Tax Knowledge Management meets Digital Revolution

1. digital technologies change knowledge management for the corporate tax department of the future

The "digital revolution" does not stop - no one should be surprised - even before or perhaps especially before knowledge management. But what exactly does tax-specific "Knowledge Management 4.0", "Digital Knowledge Management" or "Knowledge Management in Times of Digitalization" mean? Big changes are always accompanied by uncertainty and a certain risk. In the following article we will lift the veil a little. Our contribution is divided into three parts. First, we will give an introduction to classical knowledge management. Then we will explain the influence of knowledge management and the changed framework conditions in the tax field. Finally, we will present relevant new technologies and venture a look at the new, digital knowledge management world for corporate tax departments.

2. a short introduction to (tax-specific) knowledge management

Knowledge Management (KM) can be found in various disciplines such as economics, information technology, management theory, IT and communication sciences. In addition, there are knowledge managers, or "champions", who take care of the knowledge processes of the company or certain departments, in almost all industries. From our point of view, Knowledge Management serves to provide relevant knowledge and to improve workflows and processes within a company. This enables innovations, knowledge deficits or critical knowledge01 The knowledge is uncovered and an increase in productivity is brought about. Knowledge Management helps to define processes so that corporate goals are achieved sustainably, efficiently and cost-effectively. It also provides a basis for decision-making by management. It also provides a framework for evaluating and incorporating new information and experience.02 This ensures that existing knowledge is preserved and remains available to the company even in the event of retirement or employee migration/fluctuation.

Nonaka and Takeuchi explain how the creation of organizational knowledge can be seen as a spiral process.03 This process begins at the level of the individual and continues through expanding communities of interaction, thus gradually transcending cross-sectional, departmental and organizational boundaries.04 This means that contextual knowledge is first created through interpersonal exchange and personal interpretation processes and then expands within the organization.

Knowledge management is often measured by a perfect database or intranet. But it is much more than that. A good knowledge management system should aim to generate an environment in which employees can easily store, find and apply task-related knowledge related to their job. It also ensures that there is an exchange at all levels. In other words, long-serving employees can share their acquired knowledge with new employees. The necessary platforms for this are available. Furthermore, good knowledge management means that overlaps are discovered and communicated across departments. In addition, specialist cases with possible solutions can be made available to all those involved, thus avoiding mistakes.

The advantages for companies are therefore obvious. New employees can be deployed more quickly and have a lower error rate. Tasks, ideas and services can be completed or implemented faster. Sometimes difficult issues can be solved more easily and efficiently through "lessons learned". The result is lower costs and greater scope for investment for the company.

According to Haddow and Greenfield, "Tax knowledge is essential for any tax practice, no matter how large or small. KM is a recognised science, but in essence, managing tax knowledge is about identifying, collecting and organising it ...".05.

For example, tax and tax law issues and their consequences have very concrete effects on a company. The following questions, among others, arise in the event of changes in the law:

• How complex is the new legislation?
• What are the consequences for the company?
• How should changes in legislation and administrative procedures be interpreted?
• Is there any uncertainty about the tasks to be performed and the changes to be brought about?
• How much effort will be needed to be in line with the new law?

Only when the knowledge gained is put into context with the help of specialist tax knowledge and on a case-by-case basis does one obtain knowledge that is relevant to action.

3. the digital transformation is changing the framework conditions

3.1 The digitisation of the tax authorities

Digitisation" is changing the entire tax environment and thus also the framework conditions under which KM operates. In order to achieve the goals described in the first section, KM must also continue to develop. However, before we go into more detail, let us first take a brief look at the changed framework conditions.

Since the 2010s at the latest, the digital transformation has also affected the tax industry. The 2007/2008 global financial crisis and the subsequent "great recession" confronted many countries with falling tax revenues. In order to compensate for the resulting tax shortfall, tax offenses should be fought more successfully and the "tax hole" closed. In the process, many tax authorities discovered digital technologies for themselves. These promised better, more efficient and more comprehensive control mechanisms. Many companies and private individuals had now digitalised large parts of their business-related activities (communication, accounting, banking, contracting, etc.). With the help of digital technologies, the tax authorities can now effectively analyse the volumes of data generated and generate far more information than has ever been possible in the analogue world. The tax authorities are also becoming increasingly independent of the information transmitted by the taxpayer. The analysis of third-party data, for example publicly accessible sources such as patent and trademark databases, professional platforms in the social media or intermediaries (see for example DAC 6), enables the tax authorities to gain an overview of the taxpayer's activities themselves. Risk analyses and the detection of irregularities are carried out by so-called advanced analytics algorithms in new depth. The object of investigation is expanding from random samples to complete data collections. This heralds a new era of transparency in the relationship between taxpayer and state.06

At the same time, the tax authorities in the various national states are not pursuing a uniform approach to digitisation, not even within the EU. At present, each tax authority is defining its own digital "standard", for example in the design of the data format for the transmission of information. Each tax authority has different priorities and uses different technologies for digitisation. For companies, this results in very different risk scenarios in the individual countries, but these should be managed centrally. In addition, the efforts of some countries to tax the novel business models of the digital economy are leading to an increasingly complex legislative landscape. It is becoming increasingly difficult and time-consuming to keep track of this landscape.

