The book "Successful Business Intelligence" is highly recommended and worth a quick trip to the bookstore or an online order. Published in 2008, this unique book by Cindy Howson stands out for several reasons. First and foremost, Howson, despite her extensive experience, writes with humility, avoiding a condescending tone. Instead of adopting a "know-it-all" approach, she acknowledges her mistakes along the journey and shares her knowledge with readers on an equal footing.
What makes this book particularly exceptional is its departure from the predominantly technological discourse that permeates the business intelligence (BI) world. While Howson draws from her over 20 years of experience, she recognizes that BI's success hinges on factors beyond technology. She adeptly lays out and shares insights into these various components in the book.
Another distinguishing feature lies in the sources of information the book relies on, including personal knowledge, case studies from diverse companies, and a comprehensive survey. Howson conducted an extensive survey answered by 513 individuals worldwide, detailing BI activities in their organizations. The results are intelligently analyzed and integrated into the chapters, providing exciting and, in some cases, innovative contexts. Significantly, each statement is attributed to its origin, whether derived from personal experience, case studies, or survey results.
Despite its rich content, the book maintains an easy-to-read style. The introductory chapters cover the basics of BI, which we'll skip in this review and focus instead on the components of success.
Often, the key to success lies in the "what" and the "how."
The book covers the following topics:
Business Factors:
Measuring success
The LOFT effect
Relevance
The relationship between IT and business
Human Organizational Factors:
Management Support
Supporting organizational structure
Culture and promotion of use
Technical Factors:
Quality database
Flexible development
Adapting tools
Appendix: Organizations Case Study
Business Factors:
Measuring success
Let's start with the conclusion: According to the survey results, 68% of respondents characterized the BI activity within their organizations as highly successful. Another 68% deemed the success level moderate, with only 8% perceiving it as a failure. It's essential to note that this response might be biased, as those viewing BI as unsuccessful may be less inclined to participate in the survey. Despite this potential bias, the noteworthy success rates remain relatively moderate.
The measurement of success is not uniform and varies based on factors such as the company's sector and its ownership structure (public, private, governmental, etc.). Typical methods employed to assess the success of BI activity include:
Enhancing business performance (endorsed by 70% of respondents).
Leveraging the benefits of data accessibility (acknowledged by 68%).
Securing support from critical stakeholders (affirmed by 53%).
Evaluating customer perception of activity criticality (highlighted by 50%).
Analyzing Return on Investment (ROI) metrics (cited by 43%).
Monitoring the percentage of active users (recognized by 31%).
Identifying cost savings (mentioned by 31%).
Determining the number of defined users (acknowledged by 17%).
A valuable tip for success in BI activity is to consider implementation for 100% of users, extending beyond organizational boundaries to include suppliers and customers. Additionally, caution is advised against overly standardized ROI metrics. One can draw parallels with the example of a phone and question how much the investment in its existence is recouped. This seemingly absurd question can affect ROI considerations in the BI domain.
The LOFT effect
An in-depth analysis of the BI market has identified four key components consistently present in successful BI organizations:
L - Luck
O - Opportunity
F - Frustration
T - Threat
In every organization deemed successful, only two of these components were identified. Here's a breakdown of each:
Luck – Some argue that luck results from the intersection of honesty and opportunity. Timing plays a crucial role, and many successful individuals attribute luck as a significant factor. The manifestation of luck can take various forms, such as an influential sponsor, globalization, or opportune timing. Regardless, it's crucial to note that the construction of prototypes often leads to the emergence of brilliant ideas, requiring sensitivity to recognize and flexibility to adopt them.
Opportunity – It is highly recommended to be attentive to opportunities that can fuel BI's expansion within a company, create a distinct external business advantage, or tap into a new customer segment. Numerous illustrative examples are provided, including FlightStats, which harnessed collected data to become a key factor for sales agents evaluating aircraft delays and recommending flights. Another example highlights Dow Chemicals, which utilized data to enhance profits through a strategic merger.
Frustration – Initiating a new BI activity should target organizational pain. As these pain points intensify, BI often becomes a priority out of necessity. Employee frustration with unclear decision-making processes and ineffective work methods can create an opening for BI. The book presents an example of a customer service center where representatives were compensated based on activity but struggled to track, understand, and adapt. Once real-time data was provided, employees devised methods and tools to enhance performance and meet their goals.
