purpose of creating an Early Warning System. Unlike environmental scanning, business intelligence is an approach that is oriented toward gathering intelligence from the external environment of an organization.. It is defined as a strategic approach for systematically targeting, tracking, communicating and transforming relevant weak signals into actionable information on which strategic decision-making is based (Rouibah and Ould-Ali 2002). Business intelligence aims to link an organization to its external environment in a way that will assure the organization's continued success and make it secure from surprises and unexpected events. It also tends to help a company exploit information elements (in its environment) to grasp opportunities and avoid surprises when discontinuities occur. Business intelligence includes two modes of scanning: reactive (also called problematic search ) and proactive. Reactive scanning corresponds to an oriented and well defined mode of information search. The task of collecting information is limited, rational and well organized. The proactive mode is a continuous search where the scope of scanning is very large, information is little targeted, information is anticipatory and ambiguous, and tasks of scanning are complex and ill structured. This is why target information has been termed weak signals. Early warning systems are also involved in this process. These systems aim to assist an organization achieve its competitive advantage through better weak signals management. The emphasis here is on information systems that enhance strategic decision-making and that support the competitive strategy of an organization (Wiseman 1988) ٣

2.2 The nature of Weak Signals and early warning systems The main intention of Early Warning Systems is to anticipate and to detect hints of possible changes in a company s environment to maximize options for reaction. In 1967, AGUILAR published the first dissertation focussing on this topic from the point of view of business. Even if AGUILAR claims for a systematic and continuous monitoring of companies environment and involving this gained information in the decision-making process, he offered no way for implementation. Eight years later, in 1975, ANSOFF developed his concept of weak signals. Based on this concept, ANSOFF enhances the so-called Strategic Issue Management, which is very similar to the concept of weak signals. Ansoff was the first author who called senior mangers to keep track of strategic information that may create discontinuities in the company strategies and to initiate further probing based on such detection (Walls et al. 1992). He called this kind of information "weak signals" to denote information that may feed into an unexpected decision or that may lead to opportunities or avoid unexpected threats (El Sawy 1985). Weak signals were defined as uncertain and fragmented information about developments and trends. These have not been completely realized, or they have potential consequences, or are perceived to have a significant impact on organizational performance, either as threats or opportunities (Rouibah & Ould-Ali). A weak signal is a piece of information in progress and requires much more effort to be caught. Weak signals are different than strong signals, in that a strong signal is information that informs about some thing already happened, or confirms a decision already taken. Consequently, firms have no response delay. This is the case when date for a request for quote (bidding) is over. Weak signal management can be rooted in the resource-based theory. Its view of the business posits that businesses compete on the basis of unique corporate resources that are valuable, rare, difficult to imitate, and non-substitutable by other resources (Conner 1991; Schulze 1992). It includes tangible, intangible, and personnel-based resources. The resource-based view illustrated that businesses can differentiate themselves on the basis of weak signal management. The field of research of Early Warning and Early Warning Systems deals with the identification of business relevant developments and changes at an early stage. We distinguish a difference between Early Warning and Early Detection. While the wording of Early Warning just signals possible hazards, Early Detection means that there are also chances, but the company has the "will" or the motivation to identify possible dangers. However, the pure identification of chances and risks is not enough; it also is necessary to develop collect, interpret, and proposed appropriate actions/strategies to face chances and risks. Management of weak signals is recognized as an important research topic for strategic decisionmaking (El Sawy and Pauchant 1988; Martinsons 1994; Freeman 1999). Accordingly, several experts in the strategic management discipline (e.g. Martinsons 1994, Attaway 1998; Rouibah and Ould-Ali 2002) built different models to formalize the business intelligence as well as to transform the weak signal management into feasible processes. In this paper, we defend the idea that the business intelligence approach should be a continuous and proactive so that a company may be an exploiter/innovator (Mair et al., 1997). The company should collect weak signals that are essential for its survival and that have an impact on the company's performance and informs about events not yet realized. In the strategic management literature, theorists consider the business intelligence role for a business as a radar for a ship. It aims to inform managers early enough about interesting opportunities and threats so that they can cope with them and propose adequate answers. Such a strategy will help to keep a sustainable competitive advantage. The particular characteristics of weak signals make them very difficult to manage. According to Rouibah & Ould Ali (2002) a weak signal is: ٤

