Decision Making and Digital Transformation
Digital transformation is changing our lives, our jobs, our organizations, and our world. Each of us makes choices that impact how we use digital data and digital technology. Managers and organizations that do not keep up with digital transformation trends and successfully implement key transformation projects will likely suffer negative consequences, including loss of jobs, going out of business, or being acquired by a digital upstart or a more traditional competitor.1
Innovating with data, digital technologies, and data-based decision making is a major business opportunity that can change business models, improve customer experiences, reduce costs and increase agility. Such innovation is necessary to prosper in our changing global economic markets. Sadly, there is no simple formula or training program that can help managers become data-based decision makers. Indeed, most managers are trying to learn “on the job” to understand what digital transformation means for them, their team, and their organization while trying to tackle the great challenge presented by rapidly expanding organizational data. Successful digital organizational transformation requires that managers have mastered data-based decision-making skills that can help formulate, implement and manage an appropriate digital transformation strategy.
Effective data-based decision making using accurate and timely data is integral to successful digital transformation. Opportunities and obstacles created by digitalization2 and leveraging digital technologies require that managers develop sophisticated data analysis, data interpretation, and decision-making skills. The velocity at which large datasets are processed and reported requires that managers adapt to an increasingly fast paced business environment. A digital world requires decision making and leadership skills that leverage diverse data sources, data analytics know-what and know-how, and decision support capabilities that enable and support the strategic direction of an organization.
Data-based decision making is not a new idea. Decision support research began at the dawn of the digital age and the concepts of decision support and decision support systems (DSS) remain understandable and intuitively descriptive. Related terms such as high velocity decision making, data analytics, business intelligence, and big data analytics are of more recent origin and are interpreted in different ways by managers, software vendors, and consultants. Also, artificial intelligence (AI), data-based applications, and real-time analytics are accelerating the velocity of digital organizational transformation.3
Data-based decision making is a broad concept that prescribes an ongoing process of collecting and analyzing different types of data to aid in making fact-based, routine and nonroutine decisions. The use of new digital information the use of new digital information technologies to change and improve business processes, alter business models, enhance products, and change customer experience.
Many new technology developments, like the Internet of Things (IoT) are expanding the range of computing devices and expanding data collection. AI and data analytics are helping managers use the new data sources in real-time. Managers choose how to exploit and adopt these technology developments. Actions of managers disrupt existing business models and create new opportunities for businesses across industry sectors. These intertwined changes are causing significant digital disruptions.
The following five sections in this chapter discuss related topics, including: (1) data, information, and knowledge, paying particular attention to the opportunities to use new data sources as part of a digital transformation strategy, (2) understanding data-based decisions and decision support, (3) digital transformation impacts, especially upon people, processes, and strategy, (4) asking the right questions, and (5) creating data-centric organizations. Our perspective emphasizes the centrality of decision support and data-based decision making in organizations.
Understanding Data, Information, and Knowledge
To adapt and cope, organization decision makers need better, faster, more accurate data and information to make decisions. The volume, availability, and speed of real-time data continue to be a special challenge in organizations. Managers are increasingly focused on finding opportunities arising from valuable data insights. An organization’s capability to capture, store and manage, analyze and visualize large volumes of semistructured, and unstructured data is generally referred to as using “Big Data” (Chen et al. 2012). Big data refers to very large data volumes that are complex and varied, and often collected and must be analyzed in real-time. Venture Capitalist Bryce Roberts4 reminds us “Data, big, medium or small, has no value in and of itself. The value of data is unlocked through context and presentation.” How data are presented or visualized can change behavior. Managers continue to struggle with issues like managing large volumes of data and information, anticipating external environmental uncertainty, and monitoring advances in technology.
Focusing on greater use of knowledge as an organizational strategy found widespread recognition and approval after Drucker (1992) concluded that “the basic economic resource—the means of production—is no longer capital, nor natural resources, nor labor. It is and will be knowledge.” While knowledge remains a core organizational and societal resource, the notion of big data and developing and understanding an organization’s ability to extract relevant knowledge and associated insights using sophisticated technology has of necessity become a priority for managers.
A vast amount of industry, company, product, and customer data can be gathered from a wide variety of external and Internet sources including online social media forums, web blogs, social networking sites, logs of website visits, and retail transactions. Most of this data is significant in volume and it is often unstructured in nature—it is considered big data. It is difficult to prescribe a “one size fits all” approach to big data because big data for one organization may be “small data” for another. However, the key to successful big data use is a manager’s ability to identify the value in the data collected and devise ways to explore and extract value from large volumes of data for the right people at the right time.
