Theory of Digitalization and Decision Making

"Digitalization refers to the practice of taking processes, content or objects that are used to be primarily (or entirely) physical or analog and transforming them to be primarily (or entirely) digital. The effect of digitizing processes, aside from potential efficiency gains, is to make processes more tailorable and malleable". Not only based on data, moreover targeted on markets, organizations and processes digitalization will also deploy its full value to businesses and industries. The new era of digitalization has already started and shows the first step of a new business world with a change in the division of work. In the early 20th century a "computer" was an employee calculating tables the whole day. In a first step, an automated computer (as we understand today) took over this task and increased quality and efficiency of this process. Since then, the automation of our world was ongoing and influenced by computers and machines. Crucial for the development was the definition of rules for computers because computers are perfect in following algorithmic rules. Further actions took place to develop more feasibilities of machines, i.e. Apple iPhone is now with the software Siri able to understand and direct the user. It is a real quantum leap because language automatization and transformation of it into instructions was a giant step for the industry. Currently, the speed of development is increasing, either the trend of "Industry 4.0" with full automation of the production flow or "artificial intelligence" that robots tend to make autonomous decisions and developed self-awareness and self-maintenance.

Referring to the effects of digitalization, this trend will completely change the way of making business and making decisions. Flexibility and transformability are key attitudes of successful organizations in the future and drive them on the road of digitalization. Digitalization will have an effect on customer structure and behavior, increase the efficiency of operations including their supply chain and at the end may change the entire business model. It is important to understand the logic of digitalization and to realize the four levels of transformation. These four levels have to be in scope of the decision maker.


Digital data (big data)

Due to recording, processing and analysis of mass data, high-quality and more predictable forecasts and decisions are possible in organizations. The structure of big data is:

• Volume: measured in Giga- or Terabyte
• Velocity: One-time snapshot frequency streams
• Variety: structured, numeric, alpha, unstructured, text, voice / sound, image / video, genomics
• Veracity: Validation, noise level, deception, detection, relevance, ranking

Big data is a huge trend in digitalization because the usage of data is important for an organization. As "the economist" wrote in 2010, "Data are becoming the new raw material of business". And data are increasing day by day. A strong increase of data traffic had happened and will further increase; in the future driven by digitalization, e.g. machine to machine communication or the trend of mobile data from every user of the internet. This traffic is permanently increasing, and all devices are moving toward mobile and smart functions. The latest trend of data shows the movement from big to smart data which means data including utility, semantics, data quality and security.


Automation
The traditional trend of automation is an ongoing process pushed by new technology and the need for efficiency to enable a competitive cost base. The combination of traditional work and technology with artificial intelligence will enable autonomous work in self-organization systems with high quality and high efficiency. As an effect, production speed will increase and unit costs will drop. Automation has different aspects of realization. First, the work volumes between man and machine are changing. Second, the trend toward a higher automation is still ongoing, up to entire fully automated factories without human beings. Today a work flow from machine to machine without human interaction is possible and works without variances on a repeatable high-quality level. The third step, artificial intelligence, is a self-learning system, with a set of different reactions based on environmental conditions. As an example, one new technological process is rapid manufacturing, which means that traditional production types will be replaced by new technology. This new technology uses direct digital data for production, without a tooling procedure. Cost intensive tools are replaced by new manufacturing applications. These procedures are very flexible at an acceptable cost level and permit small batches. 3D printers are today on a level beyond testing and started their usage in business. Different possibilities of 3D printing enable a wide range of applications.


Integration
Connecting the entire value chain by the high-quality broad band will enable synchronized supply chains with shorter production cycles and faster innovation cycles (mobile or fiber opic net).An important integration in the era of digitalization is a deeper collaboration in the entire supply chain. Based on a more flexible consumption behavior a more agile supply chain has to deliver this flexibility. As a result, a strong cross-company collaboration is a must, and online information has to be exchanged between the different entities. Cross-linking of organizations and their IT systems are the requirement for an efficient supply chain. Today's technology enables this process, based on standard software tools and open interfaces for an optimal data exchange between them. A fast changing world, with a drop in product life cycles, the integration between supplier and customer has to be on the level of a partnership because a cross-company product development process needs to have speed in "time to market" on a cooperative relationship.


Digital customer access
New competitors, new services, and new transparency will increase competition and market position of companies and brands. For customers, the next opportunity is just one click away. Hence a well thought out strategy is necessary to position against competition, no matter from which industry it's coming from. For the car manufacturer BMW the competitor isn't anymore only Audi or Daimler, now it's Tesla and Google, with their new approach to mobility. The consumer and customer is again back in the center of scope of organizations. In the recent years, consumer behavior and hence customer activities changed dramatically toward a less loyal, more flexible and quick response relationship. Increased mobile services, e.g. via smart phone increased this trend, and Kreutzer/Land described the consumer buzzwords as follows: "me, all, immediately and everywhere"; which is exhibited in table 1 "Customer Expectation: Me, all, everywhere and immediately".

Table 1: "Customer Expectation: Me, all, everywhere and immediately"

Me All Everywhere Immediately
Appreciation as must Wide choice Time independence Instant contacting
(correct) Personalization High quality Location independence Fast transactions
Tailored offers Low prices Independence of technologies, channels, devices Short response time
Approach based on permissions Good service

 

The mentioned customer expectations are increasing, but in today's flexible world, this expectation is covered by competition. Hence this has to be the mantra in an organization. The trend of "smartization" is still ongoing and will further move on, making all devices via networks intelligent. Starting with mobile phones, developed to smart phones, now we see smart TVs and smart watches. Also, housekeeping is going smart, with refrigerators or washing machines, even energy consumption is steered with smart technology. As Porter describes, new smart products for consumers with three core elements: physical, smart and connectivity. All three elements deliver customer value improvements and are substantial for a future market position.

