Analytical Organization Questions Response

ORDER A PLAGIARISM FREE PAPER NOW

DB Hightower – The analytical organization DB pr Ma CO od ss 4 uc re 1 tio lea n se or de r R DB 1 41 CR But Netflix uses analytics in many other ways as well. For example, Netflix uses analytics to determine how much to pay for the distribution rights to DVDs, and to prioritize shipping to infrequent users, who are their most profitable customers. The use of analytics has allowed Netflix to stay ahead of competitor Blockbuster, which also delivers DVDs through the mail but has the advantage of brick and mortar stores as well. Although they undoubtedly increased the awareness of the value of analytics, Davenport and Harris did not create the phenomenon. The term business intelligence (BI) appeared in the early 1990s, but its roots go back to the 1960s with systems such as decision support systems (DSS) and executive information systems (EIS). The details of the technologies have evolved but the goals of using data to gain insight and improve performance are not new. What has changed is the pervasiveness of the technology. Today, almost every Fortune 2000 company has a data warehouse (Erickson, 2006), a key BI enabler. Furthermore, BI is one of the fastest and most consistently growing areas of the information technology industry. The worldwide market for BI technologies is nearly $100 billion, and growing at almost 10% per year (The Economist, 2010). Unlike many other segments of the information systems industry, the BI market continued to grow during the recession that began in 2008 (Henschen, 2009). Harrah’s uses data from their Total Rewards customer loyalty program to create individual relationships with the more than 40 million visitors to their properties. This individualized approach allows Harrah’s to create customized incentive plans for each member of the program. The result is that Harrah’s share of their customers discretionary gaming dollars spent versus their competitors has risen from 30% to 50%. 5 40 t 40 6 Go od sr ec ei p /IR DB Po s R Ac c. Pa y. C t p F-53 ay m en 40 Po M st IRO in vo ice CR /IR DB GR CR B CR Ca s pu t ut n nu es Preliminary Version – send comments to pml@hec.ca h In ve s Readings on Enterprise Resource Planning tm en ts DB GR In ve so nto ld ry pr ch od an uc ge t R co aw ns m um ate pt ria io l n Netflix has used analytics extensively since the company was founded. Netflix is best known for their Cinematch algorithm which matches customers’ movie rankings to clusters of movies in order make personalized recommendations. Netflix creates a personalized web page for every user based on these recommendations. This growth has a number of drivers. One important factor is the maturity of enabling technologies. Increased sophistication of information technology (IT) combined with lower costs has made it much easier to collect, access, distribute, and analyze data. Many of the components of the BI infrastructure, such as data warehouses and data networks, are already in place or are considered relatively low risk investments. Many vendors offer out-of-the-box solutions that can be deployed with minimum effort (Henschen, 2009). The key question remaining for companies is the best way to capitalize on BI and analytics, rather than how to implement the technology required. CR sa Do le m re es ve tic nu es DB Ra w m at In 1994 Continental Airlines was ranked 10th out of 10 major U.S. airlines by the U.S. Department of Transportation in several quality categories. Today Continental is considered one of the best run airlines. Continental Airlines brought itself back from near bankruptcy, in part by investing in real-time business intelligence that affected nearly every aspect of their business. For example, managers can see real-time revenue projections for every flight, and can identify who their most valuable customers are, and which ones are encountering delays. Using real-time information, managers can run what-if scenarios to determine the best way to respond to customer service and scheduling issues. Po M s IG re t go O ce o ip ds ts In re dep qu e ire nd m en en t t Pu r or cha de se r M pu Co E5 rc nv 9N ha er se t to or de r Ex M ec D ut 01 e M RP 5 C pr Con O15 od fi uc r m tio n CR Cr ea VL te 01 de N liv er y With their book, Competing on Analytics, Davenport and Harris (2007) shown a spotlight on companies who have based their strategies on analytical technologies and methods. The authors called these companies analytical competitors, and defined them as companies that make “… extensive use of data, statistical and quantitative analysis, explanatory and predictive models, and fact-based management to drive decisions and actions (Davenport and Harris, 2007, pg 7).” Analytical competitors don’t just use analytics to enhance their operations, but as their primary competitive differentiator. The use of analytics is actively encouraged by management and is embedded within the organizational culture. By now the success stories of many analytical competitors are well known. 30 In ve st DB m en ts 15.1 Introduction DB 41 3 Ca sh CR Pr od or uct de io r n This chapter describes the attributes of companies that Davenport & Harris (2007) called analytical competitors. Business intelligence is one of the fastest growing technology areas, but many organizations are not able to exploit the technology to its fullest. The attributes described in this chapter will help any organization get the most value from their business intelligence investments. M pr od Re D0 l uc e a 7 tio se C DB In v ch en DB an tor ge y CR Ross Hightower (Texas A&M) CR o DB raw ns. Fi pr nis od he uc d t Chapter 15 : The Analytical Organization m Raw at er ia C l CR Fi pr nis h o DB d ed uc ts CR P ut o ro DB utp d. R 41 0 Ac c. Pa y. C Fa c pr tory od o uc ut tio pu n t CR Re ve nu e DB ec . DB CR Ac c. R CHAPTER 15 193 DB DB R m Raw at er ia C l C Pr od or uct de io r n 1 41 CR 5 C pr Con O15 od fi uc r m tio n 30 Ca sh DB en ts CR In ve st DB m BI technologies are often divided into two broad categories : infrastructure and access. Infrastructure includes all the technologies required to manage, extract, transform, and organize data so that it can be accessed by users. Included in this category are master data management (MDM), extraction, transformation, and loading (ETL) technologies, enterprise data warehouses (EDW) and data marts. Also driving the adoption of BI is that efforts to improve performance using conventional information technology have reached the point of diminishing returns (Williams & Williams, 2007). Over the past two decades organizations have invested billions of dollars on ERP systems to automate their business processes. There remains little room for improvement through this means. Organizations are looking to analytics as a means to boost performance by leveraging information. Access technologies include the tools used to access, report, and analyze data. Davenport and Harris, as well as others, further divide this category into reporting and analytical. Reporting tools include standard reports, alerts, ad hoc reports, and dashboards. Analytical tools include data mining, statistical analysis, forecasting, and optimization. It is this latter category that Davenport and Harris called business analytics (BA) (2007). These factors, combined with a growing awareness, have driven the interest in business intelligence and business analytics. A study by Computer Economics found that 84% of organizations reported some level of business intelligence program (Computer Economics, 2008). Yet, many of these organizations have not realised significant benefits from these investments (Williams & Williams, 2007). What’s the secret ? What separates the few companies that are able, not only to capitalize on their BI investments, but are able to separate themselves from their competition by applying analytics ? Fortunately, the capabilities that allow organizations to successfully apply BI are easy to identify, if not easy to implement. This chapter describes what it takes for an organization to become an analytical competitor. While it’s not necessary or desirable for every organization to make the commitment required to be an analytical competitor, every organization can benefit from becoming more analytical. Developing the capabilities described in this chapter will help them do that. One way to understand the difference between reporting and analytical tools is the time frame of the data used, and the questions answered. Reporting tools are suitable for presenting lag information that reflects past performance. Lag information is useful for answering questions focused on “what is” or “what happened”, requiring little or no analysis. Lag information is useful as a means to monitor performance and make evolutionary improvements in existing business processes, but is not suitable for making the sort of revolutionary changes that lead to competitive advantage (Laursen & Thorlund, 2010). The majority of organizations that have deployed BI limit themselves to reporting tools (Davenport & Harris, 2007). 