There is overlap and yet differences among the concepts of digital data, information, knowledge, and big data (see Figure 2.1). It is difficult to identify where one concept begins and another ends, this challenge is reflected in the distinct lack of a common language used by stakeholders when it comes to defining and discussing these phenomena. Rather than over emphasizing the boundaries of data, knowledge, and information and when data becomes big data, managers must focus on decision support and analytics needs required to help them in achieving business objectives. Managers need relevant data, information, knowledge and decision support capabilities to meet decision making needs.
Figure 2.1 Data concept map: Big data, data, information, and knowledge
Understanding Decisions and Decision Support
Managers make many decisions and the characteristics of each decision determine if analytics and decision support are appropriate and if so what support is most useful. Decisions are made as part of processes and decisions result in outcomes. A decision may involve assessing and evaluating alternatives using data sets, variables, and algorithms. The quality of a decision is often impacted by the type of process or path that is pursued in making and implementing a decision. Organizational decision environments are typically characterized by a rational decision making approach. Rationality is the quality of being consistent with or based on logic and reason. One hopes managers and responsible decision makers attempt to be rational and thoughtful in their decision making.
Decisions vary widely in structure and complexity. Some decisions are characterized by a concise decision question, with a clear, well defined and structured choice. These are typically known as operating or function-specific decisions. This type of decision is usually routine, occurring regularly and frequently, that is, daily or weekly. Tactical decisions are typically addressing a broader decision question, and are semistructured in nature, this means that some but not all of the information necessary to make the decision is available. These decisions are mostly internally focused and may even be specific to an individual business unit. Other nonroutine decisions are more complex. In these situations, some variables may not be well understood, often information required to make the decision is unavailable, incomplete and in some situations information may be known to be inaccurate. Classified as strategic decisions, these are usually complex, unstructured decisions involving many different and connected parts. These decisions usually involve a high degree of uncertainty about outcomes. If implemented, strategic decisions often result in major changes in an organization. Pursuing digital transformation is a strategic decision.
Modern decision support is evolving rapidly in step with computing hardware and software progress. A modern decision support system is up-to-date technologically. The modern era in decision support development started in many ways in 1995 with the specification of HTML 2.0 and the introduction of handheld computing and cell phones. Since 2007, Web 2.0 technologies, mobile integrated devices, and improved software development tools have revolutionized decision support user interfaces, while the decision support data store back-end has gotten extremely powerful supporting large, real-time and complex data sets. Modern decision support is varied and increasingly widespread in use.
A DSS is a computer-based information system that supports individual or team decision making. There are five major categories or types of DSS, but some DSS do not fit neatly into one of the categories, instead they have multiple decision support functionalities and a hybrid architecture. Hybrid or complex DSS have components to provide more than one category of decision support. The five primary DSS categories include: (1) Communication-driven DSS that enable cooperation, supporting more than one person working on a shared task, (2) Data-driven DSS that emphasize access to and manipulation of internal company data integrated with external data, (3) Document-driven DSS that manage, retrieve, and manipulate semistructured or well-structured documents, (4) Knowledge-driven DSS that provide specialized problem-solving expertise stored as rules, procedures, or in similar structures, and (5) Model-driven DSS that emphasize access to and manipulation of statistical, financial, optimization, simulation, or other quantitative models.
The general, defining characteristics of DSS have not changed over the years. These systems remain characterized by facilitation, interaction, an ancillary role, repeated use, task oriented, identifiable and having a decision impact (cf. Power 2002). Some characteristics are more closely associated with one category of DSS than another, but complex DSS often have multiple subsystems that fit in different categories. For example, a complex, modern decision support capability may have a well-defined data-driven subsystem and a model-driven subsystem. Major specific characteristics of modern DSS include: (1) Broad domain of applications with diverse functionality, (2) Faster access to data stored in very large data sets, (3) Faster deployment, (4) Faster response, (5) Integrated DSS with transaction processing systems (TPS), multiple decision support subsystems, (6) Lower cost per user, (7) Multiuser and collaborative interaction, (8) Real-time data and real-time DSS use, (9) Ubiquitous, (10) User friendly and a better user experience, and (11) Visualization. Table 2.1 describes these 11 characteristics of modern decision support capabilities. These modern decision support applications enhance both data-based and data-informed decision making.