- anticipatory; it informs about changes when they begin to occur, - uncertain; it concerns events that have not yet been realized, - ambiguous; its content is usually uncertain or could be deliberately contaminated or distorted (for example, by a competitor), - fragmentary; each information element taken alone is insignificant; however, its significance gradually increases when combined with other weak signs, - dynamic; it evolves over time, - cyclical; it has a complex life cycle from growing to declining that varies in its duration and significance, and - qualitative; in most cases, numbers are not involved and information may be available in other forms such as written, verbal, or visual images. Weak signs are thus subject to perception and interpretation, and multiple meanings are possible. 2.3 Business intelligence process The assimilation of business intelligence is quite similar to knowledge management in several ways. Weak signals are tacit knowledge since they represent anticipative pieces of information in the mind of gatekeepers who believe they might be useful to senior managers. Moreover, each piece of information is insignificant unless it is interpreted and transformed into actionable information. Such activity is completed when several viewpoints are combined and opposed. With regard to weak signal management approach, companies may adopt two different policies. The first is an offensive strategy via a proactive attitude in that it seizes opportunities by having a focus on events that have not yet taken place and that are announced by weak signals. The second policy consists of adopting a defensive or reactive behavior. This is evidence when a company waits until the best company grasps the opportunity, and then it reacts after tangible benefits are achieved. Adopting one of the two previous strategies besides the particular characteristics of weak signals make their management somewhat difficult. The translation of theories on business intelligence to actionable knowledge useful for decision making requires the existence of a process model. While several processes were proposed in the past, this paper emphasizes the model developed by Rouibah and Ould-Ali (2002), see Figure 1. This process aims to translate weak signals into early warnings that keep management informed of what's looming ahead in the competitive environment. This process is oriented to ease weak signal management. This process requires several activities: (i) evaluation of the process relevancy for a kind company; (ii) the delimitation of scanning areas; (iii) the collection of target information; (iv) the selection of relevant information; (v) the routing and diffusion of collected information; (vi) the interpretation of collected pieces of information; and (vii) the integration of processed information in the decision-making process. ٥

Selection 3 Routing 4 Tracking 2 2 5 Processing interpretation Targeting 1 Action Evaluation of 6 Relevancy 0 Figure 2. Business intelligence process (Rouibah & Ould-Ali 2002) When executives try to translate Rouibah and Ould-Ali's (2002) model into practice, they may encounter several obstacles. The following sections will review these problems. 2.4 Obstacles to business intelligence In trying to review past studies on the subject, we categorized the literature into two groups: studies based on empirical studies versus studies based on a conceptual problem view. The first category includes both empirical studies and studies that only mentioned some problems related to business intelligence, scanning and competitive intelligence, and strategic management. The second category refers to studies that concentrate on the conceptualization of a specific problem related to weak signals and then try to propose guideline, methods and methodologies to overcome these problems. Identified problems were categorized according to Figure 1 that depicts phases of business intelligences process. ٦

Activities of business intelligence process Phase 1: Relevancy of weak signal management Phase 2: Targeting of weak signals Phase3; Tracking of weak signals Problems/ obstacles of weak signals management 1. Inability to perceive potential benefits of weak signals on the company 's performance 2. Difficulty shaping the meaning of several concepts which refer to "weak signals", "business intelligence" and managers do not have a clear vision about their content and usefulness 3. Company culture is not oriented toward usage of weak signals 4. Difficulties identifying the company's requirements in terms of scanning activities 5. Lack of knowledge in on how to start scanning activities 6. Uncertainty regarding governmental long-term policies 7. Lack of qualified gatekeepers to conduct scanning activities Support literature Empiricalbased. Lesca (1994); Groom & David (2001); Rouibah (2003); Voigt & Czaja (2007) El Sawy (1985) Beal (2000); Du Toit (2003) Lesca (1994); Rouibah (2003) Voigt & Czaja (2007) Lesca (1994); Rouibah & Bessam (2001) Freeman (1999) Ghoshal & Westney (1991); Lesca (1994); Yasai-Ardekani & Nystrom (1996); Elenkov (1997); Ngamkroeckjoti & Johri (2000); Sawyerr et al., (2000); Rouibah & Bessam (2001); Du Toit (2003); Xu et al., (2003) Auster & Choo (1994); Du Toit (2003); Lesca (1994); Rouibah (2003) Sawyerr et al., (2000); Beal Support literature Processbased Wang & Turban, (1991); Lesca (1994); Rouibah & Lesca (1996) Ansoff (1975) Rouibah & Lesca (1996) N.A Martinsons (1994); Attaway (1998) Aguilar (1967); Gelb et al., (1991); Rouibah & Lesca (1996); Attaway (1998) N.A. (2000); Elenkov (1997) Lesca (1994) Wang & Turban (2001); Martinsons (1994); Attaway (1998); Rouibah & Lesca (1996) 8. Problem of legal and ethical practices of scanning N.A. Gelb et al., (1991) activities 9. Inadequate management education and training Sawyerr et al., (2000), Rouibah (2003) Rouibah & Lesca (1996); Rouibah (2000); 10.Lacking of motivation and rewards for company Lesca (1992); Lesca (1994); personnel Rouibah (2003) Wang & Turban (2001); Martinsons (1994); Attaway (1998); Rouibah & Lesca (1996) 11. Lack of time allocated to scanning activities N.A. Rouibah & Lesca (1996) 12. Lack of trust and cooperation among members of the organization Lesca (1994); Rouibah (2003) Rouibah & Lesca (1996); Attaway (1998); Wang & Turban (2001) 13. Lack of support from top management Lesca (1994) Martinsons (1994); Wang & ٧