Threat – In times of company threats, such as bankruptcy or substantial losses, many organizations leverage BI for crisis management and renewed growth. While intentionally inducing frustration or threat for BI success is discouraged, these components become crucial in leveraging BI effectively. Howson addresses the question of whether achieving high-level success in BI is possible without utilizing these components. While no definitive data establishes them as mandatory, these elements are undeniably common among highly successful companies and should not be disregarded.
Relevance
In this chapter on relevance, Howson directs attention towards creating relevance for users by focusing on business results and concrete business opportunities and adopting the user's perspective (WIIFM effect – What’s In It For Me). Recognizing that organizational, departmental, and personal business analyses are far from trivial, as users may struggle to articulate their needs is crucial. While showcasing initial benefits is often straightforward at the departmental level, especially within the finance department, connecting at a personal level requires tapping into individual incentives. For instance:
The desire to excel or outperform colleagues.
The aspiration is to enhance job performance, whether cultivating better customer relationships or increasing customer satisfaction.
Inclination towards working with data, frustration over data deficiencies, and satisfaction with improved data access.
A valuable tip is to avoid exclusive focus on expert knowledge workers and extend efforts throughout the entire organization, encompassing ordinary functionaries. Success with experts alone does not constitute overall success. Additionally, it is advisable to personalize communication as much as possible to prevent individuals from getting overwhelmed by reports that do not align with their specific responsibilities.
The relationship between IT and business
Many respondents in the survey highlighted a significant communication challenge between IT professionals and businesspeople. A stereotype distinguishes the business worker from the IT person (though this is a generalization): the IT person tends to be more solitary, methodical, and risk-averse, relying heavily on email and instant messaging for communication. In contrast, the business worker is often perceived as more social and inclined toward face-to-face meetings. Additionally, the organizational compensation systems for the two roles differ.
Various tools to bridge this communication gap are available (although not all are mandatory):
Employ someone with a hybrid business-IT background to mediate between teams, understanding both business aspects and how technology can serve those needs. These individuals can identify suitable technology and highlight new opportunities for leveraging it.
Recognize and reward BI employees transitioning from the IT world based on business metrics, aligning their thinking more closely with business employees.
Subordinate the BI person originating from IT to a business unit, both organizationally and financially.
Emphasize the importance of the connection between IT and business, proactively working to strengthen this relationship.
Facilitate better mutual understanding between the two groups, acknowledging and addressing inherent gaps.
Define a vision, strategy, and BI activity that aligns with business equivalents.
It's essential to recognize the high importance of this subject. In the survey, this connection was ranked as the number one critical factor for the success of BI activity.
Human Organizational Factors:
Management Support
The assertion that management support is crucial may be a well-worn statement. Still, survey results indicate that the "management support" factor ranks near the top of the list for influencing the success of BI activity. Successful management support can come from various levels of the management spectrum, including the CEO, COO, VP of Marketing, CFO, other VP of Business, and even the CIO. While the impact of the CIO may be lower in many organizations, obtaining support from the VP of Business is crucial if support from another high-ranking VP is unavailable. To secure management support, it is advisable to:
Showcase small successes and communicate tangible business benefits.
Manage expectations carefully.
Establish a connection between BI and addressing organizational frustrations.
Management support contributes to BI success in the following ways:
Cultivating a genuine commitment to the initiative and its implications within the organization.
Articulating the BI vision in alignment with the company's strategy.
Securing budget approval.
Navigating organizational political crises.
Serving as a supreme advisory body for issues unresolved at the BI team or steering committee levels.
It's important to note that, in most cases, management support is not automatic but rather an acquired process that requires investment.
Supporting organizational structure
Initiating influence on the supporting organizational structure at the outset of BI activity is not always feasible. In some cases, it becomes necessary to commence with departmental business activities, only approaching discussions about the appropriate supporting organizational structure after achieving success. There are two fundamental approaches: a central BI team or BI personnel subordinate to the departments they serve. Each method offers distinct advantages:
Departmental BI places greater emphasis on unit needs, while Organizational BI places greater emphasis on the needs of the organization at the top level.
