Also called Decision Support Systems (DSS), these systems use a series of concepts, techniques, and tools to transform data into useful information to assist in decision-making. To develop this type of system, many professionals are typically involved, mainly in the areas of Information and Communication Technology (ICT) and Business, and specifically in the fields of Business Intelligence (BI), Business Analytics (BA), Data Mining, Machine Learning, Statistics, Databases, Software Infrastructure, Competitive Intelligence, Marketing, Engineering, and Knowledge Management, among others. The interaction of different professional profiles ensures, in addition to performance and security, that the key points for the institution (Key Performance Indicators - KPIs) are defined by the departments with the appropriate knowledge, maximizing the success of the projects.
Analytical Cube
With all Key Performance Indicators (KPIs) defined, all business rules documented, and the origin of all data known, the development of the DSS begins. The first step consists of reading data from all sources defined in the project, integrating them into a large data repository, the Data Warehouse (DW).
From there comes the great masterstroke of Business Intelligence: Analytical Cubes, also called Multidimensional Cubes, are created, one for each fact to be analyzed and with no limit on the number of Dimensions.
A single cube must be able to provide, with great agility, all the information a manager needs about the fact that the cube analyzes. #All.
The example on the side is a graphical representation (gif) of a Financial Transactions cube composed of 3 dimensions:
- Type of Financial Transaction (what);
- Date (when) and
- Location (where)
Each of the dimensions can be used to filter the data and select only the desired period. The portion in red corresponds to the selection made, and in the last frame of the gif, we can see the square that indicates the intersection of all filters, that is, the desired answer.
Filters applied: Transaction type = Sales, Date = Year 2017, Location = State of São Paulo.
Interactive Dashboards
Now that the analytical cube is capable of providing all the necessary answers to the manager, the next step is to create interactive dashboards that allow users themselves to refine their queries, simply by selecting, for example, the Year and Month, always with the possibility of filtering the data even further by clicking on the charts (bar, pie, etc.) and obtaining precise and practically automatic answers, as all the techniques and tools used allow this data to be pre-computed and easily accessible. Trends and deviations from the pattern in the institution's data are easily found, visually, in a clear and intuitive way, as shown in some examples below:
[widgetkit id=66]
Toyota Case Study
In the late 90s, the company faced major problems in its operations chain. The cost of storing cars rose, and it was no longer able to deliver products to its customers on time. It used computers that generated a huge amount of data and reports that were not used strategically because they were not always accurate and were often provided too late—which delayed decision-making.
A new CEO was hired. She identified some solutions: first, the need for a Data Warehouse—a centralized, organized, and easily accessible data repository. Second, the need to implement software tools to manipulate this data. Unfortunately, the newly implemented system did not work correctly: the historical data entry included years of undetected human errors, duplicate and inconsistent data, and a lack of important information. All of this led to hasty analyses and conclusions about the distributor's operations.
Only in 1999 did the company decide to implement a Business Intelligence platform. Within days, the system showed good results. For example, they discovered that the company was being double-billed for a special train shipment (an $800,000 error). Between 2001 and 2005, the volume of cars traded increased by 40%, and transit time was reduced by 5%. These and several other benefits helped Toyota achieve the highest profit margins in the automotive market since 2003, and they have been consistently increasing every year since. Furthermore, a study conducted by IDC Inc. in 2011 indicated that the institution achieved, in that year, a return of at least 506% on its BI investment. Source
This is just one of countless cases that illustrate the efficiency of these solutions capable of integrating and interpreting Data, transforming it, in some way, into Information relevant to your business, enabling, with proper analysis, the creation of Knowledge. Through the use and management of this knowledge, Intelligence is born.
Conclusion
More than 15 years have passed since Toyota adopted BI. Currently, more than 2.5 quintillion bytes of data are generated daily worldwide, with 90% of what exists today having been created in the last 2 years. The trend is for this number to grow at an increasingly rapid pace.
A Decision Support Solution does not need to cost millions of dollars or require the bundled sale of equipment, training, consulting, and services. On the contrary, with open-source software and the backing of a specialized consultancy, excellent and sustainable results can be achieved using existing equipment, with any operating system on the market, and accessible from cell phones and computers to large TV screens. All of this with the complete security and freedom of having full control over the source code.
And you, what are you going to do with your data?