Phase4: Selection of weak signals Phase5; Rooting & communication Phase6: Processing & interpretation of weak signals Phase 7: Action "relationship weak signals/ decision making" Phase8: Evaluation of current practices of weak signal management. 14. Frequent changes in the market resulting in a lot of information and there are no methods to identify and select relevant information 15. Difficulty analyzing the quality of information sources Turban (1991) Lesca (1994) Gelb et al., (1991) Sawyerr et al., (2000); Rouibah N.A (2003) 16. Poor quality of collected information N.A Wang & Turban (1991) 17. Lack of data sources where weak signals can be found Auster & Choo (1994); Sawyerr et al., (2000); Aguilar (1967); Martinsons (1994) 18. Data gathered is not useful since it is used to Lesca (1994) Gelb et al., (1991); Rouibah & confirm only decisions already made and not new Rouibah (1996) decisions 19. Lack of a formal process to root out & share collected information 20. Employees' resistance to business intelligence and scanning culture 21. Inability to analyze collected information and to generate actionable information 22. Collected data are not often presented in an adequate format for decision making function 23. No integration of collected information in the decision making process 24. Difficulties evaluating current practices of weak signal management and scanning activities Lesca (1994); Du Toit (2003); Rouibah & Bessam (2001) Rouibah & Bessam (2001); Rouibah (2003) Lesca (1994); Rouibah & Ould-Ali (2002); Rouibah (2003); Freeman (1999) Kourteli (2000); Rouibah & Lesca (1996); Attaway (1998) Wang & Turban (2001) Gelb et al., (1991); Rouibah & Lesca (1996); Attaway (1998); Rouibah & Ould-Ali (2002); Rouibah (2000); El Sawy & Pauchant (1988) Lesca (1992) Gelb et al., (1991) Lesca (1992); Groom & David (2001) Ghoshal & Westney (1991); Lesca (1994); Yasai-Ardekani & Nystrom (1996); Rouibah & Bessam (2001); Du Toit (2003) Gelb et al., (1991) Aguilar (1967); Martinsons (1994); Rouibah & Lesca (1996); Chouk- Kamoun & Salles (1998) Table 1. Summary of literature review on business intelligence, environmental scanning and competitive intelligence ٨

In the Arab region, few studies have focused on business intelligence. Chouk-Kamoun and Salles (1998) studied the application of business intelligence in 24 Tunisian SMEs. They found that executives don t perceive the value of weak signals as a factor that contributes to company performance. Such a result suggests that the majority of small companies did not perceive benefits from business intelligence. Rouibah and Bessam (2001) interviewed 12 Algerian SMEs about practices of environmental scanning. Results showed scanning activities lack efficiency because employees lack the knowledge to: (i) evaluate current practices, (ii) set up an efficient environmental scanning process, and (iii) share collected information. Rouibah (2003) studied the awareness of environmental scanning by executives within 86 Kuwaiti firms. In particular, the author studied which types of information (control, influence or weak signals) receive the most attention of executives. Rouibah suggested propagating the culture of "word to mouth", attention to "weak signals" among employees as well as the culture of information sharing. In addition, the author examined five problems faced by executives. Results revealed that the inability to shape the meaning of environmental scanning was the most perceived problem, followed by the confusion in the exiting of different concepts used to refer to environmental scanning. Other researchers studied problems and challenges facing Arab managers in the 19 th of the 20 th century, which inhibited the practice of modern management styles (modern development) (Atiyyah 1993). These include findings that included 1) the turbulent political environment is unfavorable to the development of modern development; 2) unethical practices of business activities; 3) managers' concentration on meeting their current needs rather than satisfying longterm objectives; 4) a shortage of time allocated to planning; 5) the resistance of managers to changes (existence of doubtful managers who don't believe on trainings effectiveness, and 6) attendance in training not for the sake to improve their skills but to earn a cash bonus or qualify for a promotion. He asserted that one of the most critical issues facing Arab managers is management development and saw a strong link between its success and economic development. Based on the previous literature review, we can observe the following remarks. First, past literature has discussed several obstacles, but none of the past studies integrated these problems in one empirical study nor validated these problems either in the West or in LDC. Second, many past studies have focused on what companies should do to solve the issues of business intelligence, rather than to look at what firms actually do. Much prior literature is devoted to prescriptions or techniques. Moreover, this study also reviewed papers that were published in Competitive Intelligence Review between 1997 and 2007, and of the 128 published papers only three have had any solid empirical research content. Third, most empirical studies on actual practices have tended to focus on the leading firms in well developed (Western) countries. By contrast, this study looks more at how firms in LDC carry out business intelligence and thus gives a better guide as to the general level of acceptance and use of this approach. Fourth, the study looks at actual business intelligence across a range of different people in the organization. The instrument-based questionnaire used in the study is completed by people selected from different management levels (senior, middle, and operational level) and thus gives a broader organizational and cultural perspective on the role and current perceptions of business intelligence in firms operating in LDC. ٩

