Corporate Data Quality

1.1 Trends in Digitization

New forms of information technology are changing every area of the economy as well as of society overall, as researchers, such as Kagermann (2014), have analyzed them from the view of the Federal Republic of Germany. We have summarized the development into four trends (Figure 1-1).

Figure 1-1: Mega-trends in Digitization (authors’ illustration)

1.1.1 Penetration into every Area of Life and Economy

According to the International Telecommunications Union, 2.9 billion people used the Internet in 2014, meaning roughly 40% of the world population (ITU 2015). The technological innovations of the last 15 years are responsible for this penetration into both the private and business areas.

· Mobility: wireless networks and miniaturization of computers and other components, like sensors and cameras, are bringing digital services to the location of use, whether in private life, such as recording a hiking route, or in business, such as remote diagnosis of a machine.

· Usability: touch screens and many other improvements in details, like logging into digital services through a Facebook account or vocal input and output systems, have drastically reduced the threshold for usage. Other efforts to ease usage, like data glasses (such as Google Glass), control using gestures and detection of eye movements, have also been distinguishing themselves.

· Content and community: whether individually (such as through blogs and tweets) or in combination (such as Facebook), innumerable people have been producing a volume of content in the form of written words, pictures, audio and video files, which can only be reviewed by machines. YouTube recorded more than one billion requests for video clips per day in June 2014[2]. Facebook recorded roughly 1.3 billion active users in March 2014[3].

· Communication: this content is being exchanged synchronously, asynchronously, privately and for business. Accessing email, text messaging, and social networking are among the top four most-popular daily activities of smartphone owners in the United States in January 2014 (statista 2015). Visual communication has been increasingly supplementing more common audio telephony and instant messaging services (such as WhatsApp) are frequently used in addition to email messages.

· Big data: unexpectedly high volumes of data are the result of the penetration of the economy and society overall by digital services, while at the same time they are the foundation for the personalization of services, especially those based on providing location information (Figure 1-2).

Figure 1-2: Online Activities for Private Purposes over the Last Three Months in Swiss Households (Froideveaux 2012 p. 25)

Almost one quarter of the world population used smartphones in 2014 and in both North America and Western Europe, about 50 percent of total population used smartphones in one way or another (statista 2015). Digital networking has had an enormous impact on the formation of people’s opinions in their political, economic and private affairs. From the view of data management, the following aspects (among others) should be taken into consideration.

· Data security: until now, the Intranet was considered the perimeter, meaning the boundary where data had to be secured. This boundary has dissolved and companies must go beyond it to protect not only networks and application systems but also enable data objects, which themselves must know who should be permitted read access and who should not (O'Brien 2014).

· Data production: classically, companies have acquired data centrally (such as customer data collected through a central, internal marketing service). Due to the spread of social media and social networks however, data consumers are increasingly becoming data producers (Strong et al. 1997). Customer data can be acquired directly from the customer or from external agencies by smartphones or onsite tablets. Employees expect that the data will be accessible from everywhere.

· Streams instead of records: millions of users generate data flows in social networks and through social media. This represents new challenges to companies, because the traditional processing of data was oriented on transactions, meaning that individual records were written persistently to databases. However, the increasing usage of data streams from social networks, such as from the cyber-physical systems in Industry 4.0, can no longer be updated incrementally, but rather must be followed continuously (BITKOM 2014).

1.1.2 Industry 4.0

The term “Industry 4.0” stands for the Fourth Industrial Revolution, meaning the merger of the physical and virtual worlds through so-called “cyber-physical systems” (Bauernhansl et al. 2014). The data will be acquired more precisely and in more detail than previously without time delays or the help of people. Machines are becoming capable of working with the Internet, assuming the tasks of production and data distribution independently and the data, which has only been available in the factory for a long time, is becoming accessible to the entire company and its business partners (Figure 1-3).