Table 2.1 Characteristics of modern decision support applications
|
Characteristic |
Description |
|
1. Broad domain of applications with diverse functionality |
Decision support user base and the rationale for DSS use has expanded. There are many use cases for decision support and we are capturing use case models. |
|
2. Faster access to data stored in very large data sets |
Data access refers to software and activities related to retrieving or acting upon data in a database or other repository. Data-driven DSS can use very large data stores. |
|
3. Faster deployment |
Software deployment is all of the activities that make a new DSS available for use. Faster deployment is partly due to the use of Web technologies, also better prototyping and templates. |
|
4. Faster response |
How quickly an interactive system responds to user input has improved significantly. In a distributed computing environment, the lag for video, voice, data retrieval or transmitting results is now negligible. |
|
5. Integrated DSS with TPS |
Enterprisewide decision support applications are increasingly common. A standardized interface and single sign-on security helps create an integrated and unified decision support/transaction processing environment (TPS). |
|
6. Lower cost per user |
Total annual cost for licensing development software on a per user basis is declining. This trend will continue given the increased open source decision support applications. |
|
7. Multiuser and collaborative interaction |
DSS are increasingly collaborative with shared decision making environments. |
|
8. Real-time data and real-time DSS use |
The classical decision support idea is an immediate real-time system that is used while action is occurring. That vision is increasingly possible and useful. |
|
9. Ubiquitous |
DSS are available and seem to be usable everywhere. DSS for a particular function can be used on mobile devices. |
|
10. User friendly and a better user experience |
Usability is the ease of using a particular tool. All DSS are much easier to use, but we can do more to improve usability and reduce information load. |
|
11. Visualization |
Creating images, diagrams, or animations to communicate a message is important. Modern DSS include capabilities to create and manipulate visualizations. |
This list of attributes and characteristics of a modern DSS is likely incomplete. Decision support is usually ahead of current practices, but the list may represent current “best practices.” Those of us interested in modern computerized decision support are promoting new ideas and approaches, and encouraging progress in supporting decision making. Building better decision support provides one of the “keys” to competing in this increasingly digital global business environment. Better decision support is a major enabler of digital transformation.
Digital Transformation Impacts
Digital transformation is a complex concept and challenging goal that holds a variety of meanings for a diverse set of stakeholders. In a recent Forbes article, Kerschberg (2017) contends that technology is central to organizational digital transformation, in particular adopting analytics, big data, mobile, cloud, IoT, and application development. While technology may be at the core of digital transformation, successful digital transformation requires excellent leadership, a supportive culture, and new business processes. Leadership should promote and cultivate a data-based decision-making culture. Digital transformation remains a complex task. It begins with strategic leadership and a commitment to a digital transformation organization strategy.
Barriers to entry for many industries have been lowered and some industries have been consolidated or forced to contract. Nontraditional competitors are entering industries and changing markets and goods and services. Few managers in traditional industries led prior transformations and some have been caught off-guard. Managers should have known digital technologies would be an enabler of change, so why were so many surprised by the suddenness and magnitude of the disruption? Lack of vision and understanding? Complacency? An inward looking attitude? Lack of knowledge? Perhaps a combination of these reasons.
A related discussion on digital transformation, cf. Power (2017), stated that a strategic vision for digital transformation is useful; however vision must be grounded in customer needs and technology possibilities. Indeed, business transformation cannot happen without people making decisions about technology. Figure 2.2 is a conceptual decision support guide for managers.
Figure 2.2 Conceptual digital transformation support guide
The conceptual model in Figure 2.2 highlights three levels of organizational tasks for implementing digital transformation including: (1) strategic tasks, (2) tactical tasks, and (3) operational tasks. Managers at each level choose from a set of tasks that should be completed as part of an organization’s approach to digital transformation. Through the completion of some or all of these tasks, organizations can move from an ad-hoc approach to a more systematic, mature approach to digital transformation. Achieving a digital transformation vision that is stable and “mature” comes from successfully completing transformation tasks.
Digital transformation tasks may be broadly characterized in terms of changing people, processes, and technology. For any business strategy to be successful activities across these dimensions need to be aligned. In a Harvard Business Review article, Trevor and Varcoe (2017) promote the notion of strategies, capabilities, and resources to achieve digital transformation including systems that “should be arranged to support the enterprise’s purpose.” Identifying appropriate transformation tasks to undertake is important.
Strategic tasks represent a high level collection of activities that implement a digital transformation vision and strategy. Some of these tasks include developing the vision, developing digital leadership capability (Westerman et al. 2014), reinventing business models, rethinking business processes, redefining stakeholder engagement (Kerschberg 2017), and developing a digital governance strategy (Ernst and Young 2017). While this list is not exhaustive, it is moving toward a more balanced and holistic approach for managers to tackle digital transformation of an organization.