Fifth, earlier studies treated the business environment as a single entity (except Xu et al., 2003), while this study is based on a multi-industry approach. For the purpose of this study, the author has summarized 24 issues that are well grounded in past literature (see Table 1). The current research is an attempt to test their validity in the Kuwaiti business environment. 3 Research methodology Since our interest focuses centers on clarifying the problems associated with business intelligence from a Kuwaiti executives point of view, the survey method, based on a questionnaire, was chosen as being most appropriate. The designed questionnaire includes three sections. The first one records data related to company profile (size, annual total revenue, number of total employees, number of employees involved in scanning activities). The second section highlights information related to respondents (department, position, and time respondent has been with the company). The third section includes major obstacles Kuwaiti executives may encounter. 3.1 Sample and procedure In an attempt to test the validity of the 24 problems (see Table 1), this study targets Kuwaiti executive as a unit for analysis. We followed a procedure involving three steps to collect the data and to approach the companies. First, we have chosen to contact companies through personal visits, using personal networks, the telephone and, in some cases, e-mail to solicit their participation. Second, the executive is identified and an appointment was made to visit him. During the appointment, the executive was delivered a letter that describes the research objectives and the questionnaire. Third, each participant was briefed on the purpose, content, and procedure for completing the questionnaire and provided with a copy of the instrument. A letter was also attached to the questionnaire that explains the main objective of the study and includes definition of key concepts (such as business intelligence, weak signals, etc.). Fourth, another appointment was made to pick up the questionnaire. The instrument was written in Arabic and a translated copy into English was also made available. It was also checked by two faculty members at the College of Business Administration. The instrument required respondents to provide information based on their perceptions of what was currently in practice in their organization. Of the 250 questionnaires distributed, 194 were retuned, giving a response rate of 77.6%. The following table provides the demographic data about the respondents. Position of respondents Frequencies Percentage Operational level 15 7.73 Middle Management 113 58.24 Strategic level 66 34.01 Total 194 100 Time respondent has been with the company Frequencies Percentage Less than 1 year 11 5.68 1-5 years 72 37.11 6-10 years 52 26.8 11-15 years 26 13.4 More than 15 years 33 17.01 Total 194 100 Table 2. Demographic data and Profile of respondents 34.7 % of participants in this study are free businesses (i.e. owned by a single family) followed by sales and marketing department (26.8%), while people belonging to the information system department are not involved in scanning activities. In addition, the majority of respondents belong to management level followed by the strategic level which depicts the importance of scanning ١٠

activities in Kuwait. 37.1% of respondents have been with their company between 1 and 5 years. 26.8% between 6 and 10 years. The profile of the respondent provides credibility and validity of the business intelligence practices in Kuwait. 3.2 Profile of companies Table 3 provides the profile of companies involved in the survey. Nature of company Frequencies Percentage Public 34 17.65 Private 160 82.47 Total 194 100 Age of company Frequencies Percentage Less than 5 years 25 13.22 5.1-10 years 21 10.35 10.1-20 years 17 8.62 More than 20 years 131 67.81 Total 194 100 Annual turnover (in million $) Frequencies Percentage Less than 1.6 17 8.88 1.6-3.3 14 7.4 3.3-16.5 45 22.96 16.5-33 70 36.29 33-108.9 24 12.59 109-165 7 2.96 Above 165 17 8.88 Total 194 100 Number of employees in the organization Frequencies Percentage Less than 100 59 30.41 101-499 40 20.61 More than 500 95 48.96 Total 194 100 Existence of R&D activities Frequencies % No 62 31.9 Yes 132 68.1 Total 194 100 Percentage of employees involved in scanning Frequencies Percentage 1-5 % 100 51.54 6-20 % 60 30.9 More than 20 % 34 17.57 Total 194 100 Table 3. Characteristics of the companies and practices of scanning Most of companies involved in scanning are in the finance, banking, and insurance sector (25.3%) followed by trading and retailers (16%), and by construction (15.3%). The majority of companies belong to the private sector (82.47%), while a small percentage (17.65%) belong to the public sector. 67.81 % of the companies have been in existence more than 20 years, while 13.22% have been in existence for less than 5 years. More than half of respondents (59.25%) work in companies with an annual turnover comprised between $3.3. millions and $33 millions. 51.04 % of companies ١١

are small and medium companies (SME), while 48.96% are large firms. Among the sampled companies, 68.1% have research & development (R&D) activities and 82.5% have less than one quarter of their employees ( 20%) involved in scanning activities. 3.3 Measures Each of the 24 obstacles to the efficient practice of business intelligence, identified from the past literature (see table 1), was measured using a five-point scale: strongly agreed (1), agreed (2), neither agreed nor disagree (3), disagree (4), and strongly disagree (5). The respondents were asked to indicate on a scale whether each of the factors had been an obstacle to efforts to collect, receive and disseminate the information from the environment. 3.4 Reliability and validity of the instrument Before data analysis, content validity and reliability of constructs were examined. To ensure content validity, a thorough examination was made of the relevant literature. Before questionnaire distribution, and to further reduce the possibility of non-random errors, a pilot study was conducted to examine the questionnaire for validity (measuring what is intended), completeness (including all relevant variable items), and readability/understandability. A pilot phase was conducted with 20 executives and the authors. This was done in order to test the applicability of such instrument. The results of the pilot study suggested several changes to the questions. These changes were incorporated in the instrument. Data analysis & results 3.5 Obstacle to business intelligence Table 4 provides the survey percentages responses. It provides the analysis of obstacles executives are facing (1 strongly agree to 5 strongly disagree) ١٢