3.2 New challenges for corporate tax departments

This development presents corporate tax departments with the challenge of simultaneously managing the increasing workload, minimizing the new types of risks and maintaining an overview of the regulatory jungle. Moreover, these factors are subject to permanent and rapid change. It can be challenging to ensure the ability of tax departments to act in this volatile, uncertain and complex environment. This includes a growing trend towards home office, "remote work" and flexible working time models. Tax-specific knowledge management is a key factor here. Both the availability and fast, uncomplicated availability of specialist knowledge within the company and the central evaluation and analysis of information arising within the company will be of great importance in the future. This is not just a matter of filtering out tax-specific specialist knowledge, but of any kind of knowledge that can have an influence on the company's tax-related facts. This is the only way to ensure the necessary efficiency and task-specific know-how to understand and well manage the upcoming changes.

What does that mean in concrete terms? In the last part of our article we try to show how specific technologies can take knowledge management to the next level in order to master the tasks at hand.

4. digital knowledge management for digital control departments

4.1 Digital technologies

What is commonly described by the collective term "digital revolution" turns out to be a whole range of very different "technologies" on closer inspection. The field of "Small Scale Automation" (or: "Simple Automation"), for example, mainly comprises Extract, Transform, Load-Tools (ETL) and Robotic Process Automation (RPA). Their main task is to perform simple, standardized and rule-based tasks faster, more continuously and with fewer errors.07 The field of "Intelligent Automation" includes technologies such as Machine Learning (ML), Optical Character Recognition (OCR) and Natural Language Processing (NLP). This area is primarily concerned with the evaluation of large amounts of data with the aim of discovering rules (patterns) or errors (anomalies), making predictions based on past experience08The new tools allow the recognition and processing of human speech (Unstructured Data)09. Another large area is the visual representation of large amounts of data and processes. New tools for visualizations allow to prepare data efficiently and time-saving. This makes it easier for people to grasp the data and provides new insights.

How can these technologies now be made fruitful for digital knowledge management? In the following we distinguish the following three categories:

• Making internal knowledge fruitful
• Preparation of external knowledge
• Provision of data/support for decision making

4.2 Making internal knowledge fruitful

In recent years, the amount of digital data generated within tax departments has increased dramatically for many reasons. In combination with the modern technologies shown, quite traditional tasks of knowledge management can now be supported and significantly improved. Tax decisions and specialist processes can be automatically enriched with data and information, thereby optimizing output.

First of all, however, the appropriate conditions must be created. Two central factors here are the standardization of processes and the creation of a uniform data basis, i.e. "clean" data management. Only when this "foundation" is in place can the truly advanced technologies unfold their effect. The goal is to transfer knowledge and processes into the area of structured data.

Standardization covers both the area of specialized processes and the area of knowledge capture. For example, similarly treated facts and cases must be identified and conforming steps for processing must be established. Subsequently, Knowledge Management can, for example, develop, provide and centrally manage "drafts" or "pre-filled forms" in order to accelerate these steps. Likewise, available knowledge must be systematically recorded in a uniform manner. One conceivable solution would be interactive chat offers, for example, which employees regularly ask questions on specific issues and automatically process the answers in a specific format. Chat bots would be available to employees if designed accordingly,10 also more entertaining and motivating than, for example, collection via traditional questionnaires or wikis.

Just as central as standardization is the storage and organization of information (data management). Newly collected data and already existing knowledge in various databases, shared spaces or intranets must be brought together and standardized in format. We know from practical experience that this is an immense task that should not be underestimated.

In both areas, Robotic Process Automation and ETA tools can help to prepare and cleanse the data accordingly. With their help, recurring processes can be automated easily and promptly. Optical Character Recognition Tools can help to translate any remaining analog information into the digital sphere.

Once this foundation is in place, algorithms can take over the task of providing relevant information, creating a whole new work experience for employees. For example, machine learning algorithms are able to distill "rules" from existing data and categorize and assign the information contained therein. In this way, the algorithms can "learn" to identify task-specific information and transfer it to the right "subject matter expert". Tedious searches in numerous different information sources are thus a thing of the past.

With the help of Natural Language Processing, which is making great strides, even unstructured data sources with significant knowledge could be automatically evaluated and processed. These include, for example, the internal circulars, newsletters, tickers or other information mails that are frequently circulating in companies.

From a purely technological point of view and from the singular perspective of knowledge management, it would theoretically be possible to extract relevant knowledge, for example, directly from communication or certain work activities (presentations, legal opinions, etc.) of employees. However, due to its proximity to the monitoring of employees, such an approach will often collide with other objectives such as data protection and is therefore rightly viewed critically.

