As a BI consultant, I have already had insights into numerous companies. There has often been a discussion as to whether the creation of reports should be the responsibility of the business department or whether it should be handled by a centrally managed organizational unit.
In the following, I will present my point of view and give a recommendation.
Advantages of Self Service:
There are no traffic jams due to orders, users can create their own reports on their own. They also do not have to wait for the report to be delivered.
As a result, many users expect improved decision-making, since the user not only consumes the data, but also engages with it in terms of content. As a result, better and, above all, faster decisions can be made. At first, the argument seems absolutely comprehensible, because not only KPIs are considered. Instead, the end user can recognize interactions and dependencies much better if he is more involved with the data.
Another argument for this is the fact that the business department often already works with the data in operational systems anyway and has a good basic understanding of the data structure. Another popular argument is “employees can make more agile and confident decisions.” I think there is actually some potential here to make the end user more focused and more accountable.
It is also worth mentioning that the business department is independently able to connect additional data sources and enrich its own workbooks.
The resulting reduction in dependency on IT and the associated reduction in workload means that the IT department can once again focus more on providing actual insights. I find it more than astonishing why I could hardly find publicly accessible industry analyses in internal reports of any company. This data is often available on the web and can be used for internal benchmark analyses via web scraping. Due to a lack of IT know-how, departments can rarely provide such data sets.
In order to successfully implement a self-service BI strategy, however, it requires some change in the information culture of the company. If you are interested in this I can recommend this article.
Drawbacks of Self Service BI:
It is important to know that there are also disadvantages of such a strategy, which I would like to present now.
It is mandatory to create a generalistic data model that applies to all end users. This is often not so easy, because different departments have different requirements and perspectives on similar topics. Here, there must first be a harmonization of the key figures and the data model. Furthermore, an SSBI strategy must never lead to an ignorance of existing compliance rules. Therefore, the development of a role-based row-level-security approach is mandatory if multiple end users are to access the same datamart or cube.
Furthermore, there is the risk that the user misinterprets the data or that deviating logics and filters are applied when creating the report. This argument should not be neglected. For example, it can happen that people from 2 departments, each with their own created reports, talk about the same key figure and show different results. This decreases the acceptance and provides a considerable additional effort for correction. Let’s talk about additional effort: With this approach, the departments have to deal with the creation of the analysis on their own instead of doing the actual work. For a sales employee, for example, the creation of dashboards should therefore not be the main focus. If a country is active in several countries, it is not necessary for each country to create and use the same reports, but it is sufficient to create the reports once and make them available to each country. This allows resources to be used much more efficiently. A final disadvantage with the self-service BI approach is that report creators are not as proficient in dashboard creation and may use less than optimal visuals. This makes it harder to interpret the bottom line of the visual. It is not for nothing that the international business communcation standards approaches exist to describe which patterns visualizations should follow.
Focus on Standard BI
Let’s now take a look at why it might make sense to use a standard BI strategy. After all, it has worked well all these years.
First things first: When you are responsible for the data and the BI application, you have a chance of controlling the quality of the results.
This means that there is an approval process that ensures that all reports are presented in a uniform corporate design, the key figures are double-checked and quality-assured. In addition, the reports can be stored on a central repository. This prevents data in lists from being transported to the outside very easily. Furthermore data is properly prepared, stored, and secured. This allows us to avoid redundancies and conserve storage space.
BI Experts can build systems that offer standardized, scalable self-service reporting, and they can give users answers to their business questions. Of course, those answers may take a little time to materialize, especially if the IT department is busy with other projects. Or if the rate of growth of data (and big data) starts to outstrip the IT department’s resources for handling it
There are many benefits to traditional BI. In particular, successful analytics requires at least some key information to be reliable, consistent, and secured. This often-painful process is one of the key areas of success of traditional BI implementations.
Organizations with traditional BI tools operate in a tightly controlled environment where analysts or the IT team have access to the data. In this case, users must request data reports or dashboards and then first wait for analysts to provide them with the reports. This process causes congestion and delays in providing information intended for timely decision making.
I hope I could show you the difficulties and advantages of a self-service Business Intelligence implementation. In general, I welcome the introduction of a self-service BI approach. However, it is important to be ready for a cultural change within the company and to have a supporting BI specialist unit for questions. Without this, there will be an unmanageable backlog of reports and the project is doomed to fail.