[1] Strongly Agree (%) [2] Agree (%) Total [1] + [2] [3] Neither Agree Nor Disagree (%) [4] Disagree (%) [5] Strongly Disagree (%) Problem of scanning 1. Difficulty shaping the meaning of several concepts which refer to "weak signals", "business intelligence" and managers do not have a clear vision about their content and usefulness 28 42.5 70.50 19.4 9.5 0.5 2. Difficulties identifying the company's requirements in terms of scanning activities 17.6 48.4 66.00 19.7 12.8 1.6 3. Lack of knowledge in how to start scanning activities 25.7 40.1 65.80 20.9 12.3 1.1 4. Inadequate management education and training 25.1 40.1 65.20 16.6 12.3 5.9 5. Difficulty analyzing the quality of information sources 19.7 44.1 63.80 14.4 18.6 3.2 6. Frequent changes in the market resulting in a lot of information and there are no methods to identify and select relevant information 16.6 42.8 59.40 20.3 18.2 2.1 7. Lack of data sources where weak signals can be found 24.2 34.2 58.40 15.3 23.2 3.2 8. Poor quality of collected information 17.2 37.1 54.30 21.5 21.5 2.7 9. Lack of a formal process to root out & share information 16 36.9 52.90 26.2 18.2 2.7 10. Collected data are not often presented in an effective format for planning function 14 38.2 52.20 21 23.7 3.2 11. Employees' resistance to business intelligence and scanning culture 11.5 37.7 49.20 26.8 17.5 6.6 12. Difficulties evaluating current practices of weak signal management and scanning activities 9.8 38.3 48.10 27.9 21.3 2.7 13. Lack of support from top management 19.3 28.3 47.60 21.4 22.5 8.6 14. No integration of collected information in the decision making process 14 32.8 46.80 19.4 26.9 7 15. Uncertainty regarding governmental long-term policies 16.8 29.6 46.40 35.2 15.6 2.8 16. Inability to perceive potential benefits of weak signals on the company 's performance 10.3 35.9 46.20 17.9 30.4 5.4 17. Lack of motivation and rewards 20.2 25.5 45.70 16 26.1 12.2 18. Lack of trust and cooperation among members of the organization 22.7 22.2 44.90 11.9 29.2 14 19. Inability to analyze information and to generate actionable information 15.2 28.8 44.00 25 26.1 4.9 20. Problem of legal and ethical practices of scanning activities 14.3 28.6 42.90 16.2 30.8 10 21. Lack of time allocated to scanning activities 5.6 36.1 41.70 32.2 22.2 3.9 22. Company culture is not oriented toward usage of weak signals 15.6 25.3 40.90 22.6 26.3 10.2 23. Data gathered is not useful since it is used to confirm only decisions already made and not new ones 9.7 28.6 38.30 24.9 31.4 5.4 24. Lack of qualified gatekeepers to conduct scanning activities 15.9 21.7 37.60 16.4 32.8 13.2 Table 4. Means, standard deviation, and internal reliability of obstacles facing executives to perform business intelligence ١٣

The above table highlights the most critical problems with "Strongly Agree" and "Agree" perceptions. It can be seen that that among the 24 problems, only 10 scored more than 50%. Such a result means that the total score of those respondents who "Strongly Agree" and "Agree" is bigger than those who were "Neither Agree & Nor Disagree", "Disagree" and "Strongly Disagree". These ten problems belong to the six phases of business intelligence. Among the three hypothesized problems related to the first phase of business intelligence, results revealed that only one is valid and ranked the most crucial problem (70.5%). It is related to the relevancy of weak signals (phase 1). Respondents have difficulties shaping the meaning of several concepts which refer to "weak signals", "business intelligence" and managers have no clear vision about their content and usefulness. This issue is supported by previous studies (Lesca 1994; Rouibah 2003; Voigt & Czaja 2007). Assessing of the relevancy and value of the business intelligence approach and weak signal is a prerequisite and first step toward efficient practice. Business intelligence may best work for company X and not for company Y. With regard to the second phase "targeting of weak signals, two problems out of five are valid. Results revealed that respondents perceived difficulties identifying the company's requirements in terms of scanning activities, i.e. how to delimit the scanning areas, (66%), and lack of knowledge in how to start scanning activities (65.8%). Scanning all areas of the company's environment is complex to achieve. It needs a careful narrowing and creation of targets. Several empirical studies support the second problem (Ghoshal & Westney 1991); Lesca 1994; Yasai-Ardekani & Nystrom 1996; Elenkov 1997; Ngamkroeckjoti & Johri 2000; Sawyerr et al., 2000; Rouibah & Bessam 2001; Du Toit 2003; and Xu et al., 2003), and the third problem (Auster & Choo 1994; Du Toit 2003; Lesca 1994; Rouibah 2003). Other studies did propose some approaches in how to facilitate companies in carrying out targeting activities (see Martinsons 1994; Rouibah & Lesca 1996; and Attaway 1998). Among the five hypothesized problems of the third phase "Tracking of weak signals", only one is perceived to be valid (scored 65.2%). Inadequate management education & training is ranked the fourth problem. Once business intelligence is viewed as important for the company, and once the targets are identified, tasks for gatekeepers need to be identified. These consist of tracking and collecting weak signals from a number of information sources (formal vs. informal). Previous studies offered support for this result (Sawyerr et al., 2000; and Rouibah 2003) while other studies proposed a framework on how to help achieve tracking of weak signals (Rouibah & Lesca 1996; Rouibah 2000). Four among five hypothesized problems related to the forth phase of business intelligence, "Selection of weak signals" is valid. Respondents perceived: Difficulty analyzing the quality of information sources (63.8%); frequent change in the market resulting in a lot of information and there is a lack of methods to identify and select relevant information (59.4%), lack of data sources where weak signals can be found (58.4%), and poor quality of collected information (52.9%). The selection of weak signals is a complex task. First, it is hard to analyze the quality of the information source where a relevant piece information can be found. Nowadays, managers are overwhelmed by the quantity of information they receive. They lack methods that help them to select the crucial weak signals. Moreover, not all information sources are reliable, and not all collected pieces of information are weak signals or reliable. Several studies confirmed these problems: difficulty to analyze the quality of information sources (Sawyerr et al., 2000; Rouibah 2003); lack of methods to identify and select relevant information (Lesca 1994); lack of data sources where weak signals can be found (Auster & Choo 1994; Sawyerr et al., 2000); and poor of quality of collected information (Sawyerr et al., 2000; Rouibah 2003). ١٤