5 t 40 6 Go od sr ec ei p t p F-53 ay m en Po s R Ac c. Pa y. C 40 Po M st IRO in vo ice CR /IR CR CR B CR Ca s pu t ut n nu es Preliminary Version – send comments to pml@hec.ca h In ve s Readings on Enterprise Resource Planning tm en ts DB GR /IR DB In ve so nto ld ry pr ch od an uc ge t DB at DB Ra w m Many authors agree that business intelligence refers to all of the technologies and methods used to organize and analyze data in order to measure performance, improve operations, and guide planning. According to this broad definition, BI includes not only the technologies required to extract, organize, report, and analyze data, but also the management processes and governance structures required to make use of the information R co aw ns m um ate pt ria io l n Questions such as “why” or “what’s next” can only be answered by lead information. Lead information results from analysis of lag information using analytical tools. Lead information can be used to create new processes, or to re-engineer existing processes in ways that provide a competitive advantage. It is this type of information that analytical competitors exploit so successfully. The use of analytical tools is usually the culmination of an organization’s evolution with BI. To successfully apply analytics, the technical infrastructure must be in place, analytical talent must be available, and users must understand how to capitalize on the information provided by the tools. Figure 1 shows the relationship between access tools, information, and competitive advantage. GR 15.2 Business Intelligence and Analytics sa Do le m re es ve tic nu es 40 Po M s IG re t go O ce o ip ds ts In re dep qu e ire nd m en en t t Pu r or cha de se r M pu Co E5 rc nv 9N ha er se t to or de r Ex M ec D ut 01 e M RP CR o DB raw ns. CR DB CR In v ch en DB an tor ge y 41 3 Cr ea VL te 01 de N liv er y gained from the technology. In the broadest sense of the term, BI includes Business Analytics (BA), Corporate Performance Management (CPM), Business Performance Management (BPM), Enterprise Information Management (EIM) and other related areas. DB Fi pr nis od he uc d t Information technology has also enabled most organizations to accumulate large data repositories to exploit. ERP, CRM, pointof-sale, and web sites all generate a wealth of data. According to Jim Goodnight, CEO of SAS Institute, in 2010 the amount of digital data in the world doubles every 11 hours (Mohammad Ali Khan, et al., 2009). This is a little easier to believe when you consider that Wal-Mart alone processes more than 1 million customer transactions per hour, storing the data in databases estimated at more than 2.5 petabytes (The Economist, 2010). The need to manage this flood of data and a rising awareness of the value that can be mined from it has helped drive BI adoption. M pr od Re D0 l uc e a 7 tio se pr Ma CO od ss 4 uc re 1 tio lea n se or de r CR P CR Fi pr nis h o DB d ed uc ts Hightower – The analytical organization ut o ro DB utp d. R 41 0 Ac c. Pa y. C Fa c pr tory od o uc ut tio pu n t CR Re ve nu e DB ec . DB CR Ac c. R CHAPTER 15 194 DB Hightower – The analytical organization DB pr Ma CO od ss 4 uc re 1 tio lea n se or de r R m Raw at er ia C l tools 1 41 CR What’s the best we can do? CR Data mining Statistical analysis Forecasting Optimization 30 • • • • Standard reports Dashboards Ad hoc reports Alerts 5 DB Ca sh In ve st DB m en ts 41 3 • • • • M pr od Re D0 l uc e a 7 tio se What will happen next? What happened? C pr Con O15 od fi uc r m tio n questions Cr ea VL te 01 de N liv er y Why is this happening? What is? Pr od or uct de io r n C DB LeAD CR In v ch en DB an tor ge y Lag CR o DB raw ns. DB Figure 1 – Analytical Tools, Questions Asked ,and Information Types CR Fi pr nis od he uc d t CR Fi pr nis h o DB d ed uc ts CR P ut o ro DB utp d. R 41 0 Ac c. Pa y. C Fa c pr tory od o uc ut tio pu n t CR Re ve nu e DB ec . DB CR Ac c. R CHAPTER 15 analytics reporting competitive potential M pu Co E5 rc nv 9N ha er se t to or de r Ex M ec D ut 01 e M RP 15.3 Becoming an Analytical Organization It’s no coincidence that analytical competitors are among the most innovative and successful companies in their industries. This success is only partly the result of using analytical techniques and decision making methods. Competing on analytics requires organizations to have a rigorously rational approach to running their companies, and embrace a culture of innovation and continuous improvement. There are no short cuts. Becoming an analytical competitor requires a long term commitment to cultural change, investments in people and infrastructure, and intense focus on extracting business value using analytical methods. Organizations should be prepared to maintain a focused, sustained effort over years to develop the technical infrastructure, internal skills, and organizational structures required. 5 40 t Go od sr ec ei p t p F-53 ay m en 40 Po M st IRO in vo ice CR Po s R Ac c. Pa y. C CR DB GR /IR 40 6 This long-term view of BI development is reflected in the models used to classify organizations’ BI maturity or readiness. One such model is the maturity model developed by Davenport & Harris (2007). This model defines five stages of maturity based on organizational, human, and technological factors. Organizations should expect to progress through the stages in order, learning lessons and building capabilities along the way. Skipping stages is difficult and often leads to failure and regression. The characteristics of each stage are described in Table 1. CR B CR Ca s pu t ut n nu es Preliminary Version – send comments to pml@hec.ca h In ve s Readings on Enterprise Resource Planning tm en ts DB GR /IR DB In ve so nto ld ry pr ch od an uc ge t R co aw ns m um ate pt ria io l n sa Do le m re es ve tic nu es DB Ra w m at On the other hand, it’s not necessary for a company to make such a commitment to benefit from BI. Much of what makes analytical competitors so successful can be applied on a smaller scale. The problem is most companies aren’t able to extract sufficient value from their BI investments because their efforts are uncoordinated, underfunded and not focused on decisions that provide significant business value. According to Williams & Williams (2007), BI provides value when it combines products, technology, processes, and people to organize information Po M s IG re t go O ce o ip ds ts In re dep qu e ire nd m en en t t Pu r or cha de se r that management needs to improve profit and performance. Analytics should be embedded in core business processes using measures that are fact-based, analytically rigorous, and repeatable. Decisions must be linked to actions that have measurable impacts on performance. Dabbling in BI with only a vague idea of what the benefits will be is a recipe for failure. To achieve success, organizations must take a systematic approach to laying the groundwork, developing managerial and technical readiness, and choosing targets that will deliver value. This is true whether BI is applied enterprise wide or on a much smaller scale. 195 DB Hightower – The analytical organization CR Pr od or uct de io r n Desire for more objective data, successes from point use of analytics start to get attention Recent transaction data un-integrated, missing important information, isolated BI / analytics efforts Mostly separate analytic processes, Building enterpriselevel plan Analysts in multiple areas of business but with limited interaction Executive support for fact-based culture-may meet considerable resistance Proliferation of BI tools, Data marts / data warehouse established expands Change program to develop integrated analytical processes and applications build analytical capabilities Some embedded analytics processes Skills exist, Broad C-suite but often not support aligned to right level / right role Change management to build a fact-based culture High-quality data, have an enterprise BI plan / strategy, IT processes and governance principles in place Deep strategic insights, continuous renewal and improvement Fully embedded and much more highly integrated Highly skilled, leveraged, mobilized, centralized, out-sourced grunt work Broadly supported fact-based culture, testing and learning culture Enterprise-wide BI/BA architecture largely implemented 6 40 Po M st IRO in vo ice CR 40 Po s CR B CR Ca s pu t ut n nu es Preliminary Version – send comments to pml@hec.ca h In ve s Readings on Enterprise Resource Planning tm en ts DB GR /IR DB In ve so nto ld ry pr ch od an uc ge t t p F-53 ay m en Figure 2 shows the Analytical Capability Assessment (ACA) model developed by Lundgren & Larsson (2009). The ACA builds on the work of several authors and is one of the most complete frameworks available for identifying the essential ingredients for becoming an analytical competitor. R CR 15.4 Analytical Capability Assessment Ac c. Pa y. C R co aw ns m um ate pt ria io l n sa Do le m re es ve tic nu es DB Ra w m So what are the characteristics of analytical competitors ? What capabilities must organizations develop to progress through the stages of BI maturity ? Fortunately, there has been enough research on this topic that the answers to these questions are known, though, not easy to implement. The next section describes a model that answers these questions. DB at GR /IR CEO passion, broad-based management commitment t 30 40 Pu r or cha de se r M pu Co E5 rc nv 9N ha er se t to or de r Executive-early stages of awareness of competitive possibilities 5 C pr Con O15 od fi uc r m tio n DB en ts CR In ve st DB m Ex M ec D ut 01 e M RP Cr ea VL te 01 de N liv er y Pockets of Functional and isolated tactical analysis (may be in finance, SCM, or marketing / CRM) Go od sr ec ei p Coordinated, established enterprise performance metrics, build analytically based insights 5- Analytical competitors Missing/poor-quality data, multiple definitions, un-integrated systems 5 3- Analytical aspirations 