istock.com/Bojan89

BI (Business Intelligence) is intended to give employees in the company comfortable access to current and decision-relevant information. Unfortunately, conventional BI solutions do not always fully meet the requirements. This article deals with the five biggest problems that users have with conventional systems.

Traditional business intelligence software is increasingly reaching its limits

In the past two decades, business intelligence has found its way into numerous companies. The primary objective was to make data from upstream instances such as ERP systems available for analysis. This relieved the previous systems and brought numerous improvements to the reporting. However, the classic tools do not only have strengths. If we look at the daily practice of users, there are five problems that occur particularly frequently:

1. Lack of self-service

2. Limited visualization options

3. Poor performance

4. External data sources are difficult to connect (lack of interoperability)

5. Lack of up-to-date data (no real-time information)

Let us consider these aspects in more detail below.

1. Lack of self-service

In classic BI scenarios, companies initially define a series of reports that are created by IT experts and included in the system – for example for the controlling department, sales, production or purchasing. With these BI reports, users are usually able to answer a large part of their business questions.

However, there are always cases in which either completely new information is required or existing reports would have to be modified in order to implement a solid decision-making basis. With traditional BI systems, users cannot do this themselves. Rather, they have to start an often cumbersome process, which can include the following steps:

– Define requirements, create technical concept

– Order implementation from IT

– Creation of the database and creation of the report by the IT department

– Test of the new report, troubleshooting if necessary

– Go live

In large organizations, this type of process can take several weeks or even months. However, at times when prompt responses to market changes are required, this is no longer acceptable. Rather, the goal must be that users can easily and quickly compile their own reports.

2. Limited visualization options

Conventional business intelligence software usually presents information and KPIs as “bare” numbers. A table with columns and rows is common. However, users want the information to be aggregated and visualized as clearly as possible. It is therefore not uncommon to export the data from the reports so that they can then be put in a clear form. This is done, for example, using Excel and corresponding diagrams, some of which are even converted into presentations. It is obvious that this procedure is time-consuming and error-prone.

3. Poor performance

Depending on the underlying database technology, the runtime of BI reports can be considerable. Both the call of reports and the internal report navigation (“breakdown”) can be extremely lengthy from a certain amount of data. Sometimes there is even a runtime-related termination if the system can no longer handle the available data volume. Especially in times of increasing amounts of data (keyword “big data”) this is of course suboptimal.

4. External data sources are difficult to connect (lack of interoperability)

Business intelligence is often based on just one data source: the upstream ERP system. In the age of big data, however, many users wish to be able to enrich the internal business data with information from other sources. CRM systems, online shops or external market data are just a few examples. However, traditional solutions do not provide for the integration of data with a different structure into the reports. From today’s perspective, their possible uses must therefore be described as severely restricted.

5. Lack of up-to-date data (no real-time information)

A significant pain point of users is also the lack of topicality of the information. Because classic business intelligence is always based on historical data. The database is often only updated at longer intervals (e.g. monthly) in order to conserve the computing resources. For users, this means that their decisions are based solely on past values. A short-term reaction to changes is not possible because there is no real-time information.

Conclusion: It’s time for new approaches

The problems mentioned make it clear that conventional business intelligence should now be critically questioned. It is still sufficient to evaluate classic past-based key figures. However, with the increasing dynamism of the markets and the increasing importance of big data analysis, it is reaching its limits. It is therefore worth taking a look at innovative solutions such as real-time business monitoring.