Among the two hypothesized problem of the fifth phase "rooting & communication of weak signals", only one was found to be valid. Lack of formal process to root out and share collected pieces of information (52.9%) is ranked the 9 th issue. It is not enough to collect and select weak signals if they are not rooted from the outside to inside of the business. They also need to be communicated to those who may interpret and use them. Without a formal process, weak signals will be lost. Such a problem has been highlighted by a few empirical studies (Lesca 1994; Rouibah & Bessam 2001; Du Toit 2003). With regard to the sixth phase "Processing & interpretation of weak signals", among the two hypothesized problems, only one is valid. Collected data are not often presented in an effective format for planning function (52.2%). Weak signals are pieces of information that might be informative about events in progress. These pieces of information need to be pulled together in order to transform them into actionable information. Thus, they require processing and interpretation. It is worthwhile to mention that not all weak signals will be transformed into strong signals. This is acknowledged in past literature (Lesca 1994; Rouibah & Ould-Ali 2002; Rouibah 2003; Freeman 1999), and some frameworks were proposed to help achieve this task (El Sawy & Pauchant 1988; Gelb et al., 1991; Rouibah & Lesca 1996; Attaway 1998; Rouibah 2000; Rouibah & Ould-Ali 2002). Finally, none of the issues related to the seventh phase " action: relationship weak signals/ decision making" and eighth phase "evaluation of current practices of weak signals" were perceived to be effective issues by respondents. Such results may be explained in that the awareness of business intelligence level is not high. 3.6 Difference in perceived problems In order to test whether there are significant differences in the 10 most perceived problems, we first assessed the normality of all these issues. For this purpose, we examined normality using the Kolmogorov-Smirnov test (Nunnally 1978). The null hypothesis that was tested was that the ten problems are normally distributed. The P-value [0.00] was found to be less than alpha [0.05], indicating that the variables are not normal distributed. Therefore, the Mann Whitney test was selected to test whether there are significant differences among the ten perceived problems based on the following seven variables. 1. Type of company (private, public, joint venture, and mixed companies). 2. Age of companies ( 5; 5-10; 10-20; 20). 3. Annual turnover ( 1.6; 1.6-3.3; 3.3-16.5; 16.5-33; 33-108.9; 109-165; and > 165). 4. Number of employees ( 100; 101-500; and > 500). 5. Percentage of employees involved in scanning (1-5 %; 6-20 %; and more than 20 %). 6. Existence of R&D activities (Yes vs. No). 7. Management level (operational, management, and strategic) The Mann Whitney test shows no statistical significance for the first five variables (type of company, age, annual turnover, number of employees, and percentage of employees involved in scanning) and perceived problems. Table 4 only reports the results for the correlation between the 10 most perceived problems and the existence of R&D activities. Note: ** Significant at p<0.05 and *Significant at p<0.10 Perceived problems Presence of R&D Mean rank P-Value related to environmental scanning 1- Difficulty shaping the Yes 89.59 meaning of several concepts which refer to "weak signals", "business intelligence" and managers do not have a clear vision about their content and No 90.88 0.870 ١٥

usefulness 2- Difficulties identifying Yes 88.77 the company's requirements No 95.73 0.371 in terms of scanning activities 3- Lack of knowledge of Yes 94.16 0.117 how to start scanning activities No 82.99 4- Inadequate management Yes 95.19 0.063 education and training No 80.38 5- Difficulty analyzing the Yes 96.79 0.024** quality of information sources No 79.03 6- Frequent changes in the Yes 94.36 market resulting in a lot of information and there are a lack of methods to identify No 82.58 0.133 and select relevant information 7-Lack of data sources Yes 94.13 where weak signals can be No 82 0.013* found 8- Poor quality of collected Yes 90 0.024** information No 88 9- Lack of formal process Yes 94.07 to root out & share No 81.52 0.145 collected pieces of information 10- Collected data are not Yes 94.5 often presented in an No 81 0.799 effective format for planning function Table 5. Differences in the perceived problems based on presence of R&D The above table reveals statistical significance differences between three problems (difficulty to analyze quality of collected information, lack of data sources where weak signals can be found, and poor quality of collected information) and the availability of R&D scanning activities. Thus, business intelligence is positively related to R&D, meaning that gathering weak signals, assessing their quality and the quality of information source is very important for R&D. This new result was not previously reported. The three problems are related to the selection phase of the business intelligence process and encountered by scanners when there are R&D activities. Note: ** Significant at p<0.05 and *Significant at p<0.10 Perceived problems Management level Mean rank Significance (P) related to environmental scanning 1- Difficulty shaping the Strategic 96.04 meaning of several Management 91.77 0.86 concepts which refer to "weak signals", "business intelligence" and manager Operational 94.92 2- Difficulties identifying Strategic 92.06 the company's requirements Management 97.29 0.63 in terms of scanning activities Operational 86.71 3- Lack of knowledge of Strategic 94.98 0.95 how to start scanning Management 93.02 activities Operational 96.13 4- Inadequate management Strategic 98.57 0.61 education and training Management 86.21 Operational 92.67 ١٦