Departmental BI utilizes technology that best serves its needs financially or in terms of capabilities; Organizational BI adheres to uniform standards.
Departmental BI generates short-term successes; Enterprise BI is sustainable for the long term.
Departmental BI is funded from dedicated resources; Enterprise BI budgets from shared resources.
The recommendation, after the initial proof of capabilities, is to operate within a centralized organizational team, except in specific cases:
When the unit functions independently as a profit center.
When there is no synergy in joint activities across business areas.
When the business unit does not leverage the use of shared enterprise resources.
When there is no link between departmental employee indemnification and organization-wide performance.
A noteworthy statistic from the survey indicates success in both configurations, with a higher success rate in the organizational configuration:
Departmental configuration: 11% failures, 72% moderate successes, 16% significant successes.
Organizational configuration: 5% failures; 64% moderate successes; 31% significant successes.
It is often observed that when there is departmental success, attempts by IT to replicate that success across other departments are usually unsuccessful due to cultural change, varying processes, and constraints. In cases of an integrated team, there is a recommendation to identify which components should be maintained organizationally and which are better managed departmentally.
Beyond the considerations mentioned above, it is advisable to establish a steering committee for knowledge management, mainly if the activity is organizational. This committee should guide and audit the activities of an independent center – the Business Intelligence Competency Center. This center includes officials in infrastructure, data, BI delivery, and business experts. It is crucial to remember that BI professionals are highly sought-after employees, necessitating hiring the best without compromise. Additionally, fostering a positive team atmosphere and providing soft rewards contribute to employee retention within the organization over the long term.
Culture and promotion of use
An intriguing aspect of this component is the author and the book's reference. On the one hand, an entire chapter emphasizes the importance of "soft" factors, including cultural elements and marketing activities, underscoring their significant impact. On the other hand, the chapter's title, both in its brief mention and within the book, is somewhat misleading, being called "Other Secrets to Success." This apparent contradiction is not entirely clear, and it might be that the author was concerned it wouldn't make this chapter more palatable for readers. It's also worth noting that the chapter's presentation in the book differs from the summary provided here at the end.
Potential pitfalls in the realm of organizational culture include:
Rejection is based on decision-making relying on intuition and gut instincts rather than data.
Using data and reports to reinforce decisions, seeking confirmation rather than exploration.
Additionally, it is essential to consider that:
People generally resist change.
Data is not always directly relevant to decision-making.
To foster a culture of data-based decision-making, the following recommendations are made:
Understand the gradual model of awareness (for the benefits of BI), knowledge (about how to work with BI), and usage.
Direct focus on business benefits rather than tool capabilities.
Develop key phrases as messages to be ingrained in the organization's employees' memory, facilitating change.
Brand the BI activity with a system name.
Promote awareness through various channels like road shows, ready-made training tools, newsletters, external articles, corporate conferences, portals, integration in regular team meetings, and more.
Prioritize tutorials, emphasizing practical training based on accurate data rather than theoretical training by the supplier company.
Refrain from copying old reports into the new system; adjust them to current needs instead.
Incorporate graphic elements where applicable.
Technical Factors:
Quality database
As indicated by survey results, a quality database stands out as the top technical component for success. Attaining a high level of data quality poses significant challenges, relying heavily on organizational considerations and data and BI ownership. Here are recommendations and tips:
It is advisable to thoroughly assess the level of data quality at the project's outset, understanding its role and significance in the overall solution.
When encountering invalid data, opting for a short-term solution of creating fixes at the BI level without addressing root causes is suboptimal. It leads to an ongoing cycle of similar compromises.
Understanding the roles of various partners influencing data quality, including those in operational processes, is crucial for knowing whom to contact in specific situations.
Only information collected should be reported and presented consistently and accurately.
Attaining a state of control and the ability to guide the backline on necessary repairs is essential, with assistance from the relevant sponsoring management body.
Defining uniform terminology within the BI framework is necessary to ensure a single truth.
The data warehouse update rate should be determined based on added business values associated with the update rate.
Continuous efforts should be directed toward improving data and its quality.
Acknowledging that achieving immaculate data is unlikely is crucial. Knowing where to set the quality demand threshold is also crucial; otherwise, BI activity will be perpetually hindered.