Decision theory is a wide area in science, with a long tradition. The first idea of decision theory reaches back over hundreds of years and is still relevant in the present. Rational models were discussed from the neoclassical economists (e.g. Adam Smith or Max Weber) with a view on rational behavior of agents which maximize their utility – the homo oeconomicus. A more scientific approach of Pascal and de Fermat shows a calculation of probabilities and Bernoulli laid the foundation of risk science by examining random events. Further developed by von Neumann/Morgenstern economic behavior in a strong rational and mathematical approach, decision making follows utility maximization. Today's view of decision theory as an interdisciplinary science (economics, psychology, sociology, philosophy, mathematics, computer science, and statistics) with different approaches is generally accepted. The most popular theory is still the theory of games and economic behavior. The theory of von Neumann/Morgenstern explains a rational behavior of market participants (either consumers or entrepreneurs). Consumers strive for a maximum utility or satisfaction and entrepreneurs strive for maximum profits. Meanwhile, a wide area of mathematical approaches and models of rational theories were further developed on the foundation of von Neumann/Morgenstern theories. The opposite of the rational view is a behavioral view on decisions. In the late 1940ies, Simon discussed the theory of bounded rationality, which means a certain influence of human attitudes with not pure rational decisions. A deeper view in the psychology science shows that theories on the behavioral economy are currently quite popular because human behavior is one part of organizational actions. In the 70ies Kahneman/Tversky developed the prospect theory. Also, Reinhard Selten contributed research to the field of behavioral decision making.


2.1 Rational View

In the classical field of the economical view on decision theory, a rational, mathematical founded approach is discussed. The process of pure calculating to find a maximum of utility or profit is a typical rational procedure. In the dimensions of decision fields, the alternatives are significant to analyze. There is only a decision problem if there are minimum two alternatives. Hence a determination of these alternatives must be reflected in the decision model. In the next step, an evaluation of these alternatives must be done. These consequences will lead to a result of the alternatives into the decision model. Important figures of the decision maker are defined as targets, these values are the result.

For a structured process, the environmental conditions are very important. Measures are not manipulable by the decision maker, these measures are called decision relevant data. These data are no variables for the decision maker. These characteristics are decision relevant environmental conditions. The illustration of figure 2 "Structures of Environmental Conditions" shows different environmental situations and a possibility of designing decision conditions.


Figure 1. Structures of Environmental Conditions

A decision structure under certainty means, that the decision maker has the real condition of the alternatives, hence all relevant information for the decision is given. Therefore the result is known and alternatives are certain. In reality, quite often decision models are formed as models of decision under certainty because the setup and the usage of this model type is easier to handle.

A decision structure under uncertainty means, that the decision maker has minimum two possible alternatives, but not all relevant information for a decision are given. Therefore the result is not known and alternatives are uncertain. In the case of uncertainty, there are two more possibilities. Either for the decision maker it's not possible to calculate a likelihood of conditions (uncertainty – narrow sense) or for the decision maker the probability of occurrence of a condition is computable (risk);.

This differentiation of the environmental conditions is important to define; especially decision making in the context of globalization and digitalization. While globalization creates more complexity and uncertainty, a more sophisticated model for decision making is substantial for an organization. In terms of digitalization, a need for algorithmic structures is a must, computers need a mathematical logic to calculate a result for the decision.


2.2 Emotional View

Significant developments in decision theory took place over the recent decades, though a trend to a behavioral approach was supported by psychological science. An outstanding contribution to the development of behavioral economics was made by Kahneman and Tversky. A collection of their scientific contribution and analysis is made in the book "Thinking, Fast and Slow". Describing decision theory with a strong psychological view makes a decision more emotional than rational. Depending on the activated system in the mind; Kahneman named it system 1 and system 2. "System 1 operates automatically and quickly, with little or no effort and no sense of voluntary control" Examples: answer of 2 + 2 =?; Drive on an empty road, orient the source of a sudden sound. "System 2 allocates attention to the effortful mental activities that demand it, including complex computations. The operations of system 2 are often associated with the subjective experience of agency, choice, and concentration". Examples are to tell someone your mobile number or fill out a form. The human structure is based on usage of system 1, only if necessary, system 2 is activated.

This concept of system 1 and system 2 set the basis for a human behavior of decision making, which absolutely defers from a rational decision view. Continuing this idea, the question what prevents a rational decision should be answered with Kahneman/Tversky's prospect theory. The fact that a lot of decisions have both elements, a risk of loss and an opportunity of gain, effects a decision to gamble or to deny. Focussing on loss aversion means, people avoid losses while there is a huge opportunity to gain this particular option. As a result, people deny this option and this is controversial to a rational choice, with a pure calculation of probability. On the other hand the "optimistic bias" means that chances for success are overestimated. Risks are undervalued or not in scope of the decision maker.

Following the process, the maximum utility for the decision should be made. The main question is, if in an organization this strict process will be executed. As H.A. Simon explains, "all decision is a matter of compromise". In an organization with different interests of members, the question of a maximum utility for the decision problem is to be questioned. In an organization, there is never a perfect achievement of targets reachable. The environment of the organization limits the alternatives and the maximum of utility. This opposite view to a rational decision process raises the question how the decision process in organizations is really made. Are organizations as rational as expected or are they emotional driven which have effects on decisions. The behavioral influence in organizations has to be respected and the result is based on this set up.