5. Difficulty analyzing the quality of information sources 6- Frequent changes in market resulting in a lot of pieces of information and there are few methods to identify and select relevant information 7- Lack of data sources where weak signals can be Strategic 102.9 Management 86.65 Operational 110.42 Strategic 102.35 Management 87.33 Operational 103.18 0.13 ١٧ 0.049** Strategic 102.02 0.38 Management 90.82 found Operational 100.50 8- Poor quality of collected Strategic 105.43 0.49** information Management 93.61 Operational 84.91 9- Lack of formal process Strategic 98.90 to root out & share Management 116.06 0.06 collected pieces of Operational 87.34 information 10- Collected data are not Strategic 101.10 0.31 often presented in an Management 88.53 effective format for Operational 95.53 planning function Table 6. Differences in the most perceived problems based on respondents' management level Table 6 reveals significant statistical significance difference between two problems (difficulty to analyze the quality of information sources and poor quality of collected information) and the respondents' management level. These two problems are related to the selection phase of the business intelligence process and means that differences exist only for weak signals selection among the three management levels. CONLCLUSION AND IMPLICATIONS This study examines the perceived problems related to weak signals and business intelligence by a sample of Kuwaiti executives and has several contributions. First, this study filled in gaps in the literature about practices of business intelligence in a developing, Arab country and highlights the perception of this approach in Kuwait. Second, this study shows that the most perceived obstacle to performing effective weak signal management are related to six phases of the business intelligence process (assessment of relevancy, targeting, tracking, selecting, rooting & communicating, and processing). Surprisingly, respondents in our study did perceive problems related to other phases "action: relationship weak signals/ decision making" and "evaluation of current practices of weak signals". The three most perceived problems are related to the assimilation of the concept "business intelligence" and "weak signals" and how scanning activities could be initiated. Finally, results reveal business intelligence is positively related to R&D and offer support to selecting critical information. A number of limitations of the study are recognized. First, we do not claim that this research is other than exploratory since the sample size was small and this limits the ability to generalize the result beyond this sample. Second, we obtained data from a single informant and a convenient sample 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. From a practical view, this study highlights the readiness of Kuwaiti companies to assimilate the concepts of weak signals and business intelligence. It suggests that approach is worthy to be propagated in Kuwait, as the world-wide economy is becoming more dynamic. Businesses are not any more secure from discontinuities, which call for business intelligence to track and interpret

weak signals. This approach requires senior managers to propagate the culture of seeing and hearing in their company. Senior managers are also responsible for initiating specific training in order to increase awareness and assimilation of weak signals and business intelligence concepts. Specifically, training should be first directed toward the assessment and significance of the environmental scanning. Training programs should also directed toward providing help for companies to assimilate how delimiting the scanning efforts and how to start these activities. Such programs will benefits the companies in term of optimizing scanning efforts, and allow achieving competitive advantage by early warning signals management. These training programs could be easily designed by available resources in Kuwait University, rather. The business intelligence process is a complex and ill structured approach, which includes several complex tasks. The input to this process is pieces of information, termed here as weak signals. These pieces are available outside the organization. These pieces of information need employees' motivation since they will not come on their own to the company. Top management support, involvement, and will throughout all the phases are necessary to make this process successful. Two interesting and future research directions are suggested. First, the study suggests continuing to study the relationship between business intelligence and R&D, in order to clarify the needed pieces of information. Second, while this study investigated the perceptions of about the 24 issues, additional research could be directed toward understanding how executives perceive how improvements of business intelligence could be achieved. References Aguilar F. J. (1967), Scanning the business environment, Macmillan, New York. Ansoff H. I. (1975), Managing strategic surprise by response to weak signals, California Management Review, XXVIII n0 2, pp. 21-33. Atiyyah H., (1993), Management development in Arab countries: the challenges of the 1990s, Journal of Management Development, vol. 12, No 1, pp. 3-12. Attaway. M.C., (1998), A Review of issues related to gathering and assessing competitive intelligence, American Business Review, Vol. 16, no1, pp. 125-35. Auster E, and Choo C.W., (1994), How senior managers acquire and use information in environmental scanning, Journal of the American Society of Information Sciences, Vol. 44, no 4, pp.194-203. Beal R. M., (2000), Competing effectively: Environmental scanning, competitive strategy, and organizational performance in small manufacturing firms, Journal of Small Business Management, 38, January, pp. 27-47. Chouk-Kamoun S., and Salles M., (1998), Définition d un dispositif de veille stratégique pour les PME tunisiennes, Colloque, VSST 98 Conner, K. R. (1991) A Historical Comparison of the Resource-Based Theory and Five Schools of Thought Within Industrial Organization Economics: Do I Have a New Theory of the Firm, Journal of Management, vol. 17, no 1,, pp.121-54. Du Toit A.S.A. (2003), Competitive intelligence in the Knowledge economy: What is in it for south African manufacturing enterprise, International Journal if Information Management, no 23, pp.111-120. El Sawy, O., (1985), Personal information systems for strategic scanning in turbulent environments: can the CEO go on-line? MIS Quarterly, March, pp. 53-60. El Sawy, O., A and Pauchant T., (1988). Triggers, templates and twitches in the tracking of emerging strategic issues, Strategic Management Journal, 9, pp. 455-473. Elenkov D. S., (1997), Strategic uncertainty and environmenta1 scanning: the case for institutional influences on scanning behaviour. Strategic Management Journal Vol. 18, No 4, pp. 287-302 Fiegenbaum A. and D. Lavie D., (2000) Strategic Management of MNCs Entry into Foreign Markets: Experience of Israel in the 1990s. European Management Journal Vol. 18, No. 1, pp. 93 105, 2000 Freeman O., (1999), Competitor intelligence: information or intelligence. Business Information Review, Vol. 16, No 2, June, pp. 71-77. Gelb B. D., Saxton M.J., Zinkhani G.M., and Albers N.D. (1991), Competitive intelligence: insight from executives. Business Horizon, January/ February, pp. 43-47. ١٨