4- Analytical companies Knowledge allergic – pride on gut-based decisions 1 Disconnected, very narrow focus None Po M s IG re t go O ce o ip ds ts Autonomous activity builds experience and confidence using analytics ; creates new analytically based insights None Culture 41 2- Localized analytics TECHNOLOGY Sponsorship CR Doesn’t exist Ca sh Limited insight into customers, markets, competitors 41 3 1- Analytically impaired Skills HUMAN M pr od Re D0 l uc e a 7 tio se DB Analytical process CR In v ch en DB an tor ge y Analytical objectives C ORGANIZATION DB STAGE In re dep qu e ire nd m en en t t o DB raw ns. CR Fi pr nis od he uc d t Table 1 – The Analytical Capability Model, reprinted from Competing on Analytics, by Davenport & Harris, 2007. pr Ma CO od ss 4 uc re 1 tio lea n se or de r DB R m Raw at er ia C l CR Fi pr nis h o DB d ed uc ts CR P ut o ro DB utp d. R 41 0 Ac c. Pa y. C Fa c pr tory od o uc ut tio pu n t CR Re ve nu e DB ec . DB CR Ac c. R CHAPTER 15 196 DB DB R m Raw at er ia C l Pr od or uct de io r n C CR o DB raw ns. CR DB CR 1 Ca sh 41 Strategic Alignement CR DB Business / IT Partnership 5 C pr Con O15 od fi uc r m tio n CR In ve st DB m en ts Ability to Align & Govern BI BI Portfolio Management Leadership 30 In v ch en DB an tor ge y Figure 2 – Analytical Capability Assessment Model Cr ea VL te 01 de N liv er y 41 3 The next three sections describe the capabilities in these three categories. DB Fi pr nis od he uc d t The model classifies the capabilities into three broad categories : the ability to align and govern, the ability to leverage BI, and the ability to deliver BI. M pr od Re D0 l uc e a 7 tio se pr Ma CO od ss 4 uc re 1 tio lea n se or de r CR P CR Fi pr nis h o DB d ed uc ts Hightower – The analytical organization ut o ro DB utp d. R 41 0 Ac c. Pa y. C Fa c pr tory od o uc ut tio pu n t CR Re ve nu e DB ec . DB CR Ac c. R CHAPTER 15 Corporate Culture Process Maturity Continuous Improvement Decision Process M pu Co E5 rc nv 9N ha er se t to or de r Ability to Deliver BI 5 Pu r or cha de se r BI Architecture 40 t p F-53 ay m en Po s R 40 Po M st IRO in vo ice CR CR B CR Ca s pu t ut n nu es Preliminary Version – send comments to pml@hec.ca h In ve s Readings on Enterprise Resource Planning tm en ts DB GR /IR DB In ve so nto ld ry pr ch od an uc ge t R co aw ns m um ate pt ria io l n Strategic alignment enables an organization to identify key targets where BI will have the greatest impact. Without this alignment BI applications will fail to live up to their hype and DB GR CR Strategic Alignment management may become reluctant to sustain BI investments. According to Davenport and Harris (2007), analytical competitors identify strategic competencies, and use analytics to develop the competencies into a strategic advantage. This explicit link between analytics and the strategic competency aligns the use of analytics with the strategic objectives of the organization, and is the primary focus of BI initiatives. It’s only after BI proves successful that analytics migrate to other, less strategically important, applications. Ac c. Pa y. C sa Do le m re es ve tic nu es DB Ra w m at The ability to align and govern refers to how successfully the organization’s analytical efforts are aligned with the strategic imperatives of the organization, and how well governance structures in the organization support the use of analytics. /IR 15.4.1 Ability to Align and Govern 6 Go od sr ec ei p t Tools & Applications 40 Data Management Po M s IG re t go O ce o ip ds ts In re dep qu e ire nd m en en t t Ex M ec D ut 01 e M RP Ability to Leverage BI 197 DB DB R m Raw at er ia C l C Pr od or uct de io r n 1 41 CR 5 C pr Con O15 od fi uc r m tio n An example of a methodology that links strategic objectives with BI applications is the BI Opportunity Analysis approach shown in Figure 3 (Williams & Williams, 2007). This method seeks to link business drivers and strategies to specific BI initiatives that affect core business processes. The analysis begins with the strategic planning process that defines the business context in terms of the company’s strategy, goals, and objectives. Once the business context is established, the strategic objectives become the basis for the business design. 30 Ca sh DB en ts Business process re-engineering is fundamental to the BI Opportunity Analysis. Analytical competitors embed analytics in core business processes where they become a routine part of doing business. Embedding analytics in business processes directly links the insights gained through the analysis with decisions and actions that capitalize on those insights (Davenport & Harris, 2010). During the business design phase, core processes are re-engineered in ways that allows optimal use of embedded analytics. BI Opportunity Analysis achieves strategic alignment by linking strategic goals and objectives with the decisions and actions taken by front line employees. CR In ve st DB m CR o DB raw ns. CR DB CR In v ch en DB an tor ge y 41 3 Cr ea VL te 01 de N liv er y The business design includes the value discipline which defines the fundamental way the organization competes in the market. The value discipline is the basis for the design of core business processes and the analytical frameworks required to support them. DB Fi pr nis od he uc d t Strategic alignment occurs when there are clear links between business strategies, business processes and BI initiatives. Achieving these links requires that the strategic planning process and the process of evaluating and choosing BI initiatives are coordinated. In many organizations individual business units invest in BI projects without regard to the organization’s overall strategic objectives, or even local strategic objectives. Even when BI projects are initiated at the enterprise level, they are often chosen based on expediency rather than for strategic reasons. For example, projects may be chosen simply because the data is available. M pr od Re D0 l uc e a 7 tio se pr Ma CO od ss 4 uc re 1 tio lea n se or de r CR P CR Fi pr nis h o DB d ed uc ts Hightower – The analytical organization ut o ro DB utp d. R 41 0 Ac c. Pa y. C Fa c pr tory od o uc ut tio pu n t CR Re ve nu e DB ec . DB CR Ac c. R CHAPTER 15 Po M st IRO in vo ice CR DB 40 CR t p F-53 ay m en CR B CR Ca s pu t ut n nu es Preliminary Version – send comments to pml@hec.ca h In ve s Readings on Enterprise Resource Planning tm en ts DB GR /IR DB In ve so nto ld ry pr ch od an uc ge t R Reprinted from The Profit Impact of Business Intelligence (p.27), by S. Williams & N. Williams, 2007, Elsevier, Inc. Reprinted with permission. Po s Ac c. Pa y. C R co aw ns m um ate pt ria io l n sa Do le m re es ve tic nu es DB Ra w m at GR /IR 40 6 Go od sr ec ei p t Po M s IG re t go O ce o ip ds ts In re dep qu e ire nd m en en t t 40 5 Pu r or cha de se r M pu Co E5 rc nv 9N ha er se t to or de r Ex M ec D ut 01 e M RP Figure 3 – BI Opportunity Analysis 198 DB DB R m Raw at er ia C l 1 5 C pr Con O15 od fi uc r m tio n CR A prerequisite for BI portfolio management is an enterprise view of the value and role of BI in the organization. Not every project can, or should, be funded. The projects that provide the greatest benefit to the enterprise should have priority over those that only benefit individual departments. An enterprise perspective allows the company to leverage resources and avoid duplicating efforts in different parts of the organization. Enterprise technology standards allow for effective support from IT, and facilitate the coordination of the BI infrastructure with the organization’s IT infrastructure. Cr ea VL te 01 de N liv er y CR 41 CR DB Along with establishing an enterprise view, the organization must establish a method for evaluating alternative BI projects. The BI Opportunity Map shown in Figure 4, an extension of the BI Opportunity Analysis, can be used to classify potential BI projects according to business impact and risk (Williams & Williams, 2007). Risk refers to the organization’s likelihood of completing the project successfully. The relevant factors to consider for risk include available data sources, technical difficulty, adequacy of infrastructure, and availability of technical and analytical talent. Projects classified as easy wins and plums are especially important in the early stage of an organization’s experience with BI, as they can help to build momentum for further efforts. 