Business Intelligence (BI) Software, BI Tools

Software Tips

Definition of Business Intelligence (BI):

Business Intelligence (BI) refers to the process of collecting, consolidating, and analyzing raw data from various sources across an organization within a central database. BI systems transform this data into meaningful insights through preparation, analysis, and reporting. Their key value lies in enabling businesses to uncover relationships and patterns across different datasets, helping decision‑makers derive well‑founded strategic and operational conclusions.

Business Intelligence Structure

Business Intelligence Structure

The basic building block of the Business Intelligence process is raw data (data sources), which is generated both inside and outside the company. Data generated within the company includes production data, sales figures, and all data from Supply Chain Management (SCM), Electronic Procurement (E-Proc.), Enterprise Resource Planning (ERP), Customer Relationship Management (CRM), etc.

Data generated outside the company includes, for example, user data from the internet, statistical data from business associations, etc. Based on this data, data extraction (Extract-Transform-Load), transformation, and loading of this data into the data warehouse are carried out. This data is then analyzed using data mining and OLAP.

These analyses (data analysis) are used to create reports, which in turn form the basis for managers' decisions (decision making).

Terms related to Business Intelligence:

Source systems:

A distinction is made between internal and external data sources. Internal data sources primarily refer to upstream operational systems such as ERP systems, inventory management systems, or CRM systems that support the execution of business processes. The data generated here is mostly transaction-oriented and has the level of detail that is typically necessary for the company’s operational activities. External data sources may include, for example, online databases for stock prices, currencies, geographic information systems, DVDs, or address databases.

ETL (Extraction, Transformation, Load):

ETL is a process in which company data is extracted (extraction) from various systems and transformed (transformation) into a predefined format and into the structure of the database (in this case, the data warehouse) and then loaded (load). In complete BI solutions, this is part of the BI platform used. This process runs automatically for company-relevant data. During this so-called ETL process (extraction, transformation, and loading), the data from the different systems is standardized, and particular attention should be paid to data quality.

ETL Process

There are so-called connectors for extracting the data and, if necessary, for correction. These connectors are used to connect the individual systems (ERP, CRM, etc.) to the ETL tool. The connectors are normally provided by the ETL tools and support common formats and systems such as relational databases, XML formats, CSV files, etc.

Data Warehouse (DWH):

A data supermarket or database in which data from different business areas is stored in a predefined, standardized format. Data is stored permanently in the data warehouse. For security and performance reasons, production data is not accessed directly; instead, redundant storage is used. The data warehouse is the fundamental basis of Business Intelligence software. A company may very well have several such data warehouses. Depending on the size of the databases, the data volume is sometimes divided by subject area, department, or group. These subsets are called data marts.

Data Warehouse (DWH)

Relational databases are generally used for a data warehouse. The data is often stored in a star schema. In a star schema, two different types of tables can be distinguished. On the one hand, there is the fact table, which is used to store key figures such as revenue or costs, and on the other hand, the dimension tables.

The dimension tables contain the business perspective on the data. This makes it possible, for example, to analyze revenue by product, region, and time. Dimensions make it possible to obtain a multidimensional view of the existing facts.

The consistency of the data between day-to-day operations and the resulting analyses and reporting is ensured by the data warehouse. A company may very well have several such data warehouses. Depending on the size of the databases, the data volume is sometimes divided by subject area, department, or group. These subsets are called data marts.

OLAP (Online Analytical Processing):

OLAP Cubes

OLAP is often referred to as BI in the narrower sense. It is a data processing process in which different dimensions of data are accessed. People often speak of OLAP cubes. The mathematical dimensions of the OLAP cube describe the data. Data can be selected and viewed flexibly via one or more axes of the cube.

There are specific OLAP operators for querying the underlying data. These allow navigation (drill-down, drill-up, etc.), selection, and rotation of the data. When rotating, key figures can be viewed from different perspectives.

Drill-Down or Drill-Up:

There are specific OLAP operators for querying the underlying data. These allow navigation (drill-down, drill-up, etc.), selection, and rotation of the data. When rotating, key figures can be viewed from different perspectives. These are filter operations in hierarchical structures in which results are viewed either at the next lower level (drill-down) or at the next higher level (drill-up).

Data Mining:

Data mining is comprehensive data analysis for identifying trends or patterns in large volumes of data. For example, relationships within data sets can be identified. The grouping of clusters (objects), their classification, and association analysis are key tasks of data mining. The individual evaluation methods are selected depending on the task. In some cases, a combination of different analysis methods may also be useful depending on the question.

Web Mining:

A sub-discipline of data mining that is limited to internet content, including user behavior and the relationships between websites. The aim of web mining is also to gain valuable information from unstructured data (e.g. industry news, press releases, social media channels) by means of analysis.

