Introducing self-service BI
Promoting democratic data analysis
The need for specialist departments to obtain, merge and analyze data independently of IT did not just emerge with the trend towards self-service BI. Since business intelligence has found its way into companies, the demand for self-determined data analysis has become louder. Their integration in the company can be successful with just a few resources.
With the introduction of central BI solutions and data warehouses, the original data sovereignty of the specialist departments was transferred to the IT and controlling departments. From then on, reports were created centrally and only adapted to the wishes of users through laborious application processes.
Gradually, this image of BI in companies has changed again. More and more users are now using modern tools for self-service BI such as Tableau or Power BI to analyze data independently and create reports. However, data provision and data modeling usually remain a task for the IT department. This restricts the freedom of analysis for experienced users. This division of tasks neither corresponds to the self-image of many business users nor to the propagated requirements for agile BI.
Requirements for a modern BI
Self-service BI should enable business users to create their own analyses and reports and they should be able to share these with other users as easily as possible, largely without recourse to IT. They should also be able to connect additional data sources to the data model provided if required. In this context, IT ideally takes on the role of an internal service provider. It provides the architecture and software environment as well as an initial data model in coordination with the specialist department.
Users need intuitive software for self-service BI. It must also provide a data basis that is prepared in such a way that the structure is understandable and the data is up-to-date. The growing needs of users in the context of digital transformation are increasingly presenting IT departments with new challenges. Specialist departments often need to expand their database at short notice and perhaps only once in order to be able to react to changing market developments. However, IT departments have standardized processes and concerns about data protection and performance. These can often put the brakes on users' desire to analyze their own data. Self-service BI does not mean that all users can access all data at will. Internal and external data protection requirements must also be guaranteed. This is ensured by a suitable authorization concept.
The task for IT is to introduce suitable architectures and processes. On the one hand, these must be able to satisfy the needs of business users, but on the other hand they must also safeguard the legitimate interests of IT. New forms of collaboration between IT and business are therefore increasingly necessary, which can be based on agile principles, for example.
Standard reporting meets freedom of analysis
On the one hand, with the increasing use of self-service BI in companies, the possibilities for carrying out ad hoc analyses are growing. On the other hand, there is still a need for standardized and reliable reporting that is made available on a regular basis. It should fulfill the basic requirements and form the basis for internal exchange. An experimental system is needed to further develop this standard. This should allow specialist users to test and further develop innovative solutions as part of a self-service BI. However, this should not mean that a shadow BI is set up in a bimodal BI at the IT department. Rather, it requires IT to provide a sufficiently broad framework in which the advantages of self-service BI can unfold freely.
Business intelligence must consider the target group
The target groups of BI and self-service BI in companies can basically be assigned to three user types, whose requirements for the autonomy of data procurement and data analysis differ fundamentally.
Above all, management wants to be provided with a quick overview of the key figures and KPIs in reports and dashboards. Due to the often low level of IT affinity, the expectations of the data basis are that it can meet the information requirements in a valid and high-performance manner.
The specialist department requires detailed information about business processes and would like to be able to view and analyze these in as much detail as possible from a wide variety of perspectives. Self-service BI and data discovery are essential requirements that the data basis must fulfill. Business users themselves should be able to link data correctly with each other so that the evaluations created contain the correct information. The database provided by IT can support this with a homogeneous nomenclature. Increasingly, modern tools for self-service BI also have options to provide users with useful support in this task.
The data scientist usually has good knowledge of data linking and data modeling. The expectation of this user group is maximum autonomy and an attribute-rich data model whose entities can be linked together as required. Large amounts of data are often retrieved within complex queries. Consideration should therefore be given to providing these users with their own playground, keyword Data Lab, which is decoupled from the actual BI system.
Infrastructure as a success factor
Various aspects should be taken into account when selecting the appropriate infrastructure for the increased introduction of self-service BI.
At the outset, it must be clarified which user group is the focus and to what level of detail they require access to data in order to fulfill their analysis needs. Does modeling according to Data Vault make sense? Or are classic Star or Snowflake models sufficient? When making a decision, it is advisable to think ahead and also take new technological developments into account.
Data growth and the required timeliness of data provision are other important influencing factors when choosing the right infrastructure for self-service BI. Should data be available in real time or near real time? In these cases, it may be advisable to virtualize data rather than provide it persistently. Technologies for implementing a logical DWH offer corresponding options and often already offer a large number of connectors.
Design of the data model
When designing the data model, particular attention should be paid to a uniform nomenclature in order to support users when creating analyses and reports. It may also be a good idea to store the definition of key figures and dimensions as well as calculations in a business intelligence wiki that is accessible to all users. This promotes a uniform understanding and avoids misinterpretations. If users are to enrich the data with additional attributes, the data model should contain the natural keys that users need to link the data with internal or external data.
When selecting a database system, business users should consider not only their expectations in terms of data volume and data growth, but above all the performance requirements of the queries. Based on these considerations, in-memory or column-oriented databases may be a good choice. If a separate data lab is set up, it should also be ensured that it is assigned its own hardware or at least its own instance so that data scientists do not drag down the performance of the BI system.
Selection of a solution for self-service BI
When selecting suitable software for self-service BI, there are a few criteria to consider. To what extent does the application fit in with the existing infrastructure and how much maintenance is required from the IT department? The extent to which the application can relieve the burden on IT by making users independent in the context of self-service BI also needs to be examined. Here, it is particularly important how intuitive the application is for users to operate, which data sources can be connected directly and to what extent the preferred solution also supports users with data linking and data preparation.
Unless the software achieves a high level of user acceptance and motivates continuous use through simplicity and speed, self-service BI initiatives can quickly fade away and lose their impact. In this context, it is also advantageous if the software supports team collaboration and the simple exchange of analyses and reports. Self-service BI can only develop its true potential through teamwork and discussion.
A suitable solution for self-service BI must at least fulfill the following requirements:
- Intuitive and fast creation of analyses and visualizations
- Connection of a variety of data sources without IT support
- In line with the IT infrastructure and requirements of the specialist departments
Finally, the decision for a specific solution for self-service BI should always be made in accordance with the decision for a suitable overall architecture in order to promote the sustainable success of the company's overall BI strategy.