30 In ve st DB m en ts 41 3 Ca sh CR Pr od or uct de io r n DB C CR o DB raw ns. In many companies, BI initiatives are scattered throughout the organization in an uncoordinated and ad hoc manner. Often there are no standard methods for evaluating alternative BI projects. Analytical competitors manage BI as a portfolio of investments (Davenport & Harris, 2007). This allows them to evaluate alternatives and prioritize them according to their potential value, as well as how well they fit with the overall BI strategy. In v ch en DB an tor ge y An enterprise view also promotes a cross-functional, process oriented view of the organization. This affords a more comprehensive view of business processes, and facilitates the re-engineering efforts required by the strategic alignment process. Departmental resistance is often a stumbling block for re-engineering efforts. An enterprise view shifts the focus from whether individual departments are winners or losers, to what is most beneficial for the enterprise as a whole. DB Fi pr nis od he uc d t BI Portfolio Management M pr od Re D0 l uc e a 7 tio se pr Ma CO od ss 4 uc re 1 tio lea n se or de r CR P CR Fi pr nis h o DB d ed uc ts Hightower – The analytical organization ut o ro DB utp d. R 41 0 Ac c. Pa y. C Fa c pr tory od o uc ut tio pu n t CR Re ve nu e DB ec . DB CR Ac c. R CHAPTER 15 Po M st IRO in vo ice CR DB 40 CR Po s CR B CR Ca s pu t ut n nu es Preliminary Version – send comments to pml@hec.ca h In ve s Readings on Enterprise Resource Planning tm en ts DB GR /IR DB In ve so nto ld ry pr ch od an uc ge t R Ac c. Pa y. C Reprinted from The Profit Impact of Business Intelligence (p. 33), by S. Williams & N. Williams, 2007, Elsevier, Inc. Reprinted with permission. t p F-53 ay m en R co aw ns m um ate pt ria io l n sa Do le m re es ve tic nu es DB Ra w m at GR /IR 40 6 Go od sr ec ei p t Po M s IG re t go O ce o ip ds ts In re dep qu e ire nd m en en t t 40 5 Pu r or cha de se r M pu Co E5 rc nv 9N ha er se t to or de r Ex M ec D ut 01 e M RP Figure 4 – BI Opportunity Map 199 DB DB DB R m Raw at er ia C l Pr od or uct de io r n Management must evangelize and communicate a vision that permeates the organization. Not only must sponsors support specific projects, they must be committed to an analytical style of management. It’s not enough to develop a BI infrastructure and analytical applications. The culture of the organization must be oriented toward the use of evidence and facts to support decisions, and a willingness to experiment and innovate using analytics. This requires much more than the garden variety executive support that many IT projects receive. 1 41 CR 5 C pr Con O15 od fi uc r m tio n For companies that don’t aspire to be analytical competitors or who have not reached that stage of BI maturity, leadership may not come from the CEO, but it must come from somewhere. Lower level sponsors can prove the value of analytics by producing highly visible, successful applications within their own departments. The Erickson (2006) study mentioned above found that projects with “very committed” sponsors were twice as likely to succeed as projects with only “fairly committed” sponsors. In fact, projects with “fairly committed” sponsors were more likely to struggle. The importance of the business/IT partnership doesn’t end when an application is deployed. Unlike operational systems, which are deployed and change little for long periods of time, analytical applications must continually adapt to changing business conditions. To stay competitive, members of the organization must constantly look for opportunities, and must be able to exploit them quickly. Thus, the business users and IT personnel must have a cooperative relationship. 15.4.2 Ability to Leverage BI Investments M pu Co E5 rc nv 9N ha er se t to or de r Ex M ec D ut 01 e M RP 30 en ts CR In ve st DB m DB Ca sh CR 41 3 Cr ea VL te 01 de N liv er y CR C CR o DB raw ns. To be an analytical competitor, analytical applications must drive the way that business is done and, unless BI initiatives are designed from the beginning by business users, this will not happen. BI projects in analytical competitors are almost always led by business users supported by strong partnerships with IT departments (Williams & Williams, 2007). According to Larsson & Lundgren (2009), the largest cause of failing BI initiatives is that they are driven by IT rather than business. This is because BI efforts driven by IT departments are often treated as purely technical projects with little thought for their business value. Erickson (2006) conducted a study of the factors that influence the success of BI in 540 organizations. This study found that BI teams that were “very aligned” with business were nearly five times more likely to be succeeding than struggling. 5 40 40 6 Go od sr ec ei p t Po M s IG re t go O ce o ip ds ts t p F-53 ay m en Po s R Ac c. Pa y. C In principle, an organization can be an analytical competitor without any of the technological gizmos that are associated with BI, such as data warehouses and analytical tools. Data and analytical services can be purchased from third parties (Larsson & Lundgren, 2009). At the same time, having an extensive BI infrastructure doesn’t make a company an analytical competitor. 40 Po M st IRO in vo ice Culture around Use of Information and Analytics CR B CR Ca s pu t ut n nu es Preliminary Version – send comments to pml@hec.ca h In ve s Readings on Enterprise Resource Planning tm en ts DB GR /IR DB In ve so nto ld ry pr ch od an uc ge t R co aw ns m um ate pt ria io l n Probably the most important capability of all is leadership. Becoming an analytical competitor is impossible without the enthusiastic support of the CEO. While department level BI projects can succeed with lower level support, analytical CR /IR GR at CR m Ra w DB sa Do le m re es ve tic nu es Leadership BI must make measurable impacts on operational processes that deliver value. Improved forecasting methods, for example, are useless unless they lead to measurable benefits such as reduced inventory or reduced order expediting costs. Many things have to happen to realize these benefits. The right data must be available and an accurate forecast must be generated. The right people must have timely access to the forecast, they must make the correct decisions based on the forecast, and take actions that influence the relevant processes. Ultimately, the value is obtained only when the processes that deliver value are affected. A technically brilliant BI application is a waste of money if it doesn’t result in improved business decisions and actions. The ability to leverage BI investments refers to the organization’s ability to use BI applications to create value. Pu r or cha de se r In re dep qu e ire nd m en en t t The philosophy of companies that are successful with BI is that there are no IT projects ; there are business projects with technical aspects. These companies integrate technology experts within the business at all levels rather than isolating them from business users (Davenport & Harris, 2010). In this way the two sides learn from one another and develop cooperative working relationships. A common way to achieve business/IT partnerships is through creating organizational and governance structures that explicitly address the IT/business partnership. Many authors recommend the formation of BI competency centers (BICC). In some cases the BICC is a separate organizational unit but in many cases it is housed within another department such as IT. The responsibilities of the BICC include promoting the use of BI, defining technical standards, training end users, ensuring the alignment of business needs with BI efforts and prioritizing BI projects. The BICC is staffed by BI analysts and IT personnel and serves as a centralized forum for business, IT, and analytical experts to facilitate knowledge sharing and coordination. DB In v ch en DB an tor ge y competitors require consistent and enthusiastic commitment from the entire organization, which is impossible without support from the highest levels. DB Fi pr nis od he uc d t Business/IT Partnership M pr od Re D0 l uc e a 7 tio se pr Ma CO od ss 4 uc re 1 tio lea n se or de r CR P CR Fi pr nis h o DB d ed uc ts Hightower – The analytical organization ut o ro DB utp d. R 41 0 Ac c. Pa y. C Fa c pr tory od o uc ut tio pu n t CR Re ve nu e DB ec . DB CR Ac c. R CHAPTER 15 200 DB R m Raw at er ia C l • Process re-engineering expertise DB Fi pr nis od he uc d t Analytical competitors successfully exploit BI because the use of analytics and information is embedded in the organizational culture. Employees believe in the power of information to effect positive change in the way the business operates, and are empowered to use information to innovate. Pr od or uct de io r n 1 41 CR DB 30 Continuous Improvement Culture Competitive advantages are fleeting, so analytical competitors are continually looking for new ways to use information and apply analytics. Analytical competitors are often the most innovative in their industries because of their reliance on experimentation to improve performance and test new ideas. In many companies employees are encouraged to ask questions and conduct experiments to find the answers. Many analytical competitors use experimentation extensively. Google conducts more than 200 experiments per day with multiple versions of ads to determine which are the most effective (Ayers, 2007). Capital One conducts many experiments on which types of direct mail offers are effective (Ayers, 2007). Once new ideas are identified through experimentation, they often must be implemented in the form of process improvements. Process Maturity Hammer defines process maturity as the ability of processes to deliver higher performance over time (Hammer, 2007). Organizations with mature processes are able to re-engineer their processes, to execute processes efficiently, to measure process effectiveness, and to assign accountability for process performance. Hammer (2007) developed the Process Maturity Model to evaluate organizations’ process maturity. In re dep qu e ire nd m en en t t • Comprehensive process design specification • Senior executive leadership to support process redesign 40 6 Go od sr ec ei p Po M st IRO in vo ice CR DB t p F-53 ay m en Po s CR B CR Ca s pu t ut n nu es Preliminary Version – send comments to pml@hec.ca h In ve s Readings on Enterprise Resource Planning tm en ts DB GR /IR DB In ve so nto ld ry pr ch od an uc ge t R Ac c. Pa y. C R co aw ns m um ate pt ria io l n • Culture that supports customer focus, teamwork, willingness to