11

KNOWLEDGE MANAGEMENT, BUSINESS INTELLIGENCE, AND ANALYTICS

Caesars Entertainment Corporation, the largest gaming company in the world by some measures, found a way to more than double revenues by collecting and then analyzing customer data. According to CEO Gary Loveman, “We've come out top in the casino wars by mining our customer data deeply, running marketing experiments, and using the results to implement finely tuned marketing and service delivery strategies that keep our customers coming back.”¹ This is more than just implementing loyalty cards to track customer activity and reward “frequent buyers.” In 2000, the Harrah's brand was valued at close to $3 billion. When it was sold 7 years later to a private equity group, it was valued at $17 billion. Much of that increase was credited to the innovative and widespread use of business analytics to turn around the gaming company. In 2010, the company changed its name from Harrah's to Caesars Entertainment Corporation.

Analytics at Caesars begins when a customer is issued a loyalty card in the Total Rewards (TR) program. Similar to the ubiquitous cards used by airlines, grocery stores, and even coffeehouses, the TR card tracks customer usage of the various games offered in their casinos. What differentiates the TR card is what Caesars does with the information they collect when customers use the card. Management uses sophisticated analytical tools to understand as much as possible about their customers. For example, they thought their best customers were high rollers. In fact, they found that 82% of revenues came from 26% of customers, and they were not the stereotypical gold cufflink-wearing, limousine-riding high rollers, but retirees who have time to spend their nest egg. The management wanted to know what motivated these customers. They conducted experiments and focus groups, using well-structured experiments designed to gather data and test hypotheses. They found that these customers were motivated by reduced rates on hotel rooms, or if they lived in the area, free chips. Special gifts and expensive rooms were not as effective as incentive.

They studied the customer's value over time and identified ways to increase spending on repeat visits. For example, when they looked at the data about their best customers, they learned that these customers wanted service quickly. So Caesars management found ways to reduce the wait at the valet parking lot and at the restaurants. Diamond customers, those that were the very best customers, rarely waited in line at all, providing a very visible “reward” for their business and motivating others to seek Diamond-level status (something they could earn through the TR program). They studied individual behaviors and created a program that was custom tailored to each customer offering specific incentives based on the results of their analytical models. As Loveman described, “If we discovered that a customer who spends $1,000 per month with us hadn't visited us in three months, a letter or telephone call would invite him back. If we learned that he lost money during his last visit, we invited him back for a special event.”² They found ways to keep the small-level gamblers in the casino longer and to lure them back again at very low costs. By understanding the limit a customer normally spends in a casino, management was able to identify when a customer was about to leave the casino, and intervene, offering him a complimentary dinner or other incentive to stay in the casino. Analytics drives their business, and the results have turned the company into a model for successfully integrating technical algorithms with marketing techniques.

As baby-boomers age, the Caesar's management team began studying the next generation of potential gamblers. By 2015, Caesars estimates that 52% of spending in Las Vegas will come from twenty- to forty-somethings.³ This means revisiting the way these gamblers spend money when they gamble, and what their preferences are. Their first experiment is Linq, an entertainment district with shops, nightclubs, bars, restaurants, and comfortable spaces to meet up scheduled to open in 2013. The prototypical Linq customer, executives say, isn't a graying slot player but rather, a thirty-something, middle-class man or woman who wants to meet up with friends for cocktails or beers. Linq is another example of Caesar's Entertainment's business intelligence at work, this time to create an experience aimed at the changing demographic of their customer base.

This chapter provides an overview of some of the ways business make decisions. Enterprises have long sought a way to harness the value locked inside the extensive data they collect and store about customers, markets, competitors, products, people, and processes. This chapter will review some of the basic concepts of knowledge management, business intelligence, analytics, and the concept of big data.

KNOWLEDGE MANAGEMENT, BUSINESS INTELLIGENCE, AND BUSINESS ANALYTICS

It's all about making better decisions. Managing knowledge is not a new concept,⁴ but it has been invigorated and enabled by new technologies for collaborative systems, the emergence of the Internet and intranets, which in themselves act as a large, geographically distributed knowledge repository, and the well-publicized successes of companies using business analytics, like Caesars. The discipline draws from many established sources, including anthropology, cognitive psychology, management, sociology, artificial intelligence, IT, and library science. Knowledge management remains, however, an emerging discipline, with few generally accepted standards or definitions of key concepts.

Knowledge management includes the processes necessary to generate, capture, codify, and transfer knowledge across the organization to achieve competitive advantage. Individuals are the ultimate source of organizational knowledge. The organization gains only limited benefit from knowledge isolated within individuals or among workgroups; to obtain the full value of knowledge, it must be captured and transferred across the organization.

Business intelligence (BI) is the term used to describe the set of technologies and processes that use data to understand and analyze business performance.⁵ It is the management strategy used to create a more structured approach to decision making based on facts that are discovered by analyzing information collected in company databases. Although some may argue with this relationship, business intelligence can be considered a component of knowledge management. Knowledge management deals with the processes necessary to capture, codify, and make sense of all types of knowledge as described earlier. Business intelligence is more specifically about extracting knowledge from data. Davenport and Harris suggest that business analytics is the term used to refer to the use of quantitative and predictive models and fact-based management to drive decisions. By this definition, business analytics is a subset of BI. Some, however, use the terms BI and analytics interchangeably.

The most profound aspect of knowledge management and business intelligence is that, ultimately, an organization's only sustainable competitive advantage lies in what its employees know and how they apply that knowledge to business problems. Exaggerated promises and heightened expectations, couched in the hyperbole of technology vendors and consultants, may create unrealistic expectations. Knowledge management is not a magic bullet, that is, an appropriate solution for all business problems. While reading this chapter, managers should consider the implications of managing knowledge, but should not believe that knowledge management by itself is the sole answer for managerial success. Knowledge must serve the broader goals of the organization, and analytics alone do not create competitive advantage. How the information is used and how the knowledge is linked back to business processes are important components of knowledge management.