Flexible development
Flexible development stands out as a less-known yet crucial factor contributing to the success of BI, derived from the analysis of standard components in organizations that have excelled in BI activities. Often omitted in surveys and BI-related study courses, this topic offers valuable insights. Recommendations include:
Given the dynamic nature of these requirements, it is important to establish a BI environment that is adaptable to changes in business needs and focus.
Adopting the spiral model over the waterfall model for initial construction allows for gradual development. Focus on an overarching definition of requirements that evolves as the project progresses.
Conducting precise analysis for each component, determining where adjustments are necessary (e.g., data, standard logical definitions, specific reports).
Collaborating with users through prototype development.
Prioritizing the main goal for users, welcoming change requests that demonstrate a better understanding and potential for added business value while adhering to budget and time constraints management rules.
Ensuring close collaboration between Business and IT.
An analysis of BI project management reveals that 44% of projects are delivered late, and 37% exceed the budget. Projects managed with flexible development demonstrate higher success rates in both aspects. A balance is crucial across three components: differentiation, resources, and time to ensure effective BI project management.
In cases where change requests impact any of these components, decisions must be made on trade-offs to prevent deviations. A tip is to conduct weekly management meetings involving users to ensure alignment on demarcation, budget, and time. This practice can contribute significantly to project success.
Adapting tools
Adapting tools may seem like a task exclusive to IT professionals, but based on insights from organizations interviewed by Howson, she recommends involving all relevant parties in the selection process. Opting for an isolated approach without collaboration often leads to rejection due to fear, defensiveness, or other organizational or political reasons. It's crucial to note that while standardization in tools is essential, it should be distinct from assuming that one tool fits all needs. Understanding specific requirements and tailoring various standard solutions accordingly is necessary. Using BI-specific applications instead of ready-made tools raises questions, with survey results indicating more failures associated with unique applications; Howson advocates using them as supplementary rather than central solutions.
Different types of solutions suit various audiences, as outlined below:
Statistical tools and report production tools: for IT developers.
BI sheets, OLAP, and business polling: for analysts and expert knowledge workers.
Scorecards, regular interactive reports, and dashboards for administrators.
BI search, integrated BI in operational systems: for front-line workers.
Ready-made reports for customers, suppliers, and regulators.
Tool suitability for user types depends on several parameters, including the percentage/number of users per group, frequency and nature of decision-making, forecast of information needs, analysis component of the job, data literacy, familiarity with the source system, technical literacy, and level of travel. Understanding these parameters and the existing level of use can indicate untapped potential.
Howson delves into discussing BI integration in Office, addressing risks associated with Excel, and sharing trends that extend integration to the Word and PowerPoint environment. Howson advises against switching tools with each new technology regarding products and suppliers. Instead, patience is recommended, and changing suppliers is justified if a product becomes an unused tool on the shelf, hindering success due to limitations in capabilities or user interface.
In conclusion, good luck! Applying the tips provided in the book can make the process less challenging.
Appendix: Organizations Case Study
Norway Post:
The activity was initiated in 1995 and refocused in 2001.
Key sponsors: CFO and CIO.
Users: 2,880 (15%).
BI Tools: Hyperon, SAS.
Continental Airlines:
The activity commenced in 1998.
Key sponsor: CIO.
Users: 1,400 + 24,000 (integrated systems) (57%).
BI Tools: Hyperon, SAS, dedicated tools.
Corporate Express:
The activity in 2000 and resumed in 2004.
Key Sponsor: CFO.
Users: 3,000 (34%) and 10,000 customers.
BI Tools: Microstrategy, SPSS.
FlightStats:
The activity was initiated in 2001.
Key sponsor: CEO.
Users: 15 internals – with 1,000,000 consumers per month.
BI Tool: JasperReports.
Medical Associates:
The activity started in 1999.
Key Sponsor: Chief Technology Officer.
Users: 220 (internal and external).
BI Tool: BusinessObjects.
1-800 Contacts:
The activity began in 2004.
Key Sponsor: CFO.
Users: 400 (60%).
BI Tool: Microsoft.
Dow Chemical:
The activity began in 1993.
Key sponsor: CIO.
Users: 12,000 (28%) and 10,000 customers.
BI Tools: Cognos, SAS.
Comments