As illustrated by the “middle out” notation and arrows used in Figure 2.2, tactical tasks (the middle) are integral to the success of a digital transformation strategy. Trevor and Varcoe (2017) refer to this mediating managerial level in terms of building organizational capabilities. At this level, it is important for managers to consider and select tasks that provide them with the means (capability) to deliver the digital transformation strategy. These tactical tasks will drive the digital transformation agenda in an organization with managers asking questions about how transformation can be achieved based on the capabilities available. Tactical tasks are concerned with designing new business processes, establishing disruptive new business models, and defining data governance processes. It is also useful to consider new mechanisms for evaluating performance in terms of achieving digital transformation. Completing tactical tasks may be an opportunity for managers to define new measures of organizational success including understanding customer engagement and customer experience. Data-based decision making should guide and help prioritize tasks.
Operational tasks are focused around the questions managers need to answer for (1) selecting and developing technologies, (2) establishing viable data integration platforms, (3) choosing necessary security controls that will balance data access with data protection, and (4) identifying and developing the right people capabilities to achieve the digital vision for the organization. Managers need to select and complete tasks that develop “assets that will be useful in a digitally transformed world” (cf. Capgemini Consulting 2011).
Completing the appropriate mix of tasks should increase the chances of a successful digital transformation. Technological maturity is at the heart of achieving real digital transformation. To begin a digital transformation journey, managers need to move beyond focusing on new individual technologies to develop a comprehensive digital technology capability that is closely aligned to a well-defined digital transformation vision and strategy.
Digital transformation can both solve problems and create new problems. Successful digital transformation creates a positive, long-term, net benefit for an organization. Applying digital technologies can create a data-based virtuous feedback cycle that leads to adopting and choosing more innovative and transformative digital solutions.
Finding Success: Asking the Right Questions
Making decisions often involves answering questions like “What should be done?” or “What alternative is best?” Asking the right questions can help people make better decisions. Kipling (1902) wrote a poem managers should remember when thinking about defining decision questions. The poem begins “I keep six honest serving-men (They taught me all I knew); Their names are What and Why and When and How and Where and Who.” These six are the primary question words. There are other question words and phrases like How many? and How much?, even the word “Is” can start a question, but Kipling’s six words (see Figure 2.3) are a good starting place for examining how decision questions differ in intent and how they are similar.
Figure 2.3 Six honest serving men (adapted from Kipling)
A decision question asks about what action(s) to take among various options or alternatives. A decision process helps create alternatives/options, find them, and sometimes eliminate options. Answering or resolving a decision question should be a thoughtful and comprehensive process. Assessing and understanding the decision question provides guidance and direction for customizing an appropriate decision process. So let’s examine Kipling’s “serving men” and evaluate what they mean for decision making processes.
What ... ? is a complex word to begin a question. A “What” question may seek an approximation, a forecast or an estimate, that is, “What will sales be in the next quarter?” Alternatively a “What” question may seek specific facts or information, that is, “What is the current profitability of XYZ?” In this second case, the “What” question does not or should not require a decision, rather a fact is sought.
Why ... ? questions are the most perplexing and most troublesome. By asking why, a person is expecting a reason for an action or event or an explanation of something that has occurred. Asking and answering why questions are important in problem solving. Understanding causes and motivations helps us understand a decision situation. In general, a “why” question supports reasoning or informs a decision question. For example, examine this question, “Why did sales decline in the last quarter?” This diagnostic question seeks to know the cause of a problem so that perhaps a decision can be made about how to remove the cause and reduce the negative consequences. A person may ask multiple “Why” questions in a decision situation prior to or after asking a key decision question, that is, “How can the decline in sales be stopped or reversed?”
When ... ? refers to time. One wants information about the time or timing of events or actions. The Cambridge English Dictionary notes “We can use when to ask for information about what time something happens.” So a decision maker may want to know about past, present or future time, that is, “When did sales start to decline?” “When should the new sales and marketing plan be implemented?” Some When questions imply a decision is needed while others are informational. Helper words like will and should can indicate whether information or a decision is sought.
How ... ? much or how many or how can. How is sometimes the indicator of a decision question. For example, similar to the prior discussion we might ask “How can we stop the sales decline?” or “How many people will attend the event?” or “How do we contact customers?” Some how questions are requesting an estimate or conclusion. For example, “How much will a new production facility cost?” Many questions that begin with the how keyword request information rather than a decision, that is, “How do I find XXX?”
Where ... ? refers to location and place. The where question word seems to be primarily an indicator of a need for information about location, that is, “Where will the event be held?” or “Where is the salesperson?” The first location question may involve a decision if phrased with the should helper word, that is, “Where should the event be held?”