Intellectual Property

Two other terms frequently encountered in discussions of knowledge are intellectual capital and intellectual property. Intellectual capital is defined as knowledge that has been identified, captured, and leveraged to produce higher-value goods or services or some other competitive advantage for the firm. Both knowledge management and intellectual capital are often used imprecisely and interchangeably to describe similar concepts. Information technology (IT) provides an infrastructure for capturing and transferring knowledge, but does not create knowledge and cannot guarantee its sharing or use.

Intellectual property allows individuals to own their creativity and innovation in the same way that they can own physical property. However, when the intellectual property is information-based, it differs from physical property in two important ways. First, information-based property is non-exclusive to the extent that when one person uses it, it can be used again by another person. Consider an MP3 file of music, which can be easily copied and shared with another without loss of the original property. Second, unlike the cost structure of physical property, the marginal cost of producing additional copies of information-based property is negligible compared with the cost of original production. These differences create differences in the ethical treatment of physical and information-based intellectual property. The economics of information versus the economics of physical property is further explored in the introduction of this text.

The concept of intellectual property makes it possible for owners to be rewarded for the use of their ideas and it allows them to have a say in how their ideas are used. To protect their ideas, owners typically apply for and are granted intellectual property rights, although some protection such as copyright arises automatically, without any registration, as soon as a record is made in some form of what has been created.

The four main types of intellectual property are patents for inventions, trademarks for brand identity, designs for product appearance, and copyrights for literary and artistic material, music, films, sound recordings, broadcasts, and software.⁶ In 2002, the music sharing Web site Napster raised controversial issues long surrounding the practice of copyright. The Audio Home Recording Act (1992) was passed in the United States to prevent serial copying, but this didn't seem to apply to Napster, who only facilitated sharing. Although the act protected intellectual property, it also confirmed the freedom to copy music for personal use.

In 1998, the more stringent Digital Millennium Copyright Act (DCMA) passed by a unanimous vote in the U.S. Senate with the active support of the entertainment industry.⁷ The DCMA makes it a crime to circumvent copy protection, even if that copy protection impairs rights established by the Audio Home Recording Act. Furthermore, the Digital Tech Corps Act of 2002, passed in the U.S. House of Representatives, seeks to protect intellectual property by placing a lifetime ban on employees from revealing trade secrets, and imposing a criminal penalty of up to five years in prison and a $50,000 fine.⁸ A senior-level position, Coordinator for International Intellectual Property Enforcement in the U.S. Department of Commerce, was created to coordinate the battle against global piracy of intellectual property.

The U.S. Congress continues to propose and discuss ways to protect intellectual property, particularly from piracy of online materials by sites and companies outside of U.S. jurisdiction. For example, though it was soundly rejected by the public and by Web sites around the world, the Stop Online Piracy Act (SOPA) and the Protect IP Act (PIPA) were introduced to the U.S. Congress in 2011 at the behest of the entertainment industry to protect intellectual property. House Judiciary Committee Chairman Lamar Smith (R-TX) postponed plans to draft a compromise bill, the Online Protection and Enforcement of Digital Trade Act (OPEN). He commented that “The committee remains committed to finding a solution to the problem of online piracy that protects American intellectual property and innovation. . . The House Judiciary Committee will postpone consideration of the legislation until there is wider agreement on a solution.”⁹

DATA, INFORMATION, AND KNOWLEDGE

The terms data, information, and knowledge are often used interchangeably, but have significant and discrete meanings within the knowledge management domain. As was first presented in the introduction of this textbook, the differences are shown in Figure 11.1. Data are specific, objective facts or observations, such as “inventory contains 45 units.” Standing alone, such facts have no intrinsic meaning, but can be easily captured, transmitted, and stored electronically.

Information is defined by Peter Drucker as “data endowed with relevance and purpose.”¹⁰ People turn data into information by organizing them into some unit of analysis (e.g., dollars, dates, or customers). Deciding on the appropriate unit of analysis involves interpreting the context of the data and summarizing them into a more condensed form. Consensus must be reached on the unit of analysis.

FIGURE 11.1 The relationships between data, information, and knowledge.

Source: Adapted from Thomas H. Davenport, Information Ecology (New York: Oxford University Press, 1997), 9.

Knowledge is a mix of contextual information, experiences, rules, and values. It is richer and deeper than information and more valuable because someone has thought deeply about that information and added his or her own unique experience, judgment, and wisdom. One way of thinking about knowledge is to consider the different types of knowing.¹¹ Knowing what often is based on assembling information and eventually applying it. It requires the ability to recognize, describe, and classify concepts and things. The process of applying knowledge helps generate knowing how to do something. This kind of knowing requires an understanding of an appropriate sequence of events or the ability to perform a particular set of actions. Sometimes the first inkling of knowing how to do something stems from an understanding of procedures, routines, and rules. Knowing how to do something often begins with following procedures and is fully learned by actually experiencing a situation. Finally knowing how and knowing what can be synthesized through a reasoning process that results in knowing why. Knowing why is the causal knowledge of why something occurs. Often reasoning applied to knowing-how can lead to the understanding of knowing-why. These types of knowing are modeled in Figure 11.2.

FIGURE 11.2 Taxonomy of knowledge.

Source: Adapted from H-W. Kim and S. M. Kwak, “Linkage of Knowledge Management to Decision Support: A System Dynamics Approach,” presented at the National University of Singapore (July 2002).

Values and beliefs are also a component of knowledge; they determine the interpretation and the organization of knowledge. Tom Davenport and Larry Prusak, experts who have written about this relationship, say, “The power of knowledge to organize, select, learn, and judge comes from values and beliefs as much as and probably more than, from information and logic.”¹² Knowledge also involves the synthesis of multiple sources of information over time.¹³ The amount of human contribution increases along the continuum from data to information to knowledge. Computers work well for managing data, but are less efficient at managing information. The more complex and ill-defined elements of knowledge (for example, “tacit” knowledge, described later in this chapter) are difficult if not impossible to capture electronically.

