Data is the fuel for business processes.
It has been so throughout time, from the ancient stone tablet to the modern era, but data is never nearly as important as it is for the digital transformation.
The digitalised company is driven from its core by data, and digital business processes bind together all partners in the respective digital ecosystem. Digital companies virtualise their offering, their customer and business partner relationships, as well as their sales channels. Similarly, data represent every interaction of the network’s participants. Therefore, data has become an essential asset for the company.
Digitally processed data have been used since the sixties to document master and transactional data of the company. Since the eighties, they have been increasingly used for ex-post analysis. Alongside digitalisation, two new key disciplines enter the stage: the prediction of future market behaviour, and automated individual decision-making in real time.
In general, data represent business objects for a defined purpose in a simplified manner. Data objects are based on a reality space, the scope of which will be defined in order to support given business goals, together with the necessary entities, attributes and activities. In the digital environment, however, this reality space tends to become larger, as business processes are transferred to the customer side and become more influenced by external actors. Every activity in digital business leaves traces in the form of data, which are stored within the company. Internal data are those resulting from the execution of transactions with customers, products, services, partners and so on. They also represent business processing, and thus status and rules of business processes themselves that run automatically in the digital ecosystem.
This way, all fully executed and also incompletely executed activities are documented. Internal data are the foundation of the business, because they are also legally relevant (e.g., for contracts, financial data, risk calculations, compliance). Beyond that, they provide detailed information on all activities associated with the business. Unstructured internal information is becoming more important than ever, as it is considered highly relevant for the appropriate interpretation of customer needs.
Data and processes are connected in a true symbiosis in order to create value.
Data last for a long time and must usually be accurate, at least when they are legally or financially relevant for the company. Due to this fact, a structured, precise and typically rather cautious data development process is required. Processes related to these data objects, however, are much more dynamic and volatile, because they are – especially in digital businesses – affected by outside factors, requiring a more agile approach. Consequently, the objects’ attributes (data) should follow a different development model from activities related to such objects (processes), allowing for different development speeds.
Publicly available information is increasingly attracting the interest of digitalised companies, in order to access both new customers in well-known target groups and even new markets with their potential for accelerated growth. At first sight, such information represents seemingly unrelated pieces of a puzzle of target objects and entities (e.g., people, products, companies). The collected pieces must be evaluated in terms of relevance and likelihood of fitting into a desired, initially unknown object. Additionally, behavioural information on the detected entities is of great importance, because it hides business-related events and is derived through analysis of previously collected raw data.
Unlike internal data, publicly available data are largely uncontrollable; data quality, reliability of the source (e.g., content changes, structural changes, temporary or permanent non-availability, update frequency), redundancy of the same information across different sources and so on, affect the collection and analysis of data, which complicates matters.
The collection and use of publicly available data require new technologies, which many companies have yet to master.
This is also true for the analysis, for example, of unstructured content from non-secure, potentially amorphous and independent data sources, often with unreproducible or non-controllable changes, possibly describing initially the ‘world’, but not necessarily the known or searched-for (yet sometimes still unknown) business opportunities.
From many seemingly unrelated pieces of information, data need to be processed into a structured and analysable form.
With self-learning mechanisms, they are merged and transformed into complex structures mirroring reality. Therefore, new disciplines are needed for processing publicly available data, disciplines which are, by the way, to a great extent similar to those necessary when creating a business ecosystem from internal unstructured content, an ecosystem which finds its cornerstones, business definitions and foundation in the new data. Such mechanisms include:
- Source intelligence to scout and make available relevant sources
- Entity intelligence to form objects from unknown, initially isolated data sources
- Context intelligence to connect external and internal objects and to associate with the business context of the digitalised company
- Continuous market monitoring to derive signals, trend values and forecasts from previously developed data structures for past, present and future business opportunities.
In such a process – which needs to be elaborated – new data are generated that must first be transferred into the right business context (business ecosystem) in order to create benefit for a specific use case. Digital enterprises must learn to deal with concepts such as the relevance of information or the probability of correctness of their business objects. This applies both to the creation and the usage of data. Derived actions based on objects that exist only with relative certainty and processes that are based on probabilistic methods cannot be classified into categories like ‘exact’ or ‘correct’. This is not so much a technical issue, but rather a cultural change for the decision makers in the company.
In the process of digitalisation, direct personal contact with targeted clients and business partners is purposefully reduced so that, in return, every potential customer on the planet can in principle be addressed.
Thus, personal judgement as well as the traditional gathering of information (e.g., recognition processes, identity and sentiment/ mood, desire to buy) must be substituted by data-driven alternatives.
‘Analytics’ is the key to creating these substitutes artificially while itself producing large amounts of derived and calculated data.