Finally, there comes Who ... ? the who keyword may indicate a request for information or a need to make a decision. For example, “Who is the sales manager?” is an information request. While the question “Who should be the sales manager?” implies the need for a decision among a set of people who have applied for a job. A Who question may refer to choosing one person or a group or team of people. Consider the question “Who will implement the new strategy?” The answer might be a single individual or a group or team.
Some decision questions are highly structured, routine and repetitive, others are semistructured or unstructured and even novel and nonroutine. The amount of structure depends to some extent upon the decision situation. For example, for a salesperson the following are usually routine and repetitive questions: “Who will buy the product or service from you and your company?” “What and how much will they buy from you?” “Why do they need the product or service?” and “How should you engage them in a meaningful conversation about the product/service?” In the context of assessing the introduction of a new product, the same questions are more unstructured and often nonroutine.
Although decision questions often begin with one of the six questions words, that is, what, why, when, how, where, and who, these question words get altered by helper words like will and should, much and many. In English, one finds decision questions are alternatively framed as “Should we” or “Should I do XX or go to XX.” Colloquial or informal language also interferes with recognizing decision questions. Knowing what the word phrase implies helps us provide decision support when that is appropriate. Asking the “right” decision questions in a solution oriented manner is an important skill for data-based decision makers. Spend some time early in a decision process to specify an appropriate decision question or questions that can be answered.
Creating a Data-Centric Organization
A supportive organizational culture is important to successful digital transformation. A data-centric organization has policies and a culture that encourage and reward the use of data in products, processes, and decision making. Using data to make decisions in organizations has long been a goal of most managers. Using “gut instincts” and limited facts has serious risks. Basing a decision on self-interest is poor practice, and even if that is the choice criterion the decision maker still requires facts. Consulting mystics/fortune-tellers fell into disfavor long ago. Today most decision makers have increased access to more and better data in near real time almost anywhere in the world. This new reality has changed the decision support possibilities. Managers can make better fact-based decisions if they choose to develop the decision support capabilities and infrastructure.
Becoming a data-informed, or data-centric organization has become a priority for many managers. Analytics and decision support must be aligned with business strategy to realize benefits from digital transformation. Pushing for more data and more analytics without a strategic fit is folly. Organizations and managers need to understand what they are trying to achieve. Decision support initiatives fail when there is poor alignment with the business strategy.
Organizations can empower employees with access to relevant data and analytics. The key is to provide relevant data when it is needed to make a decision. The decision maker remains central to decision taking, but technology and analytics support are enhanced for data-based decision making. Providing data does not mean however it will be used properly or even used. Training and reward systems are key to making the new decision support capabilities a factor in improving organizational performance.
The term digital data refers to facts, figures, and digital content captured in information systems. Raw data are the bits and bytes stored electronically. Data may be streaming to a decision maker or retrieved from a static data store. Figuring out what data is relevant and what that data means in a decision situation can be challenging. Data can overwhelm a decision maker and can mislead. Data-based decision making requires anticipating data and analysis needs and providing the opportunity to request and analyze additional data. Analytics involves processes for identifying and communicating patterns, derived conclusions and facts. Decision support and analytics must provide timely and useful information for benefits of digital transformation to be realized.
Using data in decision making must become part of an organization’s culture. The quest for understanding, formalizing, and prioritizing important decision questions, and then capturing and making available appropriate data, and relevant analysis must become an urgent requirement and ongoing priority. A data-centric organization survives and hopefully prospers based on the quality, provision, and availability of data to decision-makers. Data should be captured where it is generated and then it must be appropriately stored and managed for use in decision-making. Analytics, decision support, and data become the basis for decision making in a digitally transformed organization.
Summary
Creating a data-centric organization where managers make data-based decisions has both technology and human resource challenges. Technology challenges continue to evolve as more data and better, easier to use analytic tools become available. The human resource challenge involves retraining and motivating current employees to use analytics, model-driven, and data-driven decision support.
Digital transformation does not occur quickly, rather it is a journey. We know that factors other than data availability influence choices. Without data and facts, then luck and chance dominate outcomes in situations. Chapter 3 examines how using data and information underpin data-based decision making which is also the key to more effective decision making.
1 “Couchbase Research Reveals a Majority of Organizations Expect to Fail in Four Years if Digital Transformation Approach is Unsuccessful.” http://dssresources.com/news/4798.php
2https://i-scoop.eu/digitization-digitalization-digital-transformation-disruption/
3Dimension Data, “Artificial Intelligence and Analytics Accelerate the Pace of Digital Workplace Transformation.” http://dssresources.com/news/4789.php
4Roberts, B. February 2012. “Data Data Everywhere and Not a Drop of Value,” http://bryce.vc blog, at URL http://bryce.vc/post/15300645787/data-data-everywhere-and-not-a-drop-of-value