change and accountability 40 sa Do le m re es ve tic nu es DB The second type of characteristic is enterprise capabilities : CR Ra w m at GR /IR • A process owner who is a senior executive • Process oriented metrics to track performance 40 t • Skilled and knowledgeable performers to execute the process • Information and management infrastructure to support the process The strategies used to capitalize on analytics are similar to those used to follow lean practices including cutting waste, measuring, optimizing, and continuous change. For example, one of the more common ways to use BI to improve operational processes is to reduce process cycle time ; the time it takes to complete a process from beginning to end (Williams & Williams, 2007). Similarly, the goal of lean practices is to eliminate any activity or cost in processes that does not produce value for the customer. Analytics provides the means to identify wasteful activities and costs, and to evaluate the effectiveness of process changes. Because of these synergies, companies often implement lean practices and analytics simultaneously (Larsson and Lundgren, 2009). Po M s IG re t go O ce o ip ds ts To achieve process maturity organizations must develop two types of characteristics. The first are process enablers : 5 Pu r or cha de se r M pu Co E5 rc nv 9N ha er se t to or de r Ex M ec D ut 01 e M RP 5 C pr Con O15 od fi uc r m tio n CR In ve st DB m en ts 41 3 Ca sh One aspect of an analytical culture is an expectation that decisions should be based on facts and analytics rather than intuition. There is considerable evidence that decisions based on analytics are more often correct than those based on intuition (Ayers, 2007). Even lofty decisions such as the relative rankings of wine vintages are more accurately made using analytical models rather than the judgement of experts (Ayers, 2007). Yet, in many organizations most decisions are based on gut feelings and intuition (Davenport & Harris, 2010). Organizations operate on blind faith in the value of gut decisions because the lack of analytical focus prevents evaluation of the results. It’s not that intuition is not valued in companies that are analytically focused. However, the role of intuition is most often to determine what questions to ask rather than in providing answers (Ayers, 2007). Once the question is posed, the task is to determine what data and facts are available to obtain the answer. Cr ea VL te 01 de N liv er y CR C DB Many of these characteristics are the same as those that allow organizations to use BI effectively. The ability to re-engineer processes to embed analytics is a critical capability of the strategic alignment process. Improving process performance also requires well designed process oriented metrics, and employees who are able to use analytics effectively. Effective use of BI requires well developed infrastructure, senior leadership, and a culture amenable to change and teamwork. In fact, it would be impossible to be an analytical competitor without mature processes. CR In v ch en DB an tor ge y CR o DB raw ns. • Governance structures for managing change and project management M pr od Re D0 l uc e a 7 tio se pr Ma CO od ss 4 uc re 1 tio lea n se or de r DB CR P CR Fi pr nis h o DB d ed uc ts Hightower – The analytical organization ut o ro DB utp d. R 41 0 Ac c. Pa y. C Fa c pr tory od o uc ut tio pu n t CR Re ve nu e DB ec . DB CR Ac c. R CHAPTER 15 201 DB Hightower – The analytical organization C DB pr Ma CO od ss 4 uc re 1 tio lea n se or de r Pr od or uct de io r n 1 41 CR Ca sh DB en ts C pr Con O15 od fi uc r m tio n CR In ve st DB m 30 Po M s IG re t go O ce o ip ds ts In re dep qu e ire nd m en en t t 40 5 Pu r or cha de se r M pu Co E5 rc nv 9N ha er se t to or de r Ex M ec D ut 01 e M RP 5 Figure 5 – Decision Process Engineering Cr ea VL te 01 de N liv er y M pr od Re D0 l uc e a 7 tio se o DB raw ns. CR DB CR 41 3 CR and analytical applications. Figure 5 illustrates an engineered decision for a simple analytical application. Such a systematic approach to decision making allows the organization to track decision performance, and provides a means for systematically improving the decision process. Clearly, not all decisions can be engineered in this way, but for decisions that are repetitive or semi-repetitive this approach would improve decision performance. While most organizations have invested in systems such as ERP, SCM, and CRM to structure and standardize transaction processing, the use of analytics is often ad hoc and unstructured. Analytical competitors however, take a systematic approach to using information and making decisions based on analytics. Williams & Williams (2007) coined the phrase decision process engineering to describe a structured approach to designing decision processes to make the best use of information In v ch en DB an tor ge y R DB Fi pr nis od he uc d t Decision Process Improvement Culture m Raw at er ia C l CR Fi pr nis h o DB d ed uc ts CR P ut o ro DB utp d. R 41 0 Ac c. Pa y. C Fa c pr tory od o uc ut tio pu n t CR Re ve nu e DB Reprinted from The Profit Impact of Business Intelligence (p. 56), by S. Williams & N. Williams, 2007, Elsevier, Inc. Reprinted with permission. Go od sr ec ei p t 15.4.3 Ability to Deliver Technically Po s R 40 Po M st IRO in vo ice CR Ac c. Pa y. C CR DB GR /IR 40 6 The ability to deliver technically concerns the capabilities of the organization’s BI technical infrastructure and the availability of high quality data. Analytics are possible without extensive infrastructure. For example, some companies use externally purchased data, and there are a number of cloud-based analytical offerings that can be implemented with minimal technical infrastructure. In most cases, however, analytics requires high quality data extracted from internal sources, data warehouses to store the data, and applications to analyse the data. This requires a well-engineered and efficient technical infrastructure. CR B CR Ca s pu t ut n nu es Preliminary Version – send comments to pml@hec.ca h In ve s Readings on Enterprise Resource Planning tm en ts DB GR /IR DB In ve so nto ld ry pr ch od an uc ge t R co aw ns m um ate pt ria io l n sa Do le m re es ve tic nu es DB Ra w m at A culture oriented around improved decision making also encourages information and knowledge sharing. Organizational cultures in which information is tightly controlled or used as a political weapon to protect turf will not support fact-based decision making. In analytical organizations, information transparency is highly valued and all information, even information that questions the status quo or may be politically incorrect, is shared openly. Erickson (2006) found that BI projects in organizations that shared information “Very Openly” were more than five times more likely to be succeeding than struggling, while projects in organizations that shared information “Not Very Openly” were more than five times more likely to be struggling than succeeding. t p F-53 ay m en ec . DB CR Ac c. R CHAPTER 15 202 DB DB pr Ma CO od ss 4 uc re 1 tio lea n se or de r CR P R C Pr od or uct de io r n 1 41 CR 5 C pr Con O15 od fi uc r m tio n 30 Ca sh DB en ts CR In ve st DB m Integrating Data While establishing trusted data sources involves the same data in multiple systems, integration concerns matching data from different data domains. For example, a sales analysis application may require integrating customer buying history from an ERP system with clickstream data from a web site. Integrating data requires that the data records from different sources can be matched. 5 40 t 40 6 Go od sr ec ei p t p F-53 ay m en Po s R 40 Po M st IRO in vo ice DB By some estimates, 50% of the data records in large enterprise databases contain errors (Rettig, 2007). Because data is collected in different ways, stored in different systems, manipulated, and copied between systems there are many opportunities for errors to occur. Even when data is entered correctly, it can later become incorrect. Ac c. Pa y. C CR Improving Data Quality CR B CR Ca s pu t ut n nu es Preliminary Version – send comments to pml@hec.ca h In ve s Readings on Enterprise Resource Planning tm en ts DB GR /IR DB In ve so nto ld ry pr ch od an uc ge t R co aw ns m um ate pt ria io l n sa Do le m re es ve tic nu es DB Ra w Without good data, analytics is impossible. The first step in any BI program should be ensuring high quality data. Without strong data management, organizations will struggle with data inconsistencies, errors, and data integration issues. Every new BI project will be plagued by data issues and business users will become frustrated and will lose faith in BI. Data problems fall into the following three categories (Vayghand, et al., 2007) : CR /IR GR m at Data Management Many of the barriers to integration are a result of the data silos described above. Data formats are often tightly coupled with applications and the database systems in which they are stored. Before the data can be integrated, a series of complex extraction and transformation steps are necessary. For example, a customer id in one system may consist of ten alphanumeric characters, whereas in another system it is a five digit numeric value. It might be possible to match many of these records by using a field such as address, but, in any case, a considerable amount of manual work will be involved to ensure the matching is accurate. Po M s IG re t go O ce o ip ds ts In re dep qu e ire nd m en en t t Pu r or cha de se r M pu Co