Harnessing Cloud Analytics

Business Intelligence as a self-service (SSBI)

For several years now, companies have routinely cited the topic of Self-Service Business Intelligence (SSBI) – in other words, the independent creation of BI content in specialist departments – as being one of the top topics in the BI environment. Based on the demand for data democratisation, an international company decided to implement a cloud-based SSBI approach in the context of a cloud-first strategy.

The technical driving force came from the world of strategic planning and marketing, where competitor data (competitive intelligence) in particular needs to be quickly and flexibly available and analysed together with the company's own data. The SSBI implementation was managed by the IT department in close cooperation with the specialist department. The project took six months, and culminated in the general approval of the SSBI environment and the accompanying processes.



    In the first step, a central data pool, for which the IT department is responsible, was created based on the business requirements. It consists of ten broad data sources from different subject areas and served as a basis for the BI content that would be developed by the specialist department. The emergence of individual data processing was prevented to a large extent by central metadata definitions. Drawing on the requirements of strategic planning and marketing, the central data pool was iteratively expanded according to the needs of other specialist departments.


    The descriptions of the provided data sources were compiled in a central catalogue to give analysts an overview of existing data sources, their origin and the data structures they contain. Data sources provided by IT were marked separately to ensure that they are distinguishable from data sources created in relation to the Self-Service. Analysts can thus choose to give preference to the data created by IT, thus ensuring clear accountability. If self-created data sources develop into popular and important building blocks, there is a plan for IT to take these over and stabilise them in order to keep Self -Service Reporting scalable and stable.


    A coordinated SSBI role concept was established for the department and IT, which authorises functionality and data according to duties and responsibilities. The departments began to create the content in close cooperation with IT. This enabled the analysts, who had already been intensively trained in the implemented tool, to further deepen their knowledge of the tool, thus ensuring the semantically correct use of the centrally provided data sources.


    Based on the close cooperation with IT, the created content was the direct responsibility of the department. The following rules were established in this regard: For new analytical applications, the responsibility for accuracy lies with the department. If data sources provided by IT are used, IT only assumes responsibility for their availability and accuracy with regard to the interface requirements. The department assigns access to applications and data sources.


    Analysts exchange information in their own community at regular intervals. They discuss current problems and ideas for new analyses in these meetings. Problems can often be solved with the experience of another analyst and, when discussing future and currently developing applications and analyses, synergies can be used and knowledge can be pooled.


It can be said that establishing a cloud-based SSBI approach has significantly facilitated the accessibility of existing data and has thus significantly improved working and interacting with it. As a result, a data-driven culture has developed in the departments that also relieves the burden on the central IT department.



Thanks to the major software manufacturers' continuing development of cloud analytics products with regard to the choice of tools, especially in the public cloud, companies have new tools at their disposal whose characteristics make them suitable for use in an SSBI context.
Modern BI tools put more and more functionality at the departments' disposal; these include, on the one hand, prediction and forecasting from the world of "augmented BI" and, on the other, functions that allow them to prepare and store data themselves. That said, far-reaching implementations of SSBI environments are still rather the exception at present. SSBI projects fail and are not pursued further if SSBI is not designed holistically, based on an environment that is managed by IT and is not adapted to the company's individual situation with regard to critical topics such as role concept, content and data lifecycle, governance and tool selection.

The introduction of an SSBI approach in the specialist departments offers measurable advantages for most companies. Added value is seen above all in the flexibility gained by the departments and the simultaneous easing of the burden on the IT department:

the departments create BI content by themselves at short notice, without regard to the availability of the IT department and, where relevant, having to rely on predefined release cycles. This means that any information needs which arise are covered very promptly and the resulting freedom in the IT department can be used to focus on data provision, for example.

One additional and significant advantage of an SSBI approach is the departments' inevitable interaction with the data generated in the context of the processes for which they are responsible. The framework of an SSBI approach guarantees early involvement of the department in the definition of management-relevant KPIs and the conception of the BI application, which has proven to be one of the most critical success factors in many BI projects. It is no longer necessary to build deep process knowledge in the IT department. This can reduce the effort required for projects and save on costs.

As a consequence, the right SSBI approach not only creates flexibility and cost savings, but also leads to a general increase in analysis quality, which in turn has positive effects on the data-driven management of digital processes and thus the overall development of the company.



However, in addition to the positive effects described above, introducing an SSBI approach is also replete with various risks. Experience has shown that these things in particular are worth pointing out:

    The amount of BI content grows rapidly due to the increased number of content creators in the departments. Large portions of the created content are working versions and are not finalised. It is no longer possible to recognise which content is correct and which content can or should be used. Also the possibility and responsibility as regards deleting content is not clear. This situation makes the ongoing development of the underlying data management difficult or no longer possible.