Although knowledge has always been important to the success of an organization, it was presumed that the natural, informal flow of knowledge was sufficient to meet organizational needs. But managing knowledge has become far more complex, the amount of knowledge to manage far greater than every, and the tools to manage knowledge far more powerful. Managing knowledge provides value to organizations in several ways, as summarized in Figure 11.3.

Tacit versus Explicit Knowledge

Knowledge can be further classified into two types: tacit and explicit. Tacit knowledge was first described by philosopher Michael Polyani in his book, The Tacit Dimension, with the classic assertion that “We can know more than we can tell.”¹⁴ For example, try writing a memorandum, or even explaining verbally, how to swim or ride a bicycle. Tacit knowledge is personal, context-specific, and hard to formalize and communicate. It consists of experiences, beliefs, and skills. Tacit knowledge is entirely subjective and is often acquired through physically practicing a skill or activity.

FIGURE 11.3 The value of managing knowledge.

In 2011, quarterback Drew Brees broke the NFL single-season record for the most passing yards with 5,476 yards passed. It would be nearly impossible to verbally describe all the factors that Brees had to consider when making those passes, yet he knew who to throw the ball to, where to put the ball, and why to make that throw, all in a matter of seconds. Brees' ability to pass the football incorporates so much of his own personal experience and kinesthetic memory that it is impossible to separate that knowledge from the player himself. His bone structure, muscular development, and the nerves between his arm and his brain all make it possible for him to throw the types of passes he does.

FIGURE 11.4 Examples of explicit and tacit knowledge.

IT has traditionally focused on explicit knowledge, that is, knowledge that can be easily collected, organized, and transferred through digital means, such as a memorandum or financial report. Individuals, however, possess both tacit and explicit knowledge. Explicit knowledge, such as the knowledge gained from reading this textbook, is objective, theoretical, and codified for transmission in a formal, systematic method using grammar, syntax, and the printed word. Figure 11.4 summarizes these differences.

Knowledge conversion strategies are often of interest in the business environment. Companies often want to take an expert's tacit knowledge and make it explicit, or to take explicit, book-learning in their new hires and make it tacit. In their book The Knowledge-Creating Company, Ikujiro Nonaka and Hirotaka Takeuchi describe four different modes of knowledge conversion (see Figure 11.5). The modes are (1) from tacit knowledge to tacit knowledge, called socialization, (2) from tacit knowledge to explicit knowledge, called externalization, (3) from explicit knowledge to explicit knowledge, called combination, and (4) from explicit knowledge to tacit knowledge, called internalization.¹⁵ Socialization is the process of sharing experiences; it occurs through observation, imitation, and practice. Common examples of socialization are sharing war stories, apprenticeships, conferences, and casual, unstructured discussions in the office or “at the water cooler.”

FIGURE 11.5 The four modes of knowledge conversion.

Source: Ikujiro Nonaka and Hirotaka Takeuchi, The Knowledge-Creating Company: How Japanese Companies Create the Dynamics of Innovation (New York: Oxford University Press, 1995), 62. By permission of Oxford University Press, Inc.

KNOWLEDGE MANAGEMENT PROCESSES

Knowledge management involves four main processes: the generation, capture, codification, and transfer of knowledge. Knowledge generation includes all activities that discover “new” knowledge, whether such knowledge is new to the individual, the firm, or the entire discipline. Knowledge capture involves continuous processes of scanning, organizing, and packaging knowledge after it has been generated. Knowledge codification is the representation of knowledge in a manner that can be easily accessed and transferred. Knowledge transfer involves transmitting knowledge from one person or group to another, and the absorption of that knowledge. Nonaka's knowledge framework above in Figure 11.5 displays a form of knowledge transfer. Without absorption, a transfer of knowledge does not occur. Generation, codification, and transfer generally take place constantly without management intervention. Knowledge management systems seek to enhance the efficiency and effectiveness of these activities and leverage their value for the firm as well as the individual. But with the increasing introduction of new and more robust systems for managing and using knowledge, knowledge management processes are a dynamic and continuously evolving.

Knowledge management processes are different in the age of Web 2.0 and robust search tools such as Google. Whereas traditional knowledge management systems had well defined processes for generation, capture, codification and transfer, technologies such as large data warehouses, ubiquitous Web sites, search tools, and tagging made it possible to capture and find information without the formal processes. Tagging, where users themselves list key words that codify the information or document at hand, creates an ad-hoc codification system, sometimes referred to as a folksonomy. Search engines have changed the way information is accessed, making it possible to quickly find virtually anything on any system connected to the Internet. These technologies have replaced traditional knowledge management systems and given individuals the ability to find information that traditionally was locked within structures that had to be designed, managed, and then taught to users.

BUSINESS INTELLIGENCE

Traditional business intelligence (BI) has been associated with providing dashboards and reports to assist managers in monitoring key performance metrics. Common elements of BI systems include reporting, querying, dashboards, and scorecards. Dashboards tend to be a simple, online display of key metrics, often graphically displayed in pie charts, bar charts, red-yellow-green coded data, and other images that easily convey both the value of the metric and, with the color coding, if the metric is within acceptable parameters or not. In one example, a map of the United States was used to indicate sales performance by geography, and each state was color coded to indicate if targets were being met. Managers could drill down into each region by clicking on the state, and see the next level of detail, which provided information by region. Further drilling down indicated sales by city and ultimately by sales person. At each level, the data was presented and color coded to give a visual, and therefore quick indication of who was making targets, and who was missing them. Traditional BI is useful for strategic, tactical and operational decisions. Chapter 7 describes how dashboards and scoreboards are used for running the business of IT. The BI dashboards are similar in that they summarize information in similar ways, but the use of the BI dashboard is very different than the IT dashboard.