E5 rc nv 9N ha er se t to or de r Ex M ec D ut 01 e M RP M pr od Re D0 l uc e a 7 tio se o DB raw ns. CR DB CR 41 3 Cr ea VL te 01 de N liv er y The term architecture with regard to BI is often used in a more general sense to include the organizational processes and governance structures required to implement BI (Williams & Williams, 2007). A well-developed architecture should specify the operational processes required to manage data, implement BI applications, and support end users. Data and governance architectures should be defined in conjunction with the technical architecture. Unfortunately, in many organizations the term architecture is not an accurate description of the technical environment. BI often develops organically, with individual or departmental level efforts occurring independently. It’s not uncommon for large organizations to have hundreds of independent data marts, and dozens of incompatible platforms (Davenport & Harris, 2010). Not only does this violate the BI goal of a single version of the truth, it leads to wasted resources and an inflexible BI environment. Organizations that find themselves in the Infant stage of BI maturity with many independent BI projects will find it difficult to establish order amidst the chaos. The earlier an organization can take a systematic, coordinated approach to BI architecture, the better off they will be in the long run. Even if early efforts are single projects within a single department, a long term architectural view should guide technical decisions. One approach to ensuring a consistent architecture is to adopt the BI products offered by the organization’s ERP vendors. CR One of the overarching goals of data management is to establish a single version of the truth. This means that reports generated from an organization’s data produces identical results, regardless of how the data was accessed, where the data was obtained, and which application generated the report. The difficulties in achieving this arise because data is usually managed at the local level rather than at the enterprise level. Multiple systems, each with one or more databases, are developed and maintained independently. Not only is data duplicated in these data silos, but without established enterprise level standards, the definitions of data objects are inconsistent. For example, according to Jeanne Harris, the average company has between 15-17 definitions of customer (Larsson & Lundgren, 2009). Imagine two departments meeting to discuss marketing plans, each bringing reports showing profitability by customer. Not only are the numbers likely to be different since they are coming from different databases, but different definitions of what constitutes a customer means that meaningful comparisons are impossible. The result is often an argument over which numbers are correct. The term architecture implies that there is some overall design to BI technical components, and that they are chosen and integrated in some systematic way. At a minimum, the BI architecture should identify standard platforms, tools, and technologies used to implement BI applications. Standards allow analysts, developers, and users to leverage lessons learned, reducing the risk inherent in implementing new applications, and promoting continuous process improvement. A well designed architecture should also align with the operational IT architecture. For example, extracting data from operational system databases is complex, and compatible BI technologies should be chosen to facilitate this task as much as possible. In v ch en DB an tor ge y Establishing Trusted Data Sources for Enterprise-Critical Data DB Fi pr nis od he uc d t BI Architecture m Raw at er ia C l CR Fi pr nis h o DB d ed uc ts Hightower – The analytical organization ut o ro DB utp d. R 41 0 Ac c. Pa y. C Fa c pr tory od o uc ut tio pu n t CR Re ve nu e DB ec . DB CR Ac c. R CHAPTER 15 203 DB DB R Pr od or uct de io r n 1 41 CR 5 C pr Con O15 od fi uc r m tio n Service oriented architecture (SOA) is an essential element of the target architecture because it is a means to control data access. Services are remotely accessible programs that provide access to data and functionality. When many groups use the same data, many interfaces will be created to access the data, each with a unique view and interpretation of the data. If the data is stored on multiple systems requiring integration, the results can depend on the steps used to perform the integration. In a SOA-enabled architecture, data must be accessed through common, trusted services. The location of the data and the means to integrate it are part of the service, so the results will be the same regardless who accesses the data. Thus, SOA enforces a single version of the truth. 30 Ca sh DB Processes M pu Co E5 rc nv 9N ha er se t to or de r Ex M ec D ut 01 e M RP m Raw at er ia C l C This consisted of three parts. The first was to document the existing data environment. The second was to define a target data architecture, and finally to create a roadmap showing how to migrate from the current environment to the target architecture. CR In ve st DB m en ts 41 3 CR o DB raw ns. CR DB CR 5 t Go od sr ec ei p 40 CR B CR Ca s pu t ut n nu es Preliminary Version – send comments to pml@hec.ca h In ve s Readings on Enterprise Resource Planning tm en ts DB GR /IR DB Po s R Ac c. Pa y. C t p F-53 ay m en 40 Po M st IRO in vo ice CR /IR R co aw ns m um ate pt ria io l n IBM developed an information maturity model to establish benchmarks to measure progress. The maturity model provides a means of self-assessment to identify data quality issues, and In ve so nto ld ry pr ch od an uc ge t 6 Data governance refers to enterprise level organizational structures and processes that define and enforce data management policies, standards, and guidelines. Data definitions, business rules, and content were created for all data entities, and adherence to these standards is enforced for every new system or database. Each data element is assigned to a data steward who has the responsibility of maintaining data quality, and ensuring that data guidelines are followed. GR CR m Ra w DB sa Do le m re es ve tic nu es Maturity Model 40 Data Governance at Ensuring high quality data requires an enterprise approach to data management. This involves establishing enterprise level data governance, standardizing data definitions, and instituting global data management processes. Figure 10 shows that BI initiatives are much more successful in organizations in which data is considered an enterprise asset. The IBM internal enterprise architecture program provides a model for organizations who want to manage data as an enterprise asset (Vayghan, et. al., 2007). The goal of the program is to encourage the creation and use of enterprise data instead of local data. The program consists of six elements : Po M s IG re t go O ce o ip ds ts In re dep qu e ire nd m en en t t Pu r or cha de se r IBM established standard enterprise–wide processes for the creation and management of master data entities. Master data includes the core data entities of a business such as customers, vendors, and employees. These processes were integrated into a larger master data management (MDM) effort that includes technology and processes to manage master data. MDM establishes a single system of record for every master data element. Data elements are created only once to avoid duplication, and formalized update processes ensure the data remains correct. DB In v ch en DB an tor ge y Enterprise Data Architecture Program Data management involves improving the quality of existing data, as well as establishing data management processes and governance structures to ensure that data quality remains high. If they are starting from scratch, organizations should expect to spend up to two years or longer implementing a data management program and bringing data up to the required level of quality (Davenport & Harris, 2010). Yet, trying to eliminate all data errors in all of an organization’s data is unrealistic and unnecessary. The level of data quality must match the requirements of the application. Ironically, basic BI applications such as static reports may have the highest quality standards, because nothing will undermine the credibility of a BI application like bad data. If data is inaccurate, or if values are missing, users will lose faith in the application. More advanced analytical techniques have methods to deal with problems such as missing values or outliers. Analytical competitors don’t have pristine data, but they have data of sufficient quality for the analyses in the areas that really matter. If customer service is the company’s unique advantage, then they have high quality customer data. They also have established processes for ensuring the quality of customer data remains high, and for fixing quality problems that arise. Start with data that is required for analysis, and then keep moving. Erickson (2006) found that projects in organizations in which data was “Definitely” viewed as an enterprise asset were six times more likely to be succeeding, than in organizations in which data was “Not Really” viewed as an enterprise asset. Cr ea VL te 01 de N liv er y focuses on actions that can be taken to address the issues. The model also establishes a common framework for discussion about data within IBM. DB Fi pr nis od he uc d t Address data, for instance, is notoriously difficult to maintain because it changes relatively frequently, and is often stored in multiple systems. In many organizations, the ownership of data is ambiguous, at best, so there is no one to take responsibility for ensuring the quality of data. M pr od Re D0 l uc e a 7 tio se pr Ma CO od ss 4 uc re 1 tio lea n se or de r CR P CR Fi pr nis h o DB d ed uc ts Hightower – The analytical organization ut o ro DB utp d. R 41 0 Ac c. Pa y. C Fa c pr tory od o uc ut tio pu n t CR Re ve nu e DB ec . DB CR Ac c. R CHAPTER 15 204 DB R 1 41 CR DB The first step is to establish and communicate a vision of what BI means to the organization. This must come from business leaders. If BI is implemented at the enterprise level, then the vision must come from the CEO. If initial efforts are within a business unit, then the unit’s top business management must set the tone. The vision should be expanded into a fullfledged effort to change the organizational culture. Many of the analytical capabilities described in this chapter are grounded in organizational culture. Cultural change is difficult and requires a sustained and organized change management effort. Change management can be institutionalized within a BI Competency Center. 