The next step for your understanding of analytics
Who has not experienced this in their working life: Suddenly and at very short notice, a report must be prepared for an important management meeting.
So how to put this together quickly now? Standard reports are centrally set up and operated by IT, meaning that special reports require manual action in Excel over night?
That pretty much illustrates our dilemma.
Traditional business intelligence and analytics systems lack agility. Business departments have to put up with long waiting times for reports from IT. The introduction and use of so-called Self-Service Business Intelligence (SSBI) relieves the burden from IT on several levels, accelerates decision paths and business processes, and thus ensures a faster reaction to possible market changes. Users can independently perform their own data analyses or data modeling to create key figures or new reports. Not surprisingly, SSBI is one of the megatrends that is already highly relevant in operational terms. In the last four years, SSBI has consistently ranked second or third in the BI Trend Monitor of the Business Application Research Center.
To understand the success of SSBI, one must first understand the depth of an independent analysis format. SSBI is not available as a one-format solution, but exists as a multi-layered model. It combines various different requirements and thus different user groups in one concept. With respect to these requirements, we can distinguish three levels:
Level #1: Users only consume ready-made reports and dashboards. They have no direct access to the underlying data or data models. We can consider their independence as relatively low.
Level #2: Here it becomes more complex. End users have access to analytical data models and do not only consume reports but also modify or create them from the ground up.
Level #3: This represents the top class of SSBI with the highest effort in the area of system support and independence of users. The users can integrate completely new data sources and thus operate in the backend of self-service systems. Here, a high duty of care and the associated knowledge of these users are required.
By fulfilling these requirements, SSBI offers numerous improvements for your business processes. The SSBI approach provides for the creation of reports directly in the specialist department. The report creators, who work in the department themselves, know exactly what key figures and what visualizations they need, e.g. for a successful sales evaluation. According to the idea of SSBI, controlling still has a central responsibility. It is still crucial to ensure that departments do not leave their reporting channels and use predefined data sources. Standard reports, which management accounting distributes on a regular basis, are still in vogue. Another important object and advantage in the context of SSBI are intuitive user interfaces, which support the creation of reports and facilitate data integration as part of a simple-to-use reporting environment.
So what are the major challenges from our experience?
1. Proper governance: Transparent, clearly communicated and binding guidelines are an important prerequisite for a sensible and low risk use of SSBI in the company. The company must legally secure and live data protection, responsible handling and the transfer of data. Anyone setting up SSBI tools without the appropriate governance creates exactly the problems that they wanted to avoid in the first place by using a BI solution. There is a danger of replacing Excel hell by a data discovery hell with almost the same disadvantages.
2. Clear rights and roles: A good SSBI strategy also includes role-based data access or even restricts data access to what is necessary through individual rules. Not every department or employee needs access to all available information from the company. This restriction not only helps to comply with data protection guidelines, but also protects employees from having to handle data that is inappropriate for their purposes, the analysis of which may lead to completely wrong conclusions.
3. Data integrity: Another point where SSBI can cause problems is data integrity. According to a survey by Harvard Business Review Analytic Services, almost two-thirds of respondents use more than five data sources, 18 percent use more than 15, and more than three-quarters believe that this number will increase in the future. SSBI has covered the topic of data integrity so far only marginally. The user is still responsible for assessing the correctness and completeness of the data included in the analysis. If there is no uniform view of the data, the quality and informative value of the analyses will suffer. Two employees can come to different, even contradictory conclusions simply because they had access to completely different data pools. A data-centered architecture and a holistic data management approach are therefore prerequisites for the meaningful use of SSBI. Without clean, complete and quickly available data sets, neither employees, processes nor new technologies can generate meaningful insights and make decisions.
Let us now proceed to the implementation of SSBI. In order to address the challenges mentioned above, we suggest considering the following five golden rules:
1. Interdisciplinarity: Real benefit from data comes only from putting use cases and user needs at the center of consideration. It is therefore crucial that BI projects follow an interdisciplinary approach and include representatives from all user areas.
2. Agility: While in the classic BI world a traditional, simple project approach based on the plan-build-run scheme was common, today it is more important than ever to be able to act quickly. In this context, the use of agile methods is proving effective. After the interdisciplinary collection of use cases, an iterative process starts to prioritize them and to roll them out successively. The most important application areas are thus quickly productive. In addition, it is important to test and to optimize reports based on an agile technological setup in a productive environment. This ensures a sustainable, value-adding solution. To establish a data-driven work culture, not only the participation of all subsequent user groups in the requirements analysis is important. Individual users must be included in the overall process, and goals as well as benefits must be clear: Why is data-driven work important? Where do we expect benefits for our company? How does working with data help you personally? The entire change management process needs a dedicated person in charge. This person should pursue the goal of designing and establishing new work processes together with users and anchor the data-driven mindset culturally within the organization.
3. Organization: With SSBI, data governance is still highly relevant. More than that, if a company does not consider governance when implementing self-service, anarchy will probably result. SSBI is therefore always a question of organization. There are four organizational principles that SSBI should build on:
Easy access to data for reporting and analysis
User-friendly BI and analysis tools
Simple and adaptable surfaces of the tools
Data warehouse technologies that are easy to deploy (e.g. cloud-based systems)
4. Concept: Data collection and integration is of course essential for any form of BI, no matter how modern or traditional. However, two further fundamental factors must also be taken into account: the concrete definition of the data model such as key figures, dimensions and the logics that systematically link them together, and ensuring long-term scalability. Due to the ever-increasing volume of data and in order to avoid a subsequent cost explosion, attention must be paid at an early stage to infrastructure, database technologies and query methods that are scalable in the long term.
5. Comfort: Today, SSBI must be more than a system from which data scientists and analysts extract data for business stakeholders upon request. The requirements are too diverse and operational, and the required speed to provide information is too high. It is therefore crucial that all stakeholder groups can access the data and data tools relevant to them in a self-service manner. Furthermore, real value creation from data can only succeed if working with data becomes a natural part of the daily work processes of all users. Simple handling is therefore a prerequisite for enabling true self-service. In addition, proactivity and automation are other important keywords: If users have to compile information themselves through complex analyses, they may quickly throw good intentions overboard.
Want to know more?
Do you want to take your analytics to the next level and learn how you can use it for your marketing and sales transformations? Get in contact with us.
Written by Marvin Jagals (University Duisburg-Essen) and Dominik Tappert (Digital Sales Lab).