Data and networking: both a necessity and a choice
Data // Demand for an efficient and innovative use of data is growing steadily. Users’ desire for networked ecosystems often poses an dditional challenge for companies. A sustainable data strategy is the central building block of a company’s success in the medium term.
Data can open up new perspectives, cast doubt on what you know, and reinforce your impressions. In particular, the application of big data solutions the technology that processes amounts of data that are too complex, too big or too weakly structured to be evaluated using conventional methods is a key component of many digitalization goals.
Big data, small data, any data
Analyses of terabytes of data, made possible thanks to technical progress in infrastructure, can deliver a crucial competitive edge when used properly. By identifying patterns, it is possible to establish a variety of additional validation measures and conclusions, for example. Whereas, in the case of targeted individual insights gained from the datasets,key themes are prioritized, with big data there are often very few stipulations about the anticipated result. This means smaller datasets (small data) are often enough to provide a clear and simple analysis of, for example, correlations and prognoses concerning price, advertising measures, time point and demand for a product.
Big data analyses include additional exogenous variables such as age, gender, income and new data sources in a heterogeneous structure and form. This makes it possible to produce highly personalized offers, for example. In more ambitious projects, the data to be used still isn’t known at the outset of the analysis, and is evaluated as part of the analysis. In addition to the increased data volume, big data also deals with various data structures and types. The complexity of the data and the analyses therefore rises dramatically.
The larger, more complex and more heterogeneous the data amount to be analyzed, the harder it is to analyze. The attempt to utilize all the data available in the company (“any data”) is resource-intensive and often leads to hazy results. It is therefore vital to define the central theme for the data analysis at the start of any data initiative. If the theme can be evaluated using a smaller data set, it is particularly important to ensure that you select the right subsets.
Networking and platforms
It’s a given that data amounts grow exponentially. Where an average household produced two terabytes of data back in 2013, in 2020 this figure has risen to 14 terabytes. At present, it is expected that there will be a global data volume of 149 zettabytes (1021) by 2024. Compared to today, this is an increase of over 250 % (see chart).
The growth in data volume, however, is just one side of the coin. Individual data should never be viewed in isolation. According to an estimate by Cisco, 50 billion devices are connected today.
Radio, sensor, and near-field communication technologies make it possible to get connected in virtually every aspect of our lives: mobile payments, fitness trackers, e-ticketing, GPS navigation, you name it. We leave data behind us every-where we go, accessible to both providers and users. This can take all sorts of different forms and structures. Dependencies and ubiquitous computing also make things more complex. Platforms centralize the data and enable companies to reinvent themselves through data-driven business models.
The data jungle is getting ever more impenetrable and with this comes a greater demand for handling data in the right way in order to unleash untapped potential. A company-specific data strategy offers a platform for this potential to unfold, and is an indispensable companion in the quest for efficient data-driven value creation.
Data strategy. But how?
A functioning data strategy is based on key themes that can be allocated to strategic business aims; it originates in the corporate goals and visions. Unspecific goals for data evaluations are not (yet) fully functional when it comes to data analyses, and frequently also lead to unspecific results. A good data strategy requires a holistic picture to be drawn up. Technology, governance and processes must be developed and put into operation together:
technology requirements are geared towards the needs of the various interest groups and the business processes. And there are as many options for technical implementation as there are data points. Whether we are talking about a data lake, data marts or standardized solutions available in the cloud, it is essential to define requirements and compare the cost/benefits as well as future perspectives. A good place to start is with analyses of heterogeneous data sources, plus the necessary flexibility and options for evaluation.
Governance sets out the in-house framework conditions for the data strategy to be implemented successfully. Key elements of this include regulated responsibilities on data management and data sovereignty. The introduction of a (virtual) data organization with roles such as chief data officer, complemented by company-wide data policies on operationalization, is a valid method depending on the size of the company.
Well-defined processes are the mainstay of a successfully implemented data strategy. It is vital to develop and improve core management processes in collaboration with stakeholders. As a result, a culture of identification with data-driven processes can then be established.
Data volumes are getting bigger, and it’s getting more complex to manage them. Period. There is no one universally applicable solution for dealing with the “datafication” of the world, as Germany’s Institute for
the Future rather bluntly puts it. Through target-oriented motifs, selecting the right dataset, and an iterative approach to introducing the right technology, governance and processes, it is possible to build a sound foundation for data-driven value creation. Customized data strategies are a valid means of shaping your business.