At the SAS Global Forum in 2010,¹⁶ a discussion ensued about what has become known as BI 2.0, or collaborative BI, the next generation of business intelligence. BI 2.0 incorporates a more proactive perspective, and provides for querying of real-time data. It incorporates a number of characteristic that are seen in Web 2.0 applications such as providing BI as a service in the cloud, rather than as a software package purchased from a vendor and installed on the organization's computer; event-driven, instant access to real-time information rather than batch, after the fact report generation; mobile and ubiquitous access rather than access just from desktop computers; and mashup capability rather than static, stand-alone systems. Newer technologies have enabled BI to move to a new level with more robust user interfaces that provide visualization and analytics tools. Crowdsourcing allows the data structures and report designs to be created by the community, rather than a single designer. Data and reports are infused with narratives from the users to provide richer context. Dynamic capabilities in the BI system provide exceptions, alerts, and notifications that change based on what the system learns from the data itself. When a manager sees something in the data that requires an intervention, he will not only be able to do the intervention, but to tag it and link it with the data so that the collective knowledge grows over time.

COMPETING WITH BUSINESS ANALYTICS

In recent years, many companies have found success competing through better use of analytics. Companies such as Caesars Entertainment, as described at the beginning of this chapter, have turned around an otherwise lackluster business to become a leader in their industry. Capital One has also emerged from a crowded field of financial services firms, to become one of the industry leaders through use of extensive business analytics to continuously create and invent new products and services to reach out to new customers and reinvigorate relationships with existing customers. In their case, the company was founded on the idea that by mining data about individual customers they could create financial service products that addressed what the big players would consider “‘niche markets,” unattractive to the larger players because of the smaller number of potential customers, but profitable nonetheless. Using the customer database of a small bank, and running numerous analytical tests, they identified characteristics that would create a profitable service. They learned, for example, that the most profitable customers were ones who charged a large amount, but paid their credit cards off slowly. At the time, most credit cards companies didn't differentiate between these and other customers. The innovative idea was to create a product that catered to these customers. Today, Capital One runs hundreds of experiments, identifying new products that target individual customers. Using analytics to simulate and test is a very low-cost way to design and develop these products.¹⁷

Sports teams have propelled themselves to league success through business analytics. The systematic use of factual data in proprietary models is credited with helping the Oakland As and the Boston Red Sox. As seen in the movie Moneyball, Billy Beane was one of the first general managers in Major League Baseball to build his organization, the Oakland As, around analytics. Although this industry collected data extensively, it was mostly used to manage the game in process. The Oakland As managed by using data on things that they could measure such as the on-base percentage (the number of times a player gets on-base), instead of softer criteria such as determination or effort the player is willing to put in. They used analytics in their recruiting efforts to predict which young players had the best chances of becoming major league players. Their strategy paid off, consistently carrying them to the playoffs despite a budget for player's salaries that was a fraction of what some of their competitors had.

One reason for the rise in companies competing on analytics is that many companies in many industries offer similar products and use comparable technologies. Therefore, business processes are among the last remaining points of differentiation, and analytic competitors are wringing every last drop of value from those processes.¹⁸ Business analytics fuel fact-based decision making. For example, a company may use inventory reports to figure out what products are selling quickly and which are moving slowly, but a company that uses analytics will also know who is buying them, what price each customer pays, how many items the customer will purchase in a lifetime, what motivates each customer to purchase, and which incentives to offer to increase the revenue from each sale.

Davenport and Harris suggest that companies who successfully compete using their business analytics skills have these five capabilities:

• Hard to duplicate: Because successfully using analytics to compete means having a strong culture and organizational support system, as well as business processes that utilize the results of the analytical analyses, copying the capability is difficult, if not impossible. A competitor may have the same tools, but success comes from how they are used.
• Uniqueness: There are many ways to use business analytics to compete. A specific business will choose a path based on their business, their strategy, their market, their competitors, and their industry.
• Adaptability: Successful companies use analytics across boundaries and in creative ways. Workers are not held back from using analytics, and in fact are encouraged to find new and innovative ways to apply their tools. By creating a culture of analytics, virtually everyone in the organization seeks applications for analytics to enhance their business operations.
• Better than competition: Some organizations are better at applying analytics than others. For example, the Oakland As and the Boston Red Sox are well known for their use of analytics in an industry, Major League Baseball, well known for its data collection and statistical analysis.
• Renewability: Agility is an important characteristic of sustainable competitive advantage. Companies who use analytics for competitive advantage are exceptionally adaptable, continuously reinvest, and constantly renew their capabilities.

COMPONENTS OF BUSINESS ANALYTICS

To successfully build business analytics capabilities in the enterprise, companies make a significant investment in their technologies, their people, and their strategic decision-making processes. Four components are needed (these four components are summarized in Figure 11.6).

Data Repositories

Data used in the analytical processes must be gathered, cleaned up, common, integrated and stored for easy access. Data warehouses, or collections of data designed to support management decision making, sometimes serve as repositories of organizational knowledge. They contain a wide variety of data used to create a coherent picture of business conditions at a single point in time. In fact, the data contained in data warehouses may represent a large part of a company's knowledge, for example, the business's knowledge about its clients and their demographics.

FIGURE 11.6 Components of business analytics.