    Content creators develop their own, non-coordinated derivations of management-relevant KPIs. Comparisons lead to deviations from supposedly identical technical content. This then requires cost intensive re-engineering to determine the existing deviations and the "truth".


    The created content is exclusively geared towards specialist content and is not optimally developed with regard to high-performance processing. This causes the load on the underlying reporting systems to increase significantly.


    These concern on the one hand the unauthorised opening, editing and use of content, for example between competing departments, as well as the display unauthorised data.


    Not every user in the department is immediately suited to the creation of BI content. A lack of knowledge regarding the tools used and the company's internal data architecture can cause misinterpretations of the used KPIs and incorrect process management decisions.

    The aforementioned risks can be mitigated with a suitable combination of organisation and tools. If, in particular, underlying organisational conditions are not effectively defined, the SSBI project is destined to fail. Most notably, the properties of cloud analytics can be used to minimise risk here.



The characteristics of cloud analytics tools make them very suitable for use in an SSBI approach and help to minimise the risks described above. The following characteristics should be emphasised in this regard:

    Unlike on-premises installations, cloud analytics tools eliminate the need to purchase hardware and install/configure the used tools. The time it takes for the tools to be available to the department and the associated possibility to obtain information is significantly reduced. In addition, the risk of bad investments is minimised, since costs can be reduced to zero in the short term by not renewing with the cloud provider


    In contrast to rigid licensing models, flexible pay-per-use models allow costs to be determined by actual usage behaviour. The departments can directly influence the costs incurred via their usage behaviour.


    It is not necessary to perform sizing before installation. Cloud usage allows one to flexibly react to increases in resource requirements, both in terms of memory and CPU.
    This allows easy integration of additional departments and content creators into the SSBI approach. In addition, it is possible to react to individual requirements of individual departments; for example, if resources are only required at certain times, such as during a month-end closing.


    Cloud analytics tools are largely designed for direct use by the department. Functionalities that are typical for the IT department such as access rights administration can be hidden from the department. Operation is usually possible via the browser. Local installation and malfunctions caused by incompatible release statuses are eliminated.


    Cloud analytics products offer a central storage facility for created content. This can be organised according to the individual governance. The cloud provider ensures guaranteed availability of the content. The department and IT do not have to implement backup concepts.


    Cloud analytics products provide collaboration options for content creators out-of-the-box, for example by assigning tasks, sharing and reusing content or comment functionality.

    In addition to the numerous advantages provided by the cloud, its disadvantages must not be ignored. In particular, it must be decided whether the relevant data may leave the company network or must remain on premise.



With an increasing focus on cloud-first strategies, both on the part of software vendors and their customers, the market for public cloud and software-as-a-service solutions is clearly gaining traction. This is equally true of software solutions for a company's core systems (ERP, CRM, SCM) and the area of analytics (BI, reporting).
With cloud analytics solutions, data is stored and processed in the cloud. The services of public cloud providers such as Microsoft Azure, Google Cloud

Services or Amazon AWS are predominately used as the basis for the providers' individually provided services. The private cloud, where cloud resources are predominantly used by one organisation, is less represented in the context of cloud analytics.

By using public cloud services and software-as-a-service, specialist departments can fully benefit from the added value mentioned above. Cloud analytics has developed out of being a pure visualisation solution for existing data to become a holistic data platform that also fulfils upstream, accompanying and downstream tasks and requirements. Data integration functionalities allow data in the public cloud to be extracted, transformed, harmonised and stored. A wide range of connectors offers possibilities for connecting to the existing system landscape and third-party content providers, which can be seamlessly integrated into analyses.

Data governance functionalities make it possible for the IT department to authorise and release content created within the scope of self-service and to accompany and control the process. Since company-wide data lakes and data warehouses often have large amounts of data, data can be connected via live connections. This can reduce the use of cloud resources, which are often billed on a pay-per-use basis. In the case of smaller or weaker source systems, which continue to form an important part of the system landscape, for example in the context of legacy applications, replication of the data in the cloud analytics application can reduce the workload while at the same time speeding it up. The formation of data silos is prevented in both cases, as the source data can be automatically updated in analytics tools in the event of a change.

Increasingly, therefore, the cloud analytics application is no longer establishing itself as data management's "single point of truth", but is developing into a "single point of entry" via which a heterogeneous system landscape can be centrally connected and analysed.

A wide range of products is available to establishment cloud-based self-service environments in the company. The use of a tool and choice of vendor should always be evaluated individually. To provide a selection, we should mention the currently available and proven public cloud solutions on the market here: Tableau Online, Microsoft Power BI, SAP Analytics Cloud or Qlik Sense.