Dashboard:

A dashboard is comparable to a cockpit. It displays all important information in one overview and shows a company’s key figures in graphical form. Visualizations such as traffic lights or gauge charts are often used. An important feature of the dashboard is the ability to display data from a wide variety of sources in real time.

Ad hoc reporting:

In real time, a user can display the information that is relevant to them (ad hoc reporting or self-service analytics). Filter settings allow access to company data down to the smallest detail. In this way, the user is enabled to get answers to their own questions regarding company data without burdening the IT department with this task. This ensures that important information reaches the right people at the right time.

Data Profiling:

As a rule, data profiling is an automated analysis of substantial properties of data sets. This automated analysis is a continuously repeated (iterative) process in which the data is first integrated, then analyzed, the results are presented, and subsequently evaluated from a business perspective. Data profiling tools are used, for example, in compliance initiatives. Among other things, they are used to document and track data quality problems.

Additional specific criteria and question prompts for evaluating Business Intelligence (BI) software

• If your organisation has a wide range of users with diverse analytical needs – from business users and department heads to specialist analysts – the software should offer comprehensive analytical functions, as well as role-based configurability and customisation. Be sure to enquire about all the options the software provides.
• Be sure to ask about all the scorecarding features the software has to offer. Scorecarding enables companies to capture their business strategy and communicate it at an operational level. This allows managers and decision-makers to set quantifiable targets and objectives. It also enables them to track performance across different business units, operational subsidiaries and geographical regions, so they can quickly identify areas requiring their attention.
• It also makes sense to look for dashboard features that allow every user to access content, use it interactively and personalise it, thereby supporting their individual decision-making process. Secure access to information across all time horizons (e.g. business intelligence, real-time monitoring) means that users can move from insight to action in the shortest possible time. Users benefit from the potential of the dashboard content available to them, which contains reliable information from the user’s perspective and in the format they require.
• So-called real-time monitoring provides reliable business intelligence, as all staff members who interact with customers are provided with a wide range of graphical operational KPIs and key performance indicators. This ensures that the information required for decision-making is available in real time. Through reliable business intelligence, staff members receive up-to-date data in the correct context, thereby gaining relevant operational insights that enable faster decision-making. Faster decision-making leads to measures that have an optimal and real-time impact on business operations, e.g. for service staff in call centres, in marketing lead monitoring and SLA monitoring.
• Functions that allow statistical analyses to be incorporated into key business reports are also very important and useful. Such analyses help to reduce the time spent on data analysis and the preparation of business presentations. Statistical data should be able to be integrated into reports. Make sure you explore all the possibilities offered by such functions and options.
• Explore the possibilities for secure, accurate planning, budgeting and demand forecasting through integrated and comprehensive enterprise planning capabilities. If a company wants to achieve optimal business results, it must plan effectively and be able to make forecasts. To do this, the right information must be provided to the right people in the desired format. Furthermore, any changes must be able to be implemented in the shortest possible time. For most companies, regardless of size or sector, company-wide budgeting, demand forecasting and reporting present a huge challenge. Finance professionals and department heads alike usually describe annual planning and budgeting as tedious and time-consuming. Therefore, these support options and the associated functions should be investigated and examined in detail.
• Ensure that the software archives and manages the information you have collected internally and the data you have recorded in a centralised, secure and efficient manner. Ideally, this should be done in a database (e.g. in a data warehouse, data centre or database server).
• It is very important that the software supports the processing and analysis of data, presenting it as reports, in the form of key performance indicators or plans, particularly in graphical and tabular formats.
• Please ensure that, when analysing data, the software takes into account not only your own data but also external data from competitors and data from research or advertising (market monitoring, trends).
• It is essential to bear in mind that data backup is supported, or that there is a mechanism in place for organised data backup. In this regard, it is important, amongst other things, that a roles and permissions framework determines who can access the data, when, how, and in what form.
• Therefore, when planning your company-wide communication, consider which groups of people or organisations should or could be involved in the distribution or dissemination of your findings, and in what form – for example, decision-makers, staff (internal, external, freelance), branches or subcontractors.
• You should also consider whether the solution is capable of generating forecasts.
• Please ensure that the software supports data mining as well as other or additional analytical methods (e.g. OLAP), and that it generally provides comprehensive support for systematic analysis.
• Please ensure that the solution offers import and export interfaces with major software providers (e.g. Microsoft, Oracle, Drupal or Salesforce).
• Ask about multi-user or multi-client capability. Software is considered multi-client capable if it allows multiple clients to be managed or supports separate user accounts. This is important so that, for example, tax advisors can manage accounting data for multiple clients, or so that different staff members within a company can only access the data relevant to their respective areas.