Software Tools

At the core of business analytics are the tools. An approach that simulates business intelligence is data mining, which is the process of analyzing data warehouses for “gems” that can be used in management decision making. It identifies previously unknown relationships among data. Typically, data mining refers to the process of combing through massive amounts of customer data to understand buying habits and to identify new products, features, and enhancements. The analysis may help a business better understand its customers by answering such questions as: Which customers prefer to contact us via the Web instead through a call center? How are customers in Location X likely to react to the new product that we will introduce next month? How would a proposed change in our sales commission policy likely affect the sales of Product Y? Using data mining to answer such questions helps a business reinforce its successful practices and anticipate future customer preferences. For example, the New York Times reported that using data mining, Walmart found the unlikely fact that its Florida customers stocked up on beer and strawberry pop tarts when a hurricane was threatening. It now supplies its stores with plenty of these two items when hurricanes are on the horizon in an area.¹⁹

There are four categories of tools that are typically included under the business analytics umbrella. They include:²⁰

• Statistical Analysis—answers questions like, “Why is this happening?”
• Forecasting/Extrapolation—answers questions like, “What if these trends continue?”
• Predictive Modeling—answers questions like, “What will happen next?”
• Optimization—answers questions like, “What is the best that can happen?”

These tools are used with the data in the data warehouse to gain insights and support decision making.

Analytics Environment

Building an environment that supports and encourages analytics is a critical component. It requires aligning IS strategy and organizational strategy with the business strategy. This includes alignment of the corporate culture, the incentive systems, the metrics used to measure success of initiatives, and the processes for using analytics with the objective of building competitive advantage through analytics. For example, one financial services firm encouraged the use of analytics by changing its appraisal system so that demonstration of skills associated applying analytics was made a significant factor in compensation decisions. This is an example of aligning organizational strategy with a business strategy promoting the use of analytics to gain competitive advantage.

Although many companies have some sort of analytical tools in place, most are not used for mainstream decision making, and they certainly do not drive the strategy formulation discussions of the company. Those who gain competitive advantage from analytics use analytics as an integral component of their business.

Leadership plays a big role in creating a strong analytics environment. Leaders must move the company's culture toward an evidence-based management approach in which evidence and facts are analyzed as the first step in decision making. Those in this type of culture are encouraged to challenge others by asking for the data, and where no data is available, to experiment and learn to generate facts. Use of evidence-based management encourages decisions based on data and analysis rather than on experience and intuition.

Skilled Workforce

It's clear that to be successful with analytics, data and technology must be used. But experts point out that even with the best data and the most sophisticated analytics, people must be involved. Managers must have enough knowledge of analytics to use them in their decision making. Leaders must set examples for the organization by using analytics and requiring that decisions made by others use analytics. Perhaps the most important role is sponsorship. Davenport and Harris point out that it was the CEO-level sponsorship and the corresponding passion for analytics that enabled firms such as Caesars and Capital One to achieve the success they did.

Levels of Analytical Capabilities

All businesses have data. Some do a better job at it than others, and that can be a source of competitive advantage. Companies tend to fall into one of 5 levels of maturity with analytical capabilities. Understanding the different levels can help organization envision how to improve their capabilities to gain additional advantages. Figure 11.7 summarizes these levels.

BIG DATA

One of the impacts of our knowledge and information based economies today is the very large amount of data amassing in databases both inside companies and in the environment. Consider, for a moment, the vast amount of data Google must process every time a query is made. By some estimates there are Google tells the inquirer how many results they found, and how fast the found them. A recent query of “big data” produced “about 88,700,000 results in 0.19 seconds.” A second query of “lady gaga” produced 661,000,000 results in 0.13 seconds. Google's Web site claims to look at billions of sites with a choice of 46 languages to conduct the search.

FIGURE 11.7 Analytical capabilities maturity levels.

Sources: Adapted from S. Brobst and J. Rarey, “Five Stages of Data Warehouse Decision Support Evolution,” DSSResources.COM (January 6, 2003); and conversations with Farzad Shirzad, leader of Teradata's Center for Excellence in Analytics in 2011.

Big data is the term used to describe techniques and technologies that make it economical to deal with very large datasets at the extreme end of the scale. According to Wikipedia, big datasets are on the order of exabytes (10¹⁸ bytes, abbreviated as EB) and zettabytes (10²¹ bytes, abbreviated as ZB) of data. A megabyte, abbreviated MB, is 10⁶ bytes. Extreme datasets get that big because volumes of information are created, usually quickly, and stored for analysis. These extreme datasets create difficulties in storing, searching, sharing and analyzing; the size just cannot be handled by traditional data management tools or techniques. Having large data sets is desirable because of the potential trends and analytics that can be extracted, but when the dataset is so large that the information system cannot manage it, it's considered the “big data problem.” In those cases, specialized computers and tools are needed to help managers mine the data.

Gary King, director of Harvard's Institute for Quantitative Social Science claims that big data: “. . . is a revolution. We're really just getting under way. But the march of quantification, made possible by enormous new sources of data, will sweep through academia, business and government. There is no area that is going to be untouched.”²¹

Big data is increasingly common in part because of the rich, unstructured data streams that are created by social IT. Other examples of areas where big data problems typically occur are areas like simulations, scientific research, Internet searches, customer data management, and financial market analytics. Sensors that gather information for surveillance and sense-and-respond situations commonly create big data problems. With the growth of social IT, managers are increasingly finding that gathering all the information about their company and their customers from all the social sites available creates a data set that has the potential supply unique customer intelligence. Finding ways to manage and use the data, however, is significantly more difficult than managing more structured data sets.

Database warehouse vendors, such as Teradata. IBM and Oracle, have specially built tools for customers with big data problems. Data warehouses must be scalable, to allow capture and storage of all the data, agile, to accommodate changing requirements, mixed types of work, and quick turnaround of queries and reports, and compatible with the enterprise infrastructure to integrate with business applications and provide appropriate accessibility, backup and security.