Ghoshal S. (1988). Environmental Scanning In Korean Firms: Organizational Isomorphism in action Journal of International Business Studies; Vol. 19, no 1; pp. 69-86. Ghoshal S., and Westney D. (1991), Organizing competitor analysis systems. Strategic Management Journal; Vol. 12, no 1; pp. 17-31. Groom J.R., and David F.R, (2001), Competitive intelligence activity among small firms. SAM Advanced Management Journal, winter, pp.12-20. Hall C., 2001, The Intelligent Puzzle, Competitive Intelligence Review, vol. 12, n o. 4, pp. 3 14. Kourteli L., (2000), Scanning; the business environment: some conceptual issues, Benchmarking; Vol., 7, n05, pp. 406-413. Lesca H. (1992), Le problème crucial de la veille stratégique: la construction du puzzle. Comprendre et gérer-anale des Mines, Avril, pp. 67-71. Lesca H. (1994), Veille stratégique pour le management stratégique état de la question et axes de recherche. In Economies et Societés, série science de gestion, Vol. 5, n 20, pp.31-50. Mair J.L., Rainer R.K., and Snyder C.A. (1997), Environmental scanning for information technology: An empirical investigation. Journal of Management Information Systems, Fall Vol. 14, no2, pp. 177-200 Martinsons M. G. (1994), A strategic vision for managing business intelligence, Information Strategy, Spring, pp.17-33 Ngamkroeckjoti C. and Johri L. M., (2000), Management of environmental scanning processes in large companies in Thailand. Business Process Management Journal, Vol. 6, no. 4, pp. 331-341. Nunnally J.C., (1978) Psychometric theory, 2nd edition, New York, MacGraw-Hill Rouibah K. (2000), Un système d aide au traitement des informations de veille stratégique. Journal of Decision System Vol.9, n 2, pp.215-243. Rouibah K. and Bessam M., (2001), Practices of environmental scanning: the case of Algeria. The 4th Asian Academy of Management (AAM), Malaysia 10-13 November, pp.813-817. Rouibah K. and Lesca H. (1996): Tools for Strategic Probe Aide. 1st Annual International Conference on Industrial Engineering Applications and Practice. Houston, U SA, 4-7 December, pp. 342-349. Rouibah K. and Ould-Ali S., (2002), PUZZLE: a concept and prototype for linking business intelligence to business strategy. Journal of Strategic Information System, vol. 11/2, pp 111-130. Rouibah K., (2003), Environmental scanning, anticipatory information and associated problems: insight from Kuwaiti executives, Communication of the IIMA vol. 3, no1, pp. 47-63. Sawyerr O., Ebrahimi B., and Thibodeaux M., (2000), Executive environmental scanning, Information source utilisation, and firm performance: the case of Nigeria. Journal of Applied Management Studies, Vol. 9, no 1, pp. 95-115. Schulze, W. S. (1992). The Two Resource-Based Models of the Firm: Definitions and Implications for Research. Academy of Management Best Paper Proceedings, 1992. Voigt K., and Czaja L., (2007). The relevance of early warning systems within the top quality management. 8 th International Research Conference on Quality, Innovation and Knowledge Management, Web: www.conferences.monash.org/qik07/, New Delhi, India 11 th -14 th February 2007. Walls, J.G., Widmeyer G.R., and El Sawy, O.A., (1992). Building an information system design theory for vigilant EIS. Information System Research 3 (1), March, pp. 36-59. Wang P., and Turban L., (1991), Filtering strategic environmental information processing using EIS. Proceedings of the 27 th Hawaii IEEE Conference on System Science 3, pp. 126-134. Wiseman C., 1988. Strategic Information Systems. Irwin, Home-Wood, IL. Xu X.M., G.R., Kaye, and Duan Y., (2003), UK executives vision on business environment for information scanning A cross industry study, Information & Management 40, pp. 381 389. Yasai-Ardekani M. and Nystrom, P.C. (1996), Designs for environmental scanning systems: Tests of a contingency theory, Management Science vol. 42, no. 2, pp. 187-204. ١٩