Therefore the human aspect should not be underestimated in this area. In the practical implementation, motivation and corresponding training ("upskilling") are the decisive factor. Only if the employees in the company and in the departments share the tacit and context-based knowledge (professional experience, expertise, networks, etc.) that is often available and are also enabled and motivated by new, entertaining (keyword: "gamification"), easy-to-use tools and platforms, digital technologies are able to create a new work experience. Because, quite honestly, nobody likes to spend hours searching for information that is urgently needed at that moment. In this context, it can be the task of knowledge management to encourage agile learning and the necessary acquisition of new knowledge, for example by establishing digital learning management systems or creating incentives for "micro-learning" close to the workplace.11 Change management also plays an important role here. This should not be missing in a good knowledge management program.

While the fertilization of internal knowledge by the human component gains in complexity, the following category is more likely to face technological challenges.

4.3 Processing of external knowledge

An entity generally has no control over the format and method of providing external information. For example, courts, tax authorities, parliaments, specialist publishers, external service providers and news agencies prepare their information in very different, individual ways. Some allow connection via programming interfaces (API), others do not. At present, everyone is (still?) cooking their own little soup. This poses particular challenges for the automated collection and allocation of knowledge.

The objective for the knowledge management in this category is relatively clear. Tools and platforms whose algorithms track newly enacted laws in different countries and represent tax-relevant changes resulting from that clearly are desirable. The same applies to the evaluation of court decisions and the deduction of consequences from that. Equally desirable would be the automated checking of articles, comments and similar contributions in the numerous publications of the relevant specialist publishers as well as the categorisation and task-specific provision of this information to the Subject Matter Expert.

The processing of this type of unstructured data is primarily the field of computational linguistics (Natural Language Processing/Understanding). It is about the algorithmic processing of human language as a special field of machine learning. The application in the tax context is particularly challenging. Here the challenge is that the content of the text must be understood by the algorithm. This is different from the categorization of e-mails as "spam",12 where only certain patterns and characteristics need to be recorded, a deeper understanding of the content is required in the tax field, which is often very complex. This requires a semantic analysis of the texts. This means that sentences, phrases, expressions, formulations must be assigned certain meanings. Although great progress has been made in this area recently, it remains a complex undertaking.

In a next step, predictions about uncertain future events can also be made from this (past-related) information with the help of "predictive analytics". This applies, for example, to the calculation of the probability of a positive or negative outcome of legal proceedings, tax proceedings or other legal disputes with the tax authorities. Resources can thus be optimally planned and efforts with little prospect of success avoided.

In view of the technological complexity of this subject area, many companies will be faced with the question of whether it would not make more sense to use the products of highly specialized external service providers instead of developing their own solution. An agile market has emerged in recent years. Pure NLP providers compete with the cognitive solutions of large players, with insight platforms and the special solutions of large consulting firms, as well as with smaller providers in the tax field.

Another important point - for the reason described in the previous section - is the centralised review of the tax authorities' digitisation efforts. Which authorities in which countries are making what progress? What business data are the authorities able to collect? Which company independent information sources are accessible? What capabilities are available for the analysis of large amounts of data? How do they decide which companies should be more closely monitored or audited? In what way and in what format do they require information from the company? Having a central overview of the answers to these questions will be of central importance for companies as digitization progresses. This is the only way to minimise tax risks, react promptly to official requirements and understand or effectively challenge decisions of the tax authorities.

4.4 Provision of data / support in decision-making

In addition, knowledge management can also make a further, important contribution to the operative business and the upper management level. The data generated in the course of successful digital knowledge management can be made available to other departments and further processed.

A frequent challenge for tax departments is to raise awareness of tax issues among internal company decision-makers. Tax knowledge in particular is indispensable for successful corporate planning. Information gathered in the context of knowledge management, such as the respective tax rate or the tax expenditure incurred in different countries, can be put into relation to other information relevant to decision-making, such as the sales ratio. New visualization tools provide the management with quickly graspable, graphically prepared tax analyses for decision-making. Modern dashboards also make it easier to run through various scenarios. For example, what would be the consequences for the company if a concrete tax regulation were to be passed that is currently in the planning process, or what would be the consequence of the imposition of new sanctions in a commercial dispute?

The integration of tax information also offers new possibilities for the operative business branches. For example, tax-relevant knowledge about free trade agreements could be directly incorporated into supply chain management coordination programs, thereby significantly reducing costs and improving performance. Similarly, automatically recorded and processed internal and external knowledge could flow into warnings, instructions and instructions for action in the operating business and significantly reduce its compliance costs. Probably to the delight of all involved.

5. conclusion

In a digitally accelerated (tax) world, where data volumes are increasing rapidly, regulations and the environment are changing rapidly and technology is gaining significant influence, knowledge management must significantly support the digital transformation of the corporate tax department. Only then will it be able to master the challenges ahead. As an automated link for knowledge from many different departments, digital knowledge management represents a separate strategic business unit. Channeling all the information flows within a company, evaluating them and making them available in a structured form (and as structured data) represents a major challenge for man and machine. However, if this is achieved with the help of a modern digital knowledge management system and the corresponding programs/tools, completely new possibilities for innovative approaches open up. In any case, there is no reason for uncertainty, because in the end the human factor remains decisive.

"An investment in knowledge always pays the best interest." Benjamin Franklin (1706-1790, one of the founding fathers of the United States)