Figure 1-3: Data Acquisition at the Interface between the Virtual and Physical Worlds (Fleisch 2010; Wahlster 2011 p. 5)

Industry 4.0 scenarios are changing the basic handling of data both within and between companies. Three issues make this clear.

· Decentralization of data management: things are becoming “smart”, meaning that they produce, use and have an increasing amount of data and rely less on central control systems. As a consequence, things are assuming increasing importance in the distribution of data without requiring central computers.

· From the class to the instance: the focus of electronic data processing has been on traditional classes of things, meaning articles with a certain Global Trade Item Number (GTIN) or products with certain material numbers. Industry 4.0 now means that each instance of a class of products can be identified, meaning individual hydraulic cylinders or the individual bottles of hydraulic fluid (Österle and Otto 2014).

· Continuous combination of the flow of information and goods: traditionally, industrial data processing has targeted the flow of information and goods to certain control points. One example is the goods receipt record in the central warehouse for the delivery of goods. Industry 4.0 scenarios use RFID technologies, for example, and enable access to the status and location information for the individual products at any time (Österle and Otto 2014).

The inBin intelligent container developed by the company SICK[4] together with the Fraunhofer Institut für Materialfluss und Logistik (Fraunhofer IML, Institute for Material Flow and Logistics) is one example of an Industry 4.0 application. The inBin knows its location, records its environmental temperature and arranges for its own pickup (Figure 1-4).

Figure 1-4: The inBin intelligent container (Fraunhofer IML 2015)

A powerful data management system that fulfills the following requirements is the prerequisite for the success of Industry 4.0 in individual companies as well as across supply chains.

· Mastery of the volume of data: the data management system in the company must be capable of processing and reasonably evaluating the amounts of data (Wrobel et al. 2014).

· Decentralized data processing: when machines, containers, freight and so on become intelligent, this means that they will have to assume the tasks of processing their own data. Data analysis, aggregation and provision therefore no longer occur centrally in the Enterprise Resource Planning (ERP) and data warehousing systems, but rather locally onsite. Central corporate data processing will be supplemented by a network of decentralized intelligent devices (Aggarwal et al. 2013).

· Determination of data standards: advantages in terms of time, expenses and quality through the use of cyber-physical systems and automated data interchange can only be realized when standards for the description and exchange of data have been established. These standards must apply internally to the company, at least, and their applicability across entire supply chains would be better (Otto et al. 2014). The MobiVoc initiative developed, for example, a data vocabulary for new mobile solutions[5].

1.1.3 Consumerization

Every one of us today uses a number of different consumer services that support various situations of our lives (Österle 2014). Figure 1-5 depicts ten areas of life in which people use digital services, from support for navigation to listening to music, from comparing prices to controlling the illumination of homes remotely. As an example, the area of communication has been expanded with two additional layers in order to provide an impression of the multitude of services. A more detailed, but still incomplete mind map of digital consumer services can be found at il.iwi.unisg.ch/appmap (Amiona 2014).

Figure 1-5: Ten Areas of Life and Examples of Digital Services supporting Them (Amiona 2014)

At the same time, consumers increasingly expect digital services to be customized to their individual needs. Companies are reacting to the consumerization of information technology by orienting their business processes on the consumers’ needs, thus the consumer process. This process consists of all activities that an individual accomplishes for the fulfillment of various needs (such as purchasing, athletics and traveling) in a certain situation of the consumer’s life.

Consumerization leads to a new role for consumers in economic life (“consumer centricity”). They are no longer the terminal or transitional points in the unidirectional flow of goods and information, but rather directly affect public opinions of products and companies through platforms like FoodWatch.org and are now acting both as consumer and producer of goods and services. Examples include the floods of indignation that descended on Nestlé because of the use of palm oil in KitKat chocolate bars and the crowd sourcing of programming services.