There is a “dark side” to big data. The intense number crunching is likely to yield a number of “false discoveries.” Any results should be questioned before they are applied. Further, extensive analysis might yield a correlation and lead to a statistical inference that is unfair or discriminatory. Finally, Big Data might offer a high tech twist to an old practice “I know what the facts are—now let's find the ones we want.” Here again, care must be applied when using powerful tools.²²

SOCIAL ANALYTICS

In 2011, managers saw a rise in interest in using social IT as long as there was some way to measure the value gained from the invested time and resources. A class of tools called social analytics, or social media analytics were created to address this issue, and as expected, many vendors began offering packages that provided these tools. The goal of social analytics is to measure the impact of social IT investments on the business. At issue, however, is how to analyze conversations, tweets, blogs, and other social IT data to create meaningful, actionable facts. For example, it might be relatively easy to measure the number of hits on a Web site or the number of click-throughs from a link. But what does that information really tell a manager? What action would the manager consider taking based on this type of data? Hits and click-throughs are only meaningful in context and with other data that indicate if business value was achieved. That is, they only become information when they are processed to become relevant and purposeful.

As of the writing of this chapter, social analytics is one of the key topics discussed by managers seeking to incorporate social IT into their business. Vendors such as Google Analytics and Radian6 (acquired by Salesforce.com) offer platforms with social analytics tools. For example, Radian6's platform includes tools that enable:

• Listening to the community—identify and monitor all conversations in the social Web on a particular topic or brand.
• Learning who is in the community—learn customer demographics such as age, gender, location, and other trends to foster closer relationship with community.
• Engaging people in the community—communicate directly with customers on social platforms such as Facebook, YouTube, LinkedIn, and Twitter using a single app.
• Tracking what is being said—measure and track demographics, conversations, sentiment, status, and customer voice using a dashboard and other reporting tools.

UPS, Pizza Hut, Pepsi, AMD, and Dell Computers are examples of companies with well-know case studies about their use of social analytics and monitoring tools like Radian6 for engaging and encouraging collaboration among their customers. For example, in a presentation to the Blogwell community in 2011, a UPS manager described how the company turned around its customer service efforts using social IT and social analytics.²³ UPS studied their customer service process and monitored the social Web for comments. They noticed that some customers loved them, but others had a bad experience and wrote about it on sites like Twitter and Facebook. By using social analytics platform, they identified dissatisfied customers and addressed their problems on the social platform used by the customer. This resulted in more than 1 million positive tweets about UPS and lots of public recognition for turning around their customer service process.

Google Analytics, on the other hand, is a set of social analytics tools that enable organizations to analyze their Web site. The Google Analytics site thoroughly analyses the key words used by visitors to reach a Web site, and provides statistics to help managers understand the searches potential customers use. Some features are:

• Web site testing and optimizing—to understand traffic to Web sites and to optimize the site's content and design for increasing traffic.
• Search optimization—to understand how Google sees an organization's Web site, how other sites link to the organization's site, and how specific search queries drive traffic to the organization's site.

  Social Business Lens: Social Graphs

  Ever wonder who your connections are connected to? A social graph is a pictorial representation of relationships. Tools that create social graphs look at the network of people in a community and draw a map showing all the connections. Today it has come to mean the networks of everyone on the Web.

  While social scientists created the term long ago, Facebook made it part of the popular lexicon when CEO, Mark Zuckerberg referred to it in describing his platform. Facebook is an application built on the concept that there is value connecting the social graphs of individuals to create a large, global social graph.

  Typically the diagram uses individuals as the notes of the graphs, and lines between them to indicate some type of relationship. I might be that they ‘know’ each other or ‘work at the same place’ or ‘are connected on Facebook”. Relationships can be strong, such as a close friend, or weak, such as an acquaintance. When analyzing a social graph, understanding what the lines are depicting is important.

  Social graphs are useful for applications like Facebook and LinkedIn, who utilize the connections to help individuals grow their networks. But there are many other uses for this type of analytics. Do leaders have different social graphs than those they lead? Want to effect change? Then you want to know who the influencer is in an organization or community. Need to find expertise that is outside your network? Perhaps the extended social graph of your connections has such a person. There are ways to monetize this too. Zuckerberg shared his vision “Yelp maps out the part of the graph that relates to small businesses. Pandora maps the part that relates to music. If we can take these separate maps and pull them all together, then we can create a Web that's smarter, more social, more personalized, and more semantically aware.”

  Source: Facebook: One Social Graph to Rule Them All? (April 21, 2010), http://www.cbsnews.com/stories/2010/04/21/tech/main6418458.shtml.

• Search term interest and insights—to understand the interest over time of a search term, regional interest in the term, top searches for terms of similar category, and popularity of similar terms.
• Advertising support and management—to identify the best ways to spend advertising resources for online media.

Re/Max real estate franchise network is an example of a company using social analytics like Google analytics. With franchises in 62 countries, Re/Max is a leading provider of residential, commercial, referral, relocation, and asset management. As part of their online strategy, Re/Max created a site that listed all properties available, whether listed by Re/Max agents or others, and made it available to anyone accessing the site. They then used Google Analytics to understand consumer behavior on the site and to drive leads to agents in their franchises. Prior to this strategy, they used focus groups to understand consumer behavior, but they were expensive, limited in scope, and lacking in real data. The site gets more than 2 million hits a month, mostly from visitors who searched for “remax” in their query. Google Analytics helped managers redesign the Web site so the most used tools were on the home page, further providing value to potential customers. Ultimately, Google Analytics helped Re/Max drive more leads to agents, reducing the cost agents were used to paying for leads.²⁴

CAVEATS FOR MANAGING KNOWLEDGE AND BUSINESS INTELLIGENCE

Following such a broad survey, it seems appropriate to conclude with a few caveats. First, recall that knowledge management and business intelligence continue to be emerging disciplines. Viewing business intelligence as a process rather than an end by itself requires managers to remain flexible and open-minded.

Second, the objective of knowledge management is not always to make knowledge more visible or available. Like other assets, it is sometimes in the best interests of the firm to keep knowledge tacit, hidden, and non-transferable. Competitive advantage increasingly depends on knowledge assets that are difficult to reproduce. Retaining knowledge is as much a strategic issue as sharing knowledge. Business intelligence, on the other hand, is designed to make knowledge visible, at least inside the enterprise, so it can be analyzed and acted upon to meet business objectives.