15.5 The BI Roadmap 30 Tools and applications are the access tools and the applications provided to end users. Tools allow the user to generate ad hoc reports and analysis, while applications are generally developed for specific purposes. While standards should be established when possible, it’s more important that users have the tool that best fits their requirements. In most situations, a variety of tools will be required, and these must be compatible with the underlying BI infrastructure. 5 C pr Con O15 od fi uc r m tio n In ve st DB m Cr ea VL te 01 de N liv er y Change Management CR en ts 41 3 Ca sh New ways of using and managing data requires new skills. IBM trained employees in the skills necessary to implement the enterprise data architecture program. Tools & Applications CR Pr od or uct de io r n DB C CR Organizations should understand that effectively using BI takes a long term commitment, and that becoming an analytical competitor requires extensive organizational change. At the outset, the organization should begin to lay the groundwork that allows them to leverage BI investments. The following three initiatives should begin as soon as possible : CR Develop Analytical Talent If they are in the early stages of maturity, organizations must have both a short term and a long term plan. In the short term, it’s important to prove the value of BI with one or two successful applications. Early successes will help to overcome resistance and ensure funding for long term programs. At the same time, they must begin to lay the foundation necessary for long term success. This is true whether the impetus is coming from the enterprise management or within a single department. If upper management, including the CEO, are highly committed, then the organization can begin to lay the foundation enterprise wide. However, if early sponsors come from a business unit then that unit can follow the same path locally. The next two sections describe the processes required to lay the long term foundation and steps required to achieve success. Data Management BI is more likely to succeed in organizations that have a high percentage of analytically oriented employees. While this is never explicitly addressed by the ACA, it is implicit in many parts of the model. For example, a culture around the use of information and analytics implies that employees are able to evaluate and use the results of analytical models. The BI tools and applications capability implies that analysts and software developers are capable of developing the models and applications required. Analytical skills can be developed in current employees through training programs that are part of the change management effort. Analytical skills should also become requirements for new employees. 5 t 6 t p F-53 ay m en 40 Po M st IRO in vo ice Po s R Ac c. Pa y. C DB CR /IR B CR Ca s pu t ut n nu es Preliminary Version – send comments to pml@hec.ca h In ve s Readings on Enterprise Resource Planning tm en ts DB GR In ve so nto ld ry pr ch od an uc ge t Go od sr ec ei p CR /IR DB GR at CR m Ra w DB Another long term program that should begin early is a data management program. The program should establish the governance structures and processes required to manage data as an enterprise asset, and address data quality problems. While laying the long term foundation for enterprise management of high quality data throughout the organization, initial efforts should focus on the data that supports the projects designated as short term wins. R co aw ns m um ate pt ria io l n sa Do le m re es ve tic nu es 40 Po M s IG re t go O ce o ip ds ts In re dep qu e ire nd m en en t t Pu r or cha de se r M pu Co E5 rc nv 9N ha er se t to or de r Ex M ec D ut 01 e M RP The steps an organization takes to become more analytical depend, of course, on which stage of BI maturity they have already obtained. One of the first things an organization must do is conduct a self-assessment. There is more than one instrument available for this purpose. The TDWI BI Maturity Assessment Tool is available through the web site of The Data Warehouse Institute (TDWI). This tool allows you to assess your own organization’s readiness, and to compare your organization with others based on industry, company size, BI budget, and other variables. Another assessment tool was developed by Larsson and Lundgren (2009) to accompany their Analytical Capability Assessment framework. These tools help organizations identify gaps and determine where to focus their efforts. 40 In v ch en DB an tor ge y m Raw at er ia C l o DB raw ns. Eliminating data silos required creating a culture that views data as an enterprise asset, in which data is shared openly, and adherence to data management policies and standards is expected. Skills 15.5.1 Laying The Foundation DB Fi pr nis od he uc d t Culture M pr od Re D0 l uc e a 7 tio se pr Ma CO od ss 4 uc re 1 tio lea n se or de r DB CR P CR Fi pr nis h o DB d ed uc ts Hightower – The analytical organization ut o ro DB utp d. R 41 0 Ac c. Pa y. C Fa c pr tory od o uc ut tio pu n t CR Re ve nu e DB ec . DB CR Ac c. R CHAPTER 15 205 DB DB R m Raw at er ia C l C CR Pr od or uct de io r n DB 1 41 CR Roll it Out Enterprise-Wide After 3-5 years the organization should be ready to expand BI to the entire enterprise. Early BI wins demonstrate the value of BI to the organization, and allow the BICC to gain experience and develop skills. At the same time, long term programs have had time to mature, laying the long term foundation. Data, especially in strategically important areas, is of sufficient quality. A critical mass of analytically aware employees and analysts are available. It is still important to prove the value of BI, but the risks for larger projects with bigger payoffs are acceptable. A methodology like the BI Opportunity Analysis should be integrated into the organization’s strategic planning process to identify a portfolio of high value BI projects. To mitigate risks, the project should be implemented where it is most likely to succeed. Obviously, a business unit with many of the capabilities in the ACA should be chosen if possible. For example, a marketing/CRM group may already have a group of analysts, and may already be comfortable with the use of information and reporting tools. However, the most important capability at this early stage is leadership (Davenport & Harris, 2010). Even under ideal circumstances there will be many who question to value of BI, and obstacles will appear. Strong, influential sponsorship is critical to overcoming these obstacles. M pu Co E5 rc nv 9N ha er se t to or de r Ex M ec D ut 01 e M RP 5 C pr Con O15 od fi uc r m tio n Cr ea VL te 01 de N liv er y While taking a long term view, it is also very important to build momentum and engender support with short term wins. A short time frame may make it difficult to perform a full BI opportunity analysis, but an effort should be made to choose high value projects that can be implemented with minimum risk. The benefits of initial projects should affect the bottom line and be highly visible. 30 en ts CR In ve st DB m DB Ca sh CR 41 3 Create some Short Term Wins CR o DB raw ns. Once the value of BI has been demonstrated, there should be sufficient support for the establishment of a BICC. The role of the BICC initially will be to choose technologies, build the business/ IT partnership, educate, and build skills. The BICC also serves as a central organizing point for analytical talent scattered throughout the organization. Most organizations have pockets of analysts. Creating a BICC is a way of pooling the talent of these isolated groups. Establishing the BICC early also helps lay long term foundations for the BI architecture and change management. Detailed paths to analytical excellence can be found in Davenport & Harris (2007, 2010) and Williams & Williams (2007) among other places. Many of the steps described in those sources can be lumped into the three broad phases described below. The phases are designed to build momentum, while the long term programs allow the organization to develop BI maturity. Po M st IRO in vo ice CR DB 40 CR t p F-53 ay m en Po s CR B CR Ca s pu t ut n nu es Preliminary Version – send comments to pml@hec.ca h In ve s Readings on Enterprise Resource Planning tm en ts DB GR /IR DB In ve so nto ld ry pr ch od an uc ge t R Ac c. Pa y. C R co aw ns m um ate pt ria io l n sa Do le m re es ve tic nu es DB Ra w m at GR /IR 40 6 Go od sr ec ei p t Po M s IG re t go O ce o ip ds ts In re dep qu e ire nd m en en t t Because of the short time frame, data management initiatives will not have had time to ensure the availability of high quality data. If the organization does not already have data of sufficient quality, it may be possible to use external data (Larsson and Lundgren, 2009) or data from a corporate web site, which is often higher quality because it is collected automatically. If the technical infrastructure is not available ,then there are cloud based options available that allow development of simple analytical applications quickly at relatively low cost. The key is to move quickly on a project that will pay off. 5 Pu r or cha de se r Although the road to becoming an analytical competitor is long with many obstacles, organizations will gain many benefits along the way. Few companies are able to, or should become, full-fledged analytical competitors, but almost every organization would benefit from acquiring the capabilities described in this chapter. 