Figure 1-6: Network Analysis of the Flow of Product Information at Beiersdorf (Schierning 2012 p. 9)

Figure 1-6 depicts an example of how the flow of product information has changed over a period of five years at Beiersdorf, a manufacturer of consumer goods. On the one hand, the number of participants in the company network increased from 2007 to 2012, because companies like Apple and Google, as well as online retailers like Zalando use and distribute Nivea, for example. Borrowing a term from ecology, the expanded corporate network can also be viewed as an ecosystem. On the other hand, consumers have moved from the periphery to center with regard to control of the data, since nearly every company in the network interacts with consumers (Schierning 2012).

Nestlé not only maintain classic corporate data systems, but also consumer data. Nestlé had 94 million fans on Facebook and 16 million clicks to view their Contrex video on YouTube. Data from online shops, where Nespresso sold more than 50% of their coffee packets for example, should be added to these figures.

Consumer centricity means a rejection of the traditional corporate-centered view of the end customers for companies. The design and improvement of interaction with consumers is no longer the only focus from the view of the companies (Inside-out approach), but also the integral consumer process across the boundaries of individual companies (Outside-in approach).

Consumerization places new requirements on the management of data.

· Data ownership: who does the data belong to? This multi-facetted discussion about data protection and statements like those of Mark Zuckerberg of Facebook, that data security is no longer a social standard (Johnson 2010), indicate that the trend in consumerization has surpassed the traditional understanding of ownership and possession of immaterial goods. So-called “data brokers” collect personal Internet data in legal gray areas (Anthes 2015). For companies, this means that they must formulate a uniform legal position in regards to data protection. Legislatures are being asked to crate uniform frameworks.

· Data integration: people no longer use a single channel for communication in order to connect to a company, but rather use multiple channels. The Swiss retailer, Migros, identified nine different channels (offline and online) through which they communicate with consumers. This diversity includes traditional letters, online shops and email and text messages. Because consumers expect to be uniquely identified through all channels and to get the same prices and rebates on Migros products, the company had to provide consistent, current and complete data about their customers and products across all channels (Schemm 2012).

· Combinations of “structured” and “unstructured” data: as a consequence of consumerization, companies are no longer only providing information in traditional alphanumeric formats, such as descriptions, weights and prices about the products, but are more often providing product video clips, marketing information and lists of active ingredients. The differences between product data (which is generally stored in central ERP or Product Lifecycle Management (PLM) systems) and multimedia product information (which is frequently distributed using a number of internal application systems and external service providers, such as advertising agencies) can no longer be maintained (Österle and Otto 2014).

1.1.4 Digital Business Models

The penetration of digital services into the economy and society overall and in particular, of the industrial and consumer sectors, will lead to new types of business models outside of classic companies[6]. Examples from the area of consumer services include Google as well as Airbnb, idealo and many other companies that bring a large number of consumers and business customers together with a large number of providers. These companies have been assuming a role as brokers between the supply and demand for services from a variety of participants. From a more technical point of view, one frequently speaks of the “Internet of services”. Four developments characterize these business models.

· Focus on data: new business models for the Internet-based service economy use data as a strategic resource (see Figure 1-7). For example, Deutsche Post provides high-resolution geographic information for retailers, insurers, real estate agents and public administration and other customers (data as the product)[7] through the GEOVISTA service.

· Industrial convergence: traditional sector boundaries are losing their significance. Google is one of the innovation drivers for autonomous cars; classic vehicle manufacturers are potential licensees for this technology. Amazon has transformed itself from a book retailer into a fulfillment expert, who offers special capabilities like scalable IT infrastructure services or provides logistics service to companies from many sectors and even consumers.

· Hybrid services: often, digital business models combine digital services with classic offline services. One example involves the car-sharing models that combine digital rental and provision of cars including payment (generally supported by smartphone apps) with the classic services of mobility.

· Consumer process: the Internet of services is oriented on the individual, meaning the individual consumer, the patient, the service technician or the shopper. The goal is “end-to-end” support for life situations, such as purchasing, jobs, mobility, therapy and health care (Österle and Senger 2011).