Third, knowledge can create a shared context for thinking about the future. If the purpose of knowledge management and business intelligence is to help make better decisions, then it must provide value for future events, not just views of the past history. The goal is to use data to identify trends and environmental changes, then create predictions that help inform business strategy and long-term goal setting.

Finally, people lie at the heart of knowledge management and business intelligence. Establishing and nurturing a culture that values learning and sharing of knowledge enables effective and efficient knowledge management. Knowledge sharing—subject, of course, to the second caveat already described—must be valued and practiced by all employees for knowledge management to work. The success of knowledge management ultimately depends on a personal and organizational willingness to learn.

SUMMARY

• Knowledge management includes the processes necessary to generate, capture, codify and transfer knowledge across organizations. Business intelligence is the set of technologies and practices used to analyze and understand data and to use it in making decisions about future actions. Business analytics is the set of quantitative and predictive models used to drive decisions.
• The four main types of intellectual property are patents, trademarks, designs, and copyrights.
• Data, information, and knowledge should not be viewed as interchangeable. Knowledge is more valuable than information, which is more valuable than data because of the human contributions involved.
• The two kinds of knowledge are tacit and explicit. Tacit knowledge is personal, context-specific, and hard to formalize and communicate. Explicit knowledge is easily collected, organized, and transferred through digital means.
• Knowledge management is a dynamic and continuously evolving process that involves knowledge generation, capture, codification, and transfer. Traditional business intelligence includes reporting, querying, dashboards and scorecards.
• Business intelligence 2.0 integrates Web 2.0 capabilities and features into traditional BI systems and creating instant access to real-time information and data.
• Successfully competing with business analytics means that an organization has these five capabilities: hard to duplicate, uniqueness, adaptability, better than competition, and renewability.
• There are five levels of analytics maturity: reporting, analyzing, predicting, operationalizing, and activating.
• Big data refers to very large data repositories often found in environments where information is created quickly. Tools for managing big data sets are different than those for other data sets.
• Social analytics provide companies with tools to monitor and engage their communities and to evaluate the success of their investment in social IT.

KEY TERMS

big data (p. 341)

business analytics (p. 327)

business intelligence (p. 327)

data (p. 330)

data mining (p. 339)

data warehouses (p. 338)

explicit knowledge (p. 334)

externalization (p. 334)

evidence-based management (p. 340)

folksonomy (p. 335)

information (p. 330)

intellectual capital (p. 328)

intellectual property (p. 328)

knowledge (p. 331)

knowledge capture (p. 335)

knowledge codification (p. 335)

knowledge generation (p. 335)

knowledge management (p. 327)

knowledge transfer (p. 335)

social analytics (p. 342)

socialization (p. 335)

tacit knowledge (p. 332)

tagging (p. 335)

DISCUSSION QUESTIONS

1. The terms data, information, and knowledge are often used interchangeably. But as this chapter discussed, they can be seen as three points on a continuum. What, in your opinion, comes after knowledge on this continuum?
2. What is the difference between tacit and explicit knowledge? From your own experience, describe an example of each. How might an organization manage tacit knowledge?
3. What does it take to be a successful competitor using business analytics? What is IT's role in helping build this competence for the enterprise?
4. How do social analytics aid an organization?
5. Why is it so difficult to protect intellectual property? Do you think that the Digital Millennium Copyright Act is the type of legislation that should be enacted to protect intellectual property? Why or why not?
6. PricewaterhouseCoopers has an elegant, powerful intranet knowledge management system called Knowledge Curve. Knowledge Curve makes available to its consultants and auditors a compendium of best practices, consulting methodologies, new tax and audit insights, links to external Web sites and news services, online training courses, directories of in-house experts, and other forms of explicit knowledge. Yet, according to one of the firm's managing partners, “There's a feeling it's underutilized. Everybody goes there sometimes, but when they're looking for expertise, most people go down the hall.”²⁵ Why do you think that Knowledge Curve is underutilized?

CASE STUDY 11-1
STOP & SHOP'S SCAN IT! APP

The grocery store and supermarket shopping industries have combined annual revenues in the hundreds of billions of dollars. Industry guru Phil Lembert, estimated that by 2015, $706 billion dollars will be spent on groceries annually. Grocery shopping was a highly commoditized industry with over 85,000 stores in the United States. With little variation in available item selection and less money being spent on groceries in the down economy, competition for customer loyalty was at an all time high in 2012. By using business analytics to help process buying habits of its customers, Stop & Shop, a Quincy, Massachusetts-based grocer, tried to get a better grasp on the hard-to-understand concept of customer loyalty in grocery shopping.

In 2009, Stop & Shop introduced Scan It!, a portable electronic device for customers shopping in their stores. The device allowed customers to “scan and bag” products, expediting check out times at the end of their shopping trip. Additionally, the device offered deals based on the location of the scanner (and therefore the customer) in the store. Location-specific discounts in real time became increasingly popular to customers, as usage of Scan It! grew by 10% in both the first and second quarters of 2009. The most beneficial aspect of the Scan It!, however, came with the powerful analytics software built into the device by Modiv Media, in which Stop & Shop owns a minority interest. The software kept track of each customer's purchasing habits both past and present, to individualize coupons in real time for the customer.

The scanner resulted in three positive trends for Shop & Stop. Customer loyalty grew, which allowed Stop & Shop to secure a greater customer base than area demographics would predict. Additionally, each shopper's basket size increased as individually tailored coupons enticed customers to buy more. Lastly, Stop & Shop saw its customer base grow, as word of mouth marketing brought in more customers to try the state of the art device.