40 In v ch en DB an tor ge y Establish a BICC DB Fi pr nis od he uc d t 15.5.2 Path to Success M pr od Re D0 l uc e a 7 tio se pr Ma CO od ss 4 uc re 1 tio lea n se or de r CR P CR Fi pr nis h o DB d ed uc ts Hightower – The analytical organization ut o ro DB utp d. R 41 0 Ac c. Pa y. C Fa c pr tory od o uc ut tio pu n t CR Re ve nu e DB ec . DB CR Ac c. R CHAPTER 15 206 DB Hightower – The analytical organization DB pr Ma CO od ss 4 uc re 1 tio lea n se or de r R M pr od Re D0 l uc e a 7 tio se CR C CR o DB raw ns. DB Fi pr nis od he uc d t References m Raw at er ia C l CR Fi pr nis h o DB d ed uc ts CR P ut o ro DB utp d. R 41 0 Ac c. Pa y. C Fa c pr tory od o uc ut tio pu n t CR Re ve nu e DB ec . DB CR Ac c. R CHAPTER 15 Pr od or uct de io r n DB In v ch en DB an tor ge y Ayres, I. 2007. Supercrunchers : Why Thinking-by-Numbers Is the New Way To Be Smart, New York : Bantam Dell. CR Blackman, S. 2010. “Are Analytics Responsible for the Financial Crises?,” BNET.com. Retrieved from http ://www.bnet.com/blog/mba/are-analytics-responsible-for-the-financial-crisis/1834. 1 41 CR 41 3 Ca sh Anonymous. 2008. “Business Intelligence : Bright Spot in IT Investment,” Computer Economics, February. Retrieved from http ://www.computereconomics.com/article.cfm?id=1406. DB 5 C pr Con O15 od fi uc r m tio n CR Cr ea VL te 01 de N liv er y In ve st DB m en ts Anonymous. 2010. “Data, Data Everywhere,” Economist.com, publisher : London. Retrieved from http ://www.economist.com/node/15557443?story_id=15557443 30 Davenport, T., and Harris, J. 2010. Analytics at Work : Smarter Decisions, Better Results, Boston : Harvard Business School Publishing. Davenport, T., and Harris, J. 2007. Competing on Analytics : The New Science of Winning, Boston : Harvard Business School Publishing. Erickson, W. 2006. Performance Dashboards : Measuring, Monitoring, and Managing your Business, Hoboken, NJ : John Wiley & Sons, Inc. Hammer, M. 2007. “The Process Audit,” Harvard Business Review (85:4), April, pp. 1-16. Henschen, D. 2009. “Gartner Business Intelligence Summit : Donald Feinberg on the CIO’s View of BI,”. InformationWeek. Retrieved from http ://www.informationweek.com/news/software/bi/215801184?pgno=1. M pu Co E5 rc nv 9N ha er se t to or de r Ex M ec D ut 01 e M RP Larsson, T., and Lundgren, R. 2009. The Power of Knowing : A Case Study on Data Driven Management, Lund University, Lund, Sweden. Retrieved from http ://www.fek.lu.se/supp/supp_download.asp?EB_iid={CBFFB5F5-EF15-462A-9358-8AD845DBD559}&id=4078&filename=F EK-00013826.pdf Laursen, G., and Thorlund, J. 2010. Business Analytics for Managers : Taking Business Intelligence Beyond Reporting, Hoboken, NJ : John Wiley & Sons, Inc. 40 5 Pu r or cha de se r Muhammad Ali Khan, A., Amin, N., and Lambrou, N. 2009. “Driver and Barriers to Business Intelligence Adoption : A Case of Pakistan,” European and Mediterranean Conference on Information Systems 2010, Abu Dhabi, UAE. Retrieved from http ://www.iseing.org/emcis/EMCIS2010/Proceedings/Accepted%20Refereed%20Papers/C81.pdf on November 18, 2010. Po M s IG re t go O ce o ip ds ts In re dep qu e ire nd m en en t t Paulk, M., Curtis, B., Chrissis, M., and Weber, C. 1993. Capability Maturity ModelSM for Software, Version 1.1, Software Engineering Institute, Carnegie Mellon University. Retrieved from http ://www.sei.cmu.edu/reports/93tr024.pdf. Rettig, C. 2007. “The Trouble with Enterprise Software,” MIT Sloan Management Review (49:1), Fall, pp. 20-27. 40 6 Go od sr ec ei p t Sommer, D., and Sood, B. 2010. Market Share : Business Intelligence, Analytics and Performance Management Software, Worldwide, 2009, (Gartner Newsroom Press Release). Stamford, CT : Gartner, Inc. Retrieved from http ://www.gartner.com/it/page.jsp?id=1357514. Po M st IRO in vo ice CR DB GR /IR Vayghian, J.A., Garfinkle, S.M., Walenta, C., Healy, D.C., and Valentin, Z. 2007. “The Internal Information Transformation of IBM,” IBM Systems Journal (46:4), October-December, pp. 669-683. 40 CR t p F-53 ay m en Po s CR B CR Ca s pu t ut n nu es Preliminary Version – send comments to pml@hec.ca h In ve s Readings on Enterprise Resource Planning tm en ts DB GR /IR DB In ve so nto ld ry pr ch od an uc ge t R Ac c. Pa y. C R co aw ns m um ate pt ria io l n sa Do le m re es ve tic nu es DB Ra w m at Williams, S. and Williams, N. 2007. The Profit Impact of Business Intelligence, Amsterdam : Elsevier. 207 DB Hightower – The analytical organization DB pr Ma CO od ss 4 uc re 1 tio lea n se or de r 1 41 www.sap.com www.invensys.com I2 www.i2.com Microsoft www.microsoft.com Epicor www.epicor.com 30 Invensys 5 C pr Con O15 od fi uc r m tio n www.oracle.com CR Cr ea VL te 01 de N liv er y In ve st DB m Oracle DB en ts SAP AG, Inc. CR 41 3 Ca sh CR Pr od or uct de io r n 1. The capabilities included in the ACA can be augmented to varying degrees by the right technologies. ERP vendors offer a range of products that organizations might find useful. Review the products offered by an ERP vendors and analyze how their products would support the capabilities of the ACA. You can choose one of the following or select another. M pr od Re D0 l uc e a 7 tio se C CR o DB raw ns. CR DB In v ch en DB an tor ge y Questions DB Fi pr nis od he uc d t R m Raw at er ia C l CR Fi pr nis h o DB d ed uc ts CR P ut o ro DB utp d. R 41 0 Ac c. Pa y. C Fa c pr tory od o uc ut tio pu n t CR Re ve nu e DB ec . DB CR Ac c. R CHAPTER 15 2. The Data Warehouse Institute (TDWI) provides a BI readiness assessment tool. To find the tool, search for “assessment” at tdwi.org. Use the tool to evaluate your current organization, or an organization you’ve been associated with in the past. What steps would you suggest for the evaluated organization to increase its BI maturity ? 3. Locate the website of a BI software as a service (SaaS) solution provider. Review the website and determine what services they offer. Do they offer both reporting and analytical tools? What types of data connectivity options do they offer ? M pu Co E5 rc nv 9N ha er se t to or de r Ex M ec D ut 01 e M RP 4. Avinash Kaushik, Analytics Evangelist at Google, says of BI programs, “Never, never make it an IT matter…they don’t have the power to drive change” (as cited in Larsson & Lundgren, 2009). Do you believe this is correct? Why or why not ? Of all the analytical capabilities described in this chapter, which ones should be primarily the responsibility of IT departments ? 5 40 6. Thomas Davenport has stated that analytics played a significant role in the financial crises of 2008 (Davenport & Harris, 2010). In an interview for BNET.com (Blackman, 2010), he said : Po M s IG re t go O ce o ip ds ts In re dep qu e ire nd m en en t t Pu r or cha de se r 5. Another model of BI maturity is the BI Maturity Model developed by Erickson (2006). Review the model at http://knowledgeworks.wordpress.com/2009/02/17/the-tdwi-bi-maturity-model/. Compare and contrast this model with Davenport & Harris’ (2007) maturity model described in this chapter. DB 6 40 Po M st IRO in vo ice CR 40 sa Do le m re es ve tic nu es CR DB Ra w m at GR /IR Go od sr ec ei p t “It’s a very critical issue for business education. We basically have created two classes of people: quants and non-quants. They didn’t communicate very well in a number of companies during the financial crisis. The quants could develop these financial algorithms that made it look highly desirable to enter into a series of complicated derivatives and things like that. The non-quants — who tend to be senior managers — didn’t understand that, didn’t understand the assumptions behind them and didn’t know when the world was changing, so as to make those algorithms invalid. The quants haven’t been good at explaining what they’re doing in simple terms. Non-quants haven’t been diligent enough to delve into the assumptions behind the models and to know when they work and when they don’t. As a result, we have a really severe recession and a number of firms went out of business. So the stakes are high.” t p F-53 ay m en Po s CR B CR Ca s pu t ut n nu es Preliminary Version – send comments to pml@hec.ca h In ve s Readings on Enterprise Resource Planning tm en ts DB GR /IR DB In ve so nto ld ry pr ch od an uc ge t R Ac c. Pa y. C R co aw ns m um ate pt ria io l n Are there any analytical capabilities described in the ACA that may have prevented this failure ? 208
Purchase answer to see full attachment