Figure 1-7: Digital Business Models (Brenner and Herrmann 2012 p. 20)

1.2 Data Quality Drivers

Digital business models and the Internet of services are based on the resource of data. Data quality is therefore no longer a question of “hygiene” or even of internal use by line departments, but rather has become critical for operational excellence. Data quality is defined as a measure of the applicability of data for fulfilling certain requirements in business processes, where it is used (Otto et al. 2011). The following material will constantly treat “data management” with special consideration of data quality management.

The most important drivers for quality-oriented data management include:

· 360-degree view of the customers

· Corporate mergers and acquisitions

· Compliance

· Reporting systems

· Operational excellence

· Data protection and privacy

1.2.1 A 360-degree View of the Customers

Knowledge about the customers is the starting point for marketing and sales, as well as for the development of products and services. For this reason, companies must be capable of gaining access to all information about the customers’ needs. For consumers, such information includes Internet surfing behavior, purchases and peer groups in social networks; for business customers, their addresses, subsidiaries, contact information and the name of their contact people as well as data about purchased products and existing contracts.

Bühler is a globally active manufacturer of production systems specializing in the industry of food. They make digital customer profiles available to their employees in the Customer Service and Marketing departments. These profiles answer questions such as:

· How much revenue has been made with the customer (and all of its subsidiaries) in the current fiscal year?

· Which of our systems and services are used by the customer and at which locations?

· When will maintenance contracts expire?

· Which employees made contact with which customer employees in the last three months? What were the results of such contacts?

· How profitable is the customer relationship?

The 360-degree view of the customers places many requirements on quality-oriented data management.

· Data quality: customer data must be consistently, currently and completely available for all functional departments (Marketing, Services, etc.).

· Data lifecycle: how customer data is acquired by the company, where it was acquired and stored, who will modify and change the data and which business processes and systems use the data must be clearly defined.

· Data security: with consumer data, provisions must be made to ensure that data protection provisions will be maintained, including that customer data will be deleted upon request.

· Data governance: companies must clearly determine who will be responsible for which customer data in the company. Is the field staff responsible for customer addresses or the internal Marketing department? Can service employees change the customer status to active? Who will collect email messages with this customer or their Facebook pictures?

1.2.2 Corporate Mergers and Acquisitions

Corporate mergers and acquisitions are important tools for corporate strategies. In the chemical industry for example, BASF has taken over the electro-chemical division of Merck, the fine chemistry company, Orgamol, the catalyzer manufacturer, Engelhard, the building chemistry division of Degussa and the special chemical group, Ciba, since 2005. These acquisition have been integrated in the uniform application systems and business processes.

Nestlé represents another example of corporate integration. The company operates more than 2000 different brands, which are produced in almost 90 countries and sold in more than 190 countries[8]. Of their total revenue amounting to more than 92 billion Swiss franks in 2013, 93% are processed through the GLOBE central Enterprise Resource Planning system. Figure 1-8 depicts several pieces of important information about GLOBE.

Figure 1-8: Important Information about the GLOBE Central System at Nestlé (according to Muthreich 2013 p. 18)

The GLOBE program has pursued three goals since it began operations in 2001, specifically: the company-wide use of best practices based on shared business processes, the introduction of standardized application systems and the use of data as an asset. The prerequisite for this is a powerful data management system, which has integrated many corporate acquisitions over the last few years in particular.

· Data standards: binding specifications for the acquisition, maintenance and use of master data, such as customer, supplier and material and product information, must be applied.

· Data acquisition at the source: due to the size and complexity of the company, data cannot be acquired centrally, but rather as close as possible to its source.

· Data quality: the size of the GLOBE system does not allow contaminated data to be introduced into the system and then to be cleaned up afterwards. Instead, the data must be correct when it is entered the first time (first time right principle).

· Data integration: an integrated system like GLOBE does not allow for “data silos”, but rather must work with all business areas, functions and markets using an integrated database. However data integration can only be developed after the company has changed their approach to data from “my data” to “our data”.