Stop & Shop saw customer adoption plateau, however, as a couple of years passed, and the age of mobile apps increased ease of use. In October 2011, the grocer created the Scan It! app for the iPhone and Android. By eliminating the need to sign in and retrieve a scanner at the store, customer adoption continued its upward climb. Additionally, as customers became increasingly concerned about saving money while shopping, Stop & Shop built in budgeting software to allow customers to track their spending more effectively. Ads for the new app proclaimed, “New Mobile App Allows Customers to Shop, Bag, and Tally Their Grocery Order with Their Personal iPhone® and Android™ Devices”. Scan It! was heralded as “a first of its kind grocery app that allows customers to use their personal mobile device to scan, tally, and bag their groceries while they shop.”²⁶

Stop & Shop bundled an app that not only rewarded customers who shopped at their stores by helping them save money, but it provided additional functionality to the customers and tracked information on sales, which Stop & Shop loaded into its data warehouse and used to understand its customers. Analytics then helped Stop & Shop put the right items on its shelves to maximize sales and create customer loyalty.

Discussion Questions

1. What is the benefit of the Scan It! data to Stop & Shop? What are some of the questions can they now answer about their customers?
2. How would you assess the level of maturity of Stop & Shop's use of analytics? What might they do differently with the data to gain more value?
3. What concerns might shoppers have about their privacy? How would you advise Stop & Shop management to respond to these concerns?

CASE STUDY 11-2
BUSINESS INTELLIGENCE AT CKE RESTAURANTS

At a time when most fast-food restaurants were touting nutrition, Hardee's proudly introduced the Monster Thickburger. This burger boasts a phenomenal 1,420 calories and 107 grams of fat. It consists of two, one-third-pound charbroiled 100% Angus beef patties, three slices of American cheese, a dollop of mayonnaise, and four crispy strips of bacon on a toasted buttery sesame seed bun. What on earth was CKE Restaurants, the owners of the Hardee's chain, thinking?

Because of its Business Intelligence System (BIS), CKE was confident about introducing the Monster Thickburger across the United States. A BIS uses data mining, analytical processing, querying, and reporting to process a business's data and derive insights from it. CKE's BIS, known ironically inside the company as CPR (CKE Performance Reporting) monitored the performance of its Monster Thickburger in test markets to ensure that the burger contributed to increases in sales and profits at restaurants without cannibalizing sales of other more modest burgers. To do so, CKE's BIS studied a variety of factors—such as menu mixes, Monster Thickburger production costs, average unit volumes for the Monster Thickburger compared with other burgers, gross profits and total sales for each of the test stores, and the contribution that each menu item (including the Monster Thickburger) made to total sales. Because the sales of Monster Thickburger exceeded expectations in the test markets, CKE developed a $7 million dollar advertising campaign to launch its nationwide introduction. Monster Thickburger sales exceeded expectations, and Hardee's sales revenues increased immediately, eventually growing by 8%. “The Monster Thickburger was directly responsible for a good deal of that increase,” says Brad Haley, Hardee's Executive Vice President of Marketing.

CKE, partially because of its reliance on CPR, was rescued from the brink of bankruptcy. It increased sales at restaurants open more than a year, narrowed its overall losses, and finally turned a profit after three years. CPR, its proprietary system, consists of a Microsoft SQL server database and uses Microsoft development tools to parse and display analytical information. It uses econometric models to provide context and to explain performance. The company reviews and refines these models each month. The econometric models take into consideration 44 factors, including the weather, holidays, coupon activity, discounting, free giveaways, and new products. With the click of a button, for example, a sales downturn can be explained on a screen that shows that 5% of the 8% decrease was due to torrential rain in the Northeast and 2% was due to free giveaways.

In the competitive restaurant chain industry, companies have to be agile and responsive to the dynamic environment that they face. They must match their BIS initiatives to their business strategies in order to improve operations and their bottom lines. BISs assist them in making strategic decisions about menu items and closures of underperforming stores, as well as tactical matters such as renegotiating contracts with food suppliers, monitoring food costs, and identifying opportunities to improve inefficient processes. To derive value from their BISs, many restaurant chains have successfully reduced the three biggest barriers to BIS success: voluminous amounts of irrelevant data, poor data quality, and user resistance.

CKE's CIO and Executive Vice President of Strategic Planning, Jeff Chasney, states: “If you're just presenting information that's neat and nice but doesn't evoke a decision or impart important knowledge, then it's noise. You have to focus on what are the really important things going on in your business.”

Chasney stresses a BIS should be different from the plain-vanilla standard corporate reporting tools of old. Rather, a BIS should provide managers with insights rather than just data. He believes that the context from which the data was collected significantly impacts how that data should be interpreted. Systems that just report changes without enough background or information on what caused those changes are not very useful. Managers don't know what data to trust. Chasney explained: “If your business intelligence system is not going to improve your decision making and find problem areas to correct and new directions to take, nobody's going to bother to look at it.”

The first step to developing a BIS is to understand the company's decision-making processes. Before information is collected, analyzed and used in the BIS, someone has to identify what information is needed to confidently make decisions. For instance, the CEOs of CKE's three restaurant chains wanted to understand what made sales fluctuate, while the COOs wanted to know how to recognize good business opportunities as well as underperforming properties. Then the BIS designer must determine the appropriate presentation format, be it a report, a chart, or a Web site.

BIS must add value to the executive's decision-making processes. To do that, attention must be paid to the critical performance indicators. For CKE, as Chasney learned, those are sales, cost of sales, exceptions (such as high-performing or underperforming areas), and business trends.

Discussion Questions

1. How does the Business Intelligence System (BIS) at CKE add value to the business?
2. What are some tips for developing and using the BIS described in this case?
3. Was the introduction of the Monster Thickburger a good idea or an example of information leading to a wrong decision?

Sources: Christine Lagorio, “Man vs. Monster Thickburger,” CBS News (February 11, 2009), http://www.cbsnews.com; and Meredith Levinson, “The Brain Behind the Big, Bad Burger and Other Tales of Business Intelligence,” CIO Magazine (May 15, 2007), http://www.cio.com/article/109454/The_Brain_Behind_the_Big_Bad_Burger_and_Other_Tales_of_Business_Intelligence.