What about Business Intelligence?

Business Intelligence: a seven-headed beast? Really?

Let’s talk about Business Intelligence and the basics concepts to comprehend to be ready to start in the BI world. The next lines do not have any technical and practical guidance, but actually a minimalist conceptual presentation recommends before download and install a BI tool for your first project.

Don’t worry! It will be short, not painful and I hope, very useful to open your mind regards this topic.

What a hell is Business Intelligence?

Looking for some historical documentation, it was possible to find millions of different ways to define Business Intelligence, and due to that, I picked up someones to make you, if you are not, more introduced to that terminology. Check this out:

Generic term introduced by Howard Dresner in 1989. Process of collecting, analyzing, and distributing data to improve business decisions.

Computer World Magazine

Process of transforming data into information and through discovery, transforming information into knowledge. A comprehensive term that includes applications, infrastructure, tools, and best practices that allow access and analysis of information to improve and optimize decisions and performance.

Gartner Group

The ability to collect and react accordingly based on the information retrieved.

Richard Millar Devens

The ability to apprehend the interrelationships of presented facts in such a way as to guide action towards a desired goal.

Hans Peter Luhn

A set of methodologies, processes, architectures, and technologies that transform raw data into meaningful and useful information used to enable more effective strategic, tactical, and operational insights and decision-making.

Forrester Research

We could spend many lines describing Business Intelligence, but I believe after these several definitions you should be a little more familiarized with it. The high important point is to understand that Business Intelligence does not should be directly and exclusively related to technology tools. Of course, the use of is has really significant importance for the process but is not enough if the concepts and methodologies are not properly practiced. It’s possible to implement business intelligence practices without having a computer but is not possible to implement a business intelligence process without following the proper practices and methodologies. It’s necessary at least to do the basic parts of the process, even inconsistently.

---

Business Intelligence in 4 steps

I’m possible to understand the BI importance in a short 4 steps loop sequence:

These are the basic principles to understand the difference between data, information, and knowledge. From it, we can describe a little more data process, as for example one of the most used ones: ETL.

Ok, and ETL is…?

ETL is an abbreviation for extraction, transformation, and load, one of the frequently and mandatory processes in many kinds of different Business Intelligence contexts, as Data Warehouses, Data Lakes, and even in some BI self-service tools. Always that is possible, I like to present the ETL process to my students showing them a small German video that explains in a funny and practical way what is that process in the end of the day. Calm down! You don’t have to speak German to understand it, just comprehend the images should be enough.

How would you put some SAP data, e-mails contents, excel, and text files, all together in the same database? Would that be possible? Yes, that is possible. Sometimes not easy, but possible and ETL has huge importance in that process for sure.

Extraction is how to bring some raw from someplace to your project workplace, even if it’s a Power BI, an advanced Data Warehouse, or simply an excel file. The point is the answer question as:

• Which data will be extracted?
• Where are my data sources?
• How can I connect to them?

Transformation is the second step of that process, where a big part of the logic is implemented. Can contain business requirements and technical requirements and should be always them possible to be very organized and enough documented. For common software developers, its the place where we “feel free” to implement codes (when possible) with more flexibility, playing with the data the much we need. That step will handle questions as:

• Are my data clear? Is there some error or historical thing that should be corrected to avoid mistakes?
• Are all my data in the same format? Do all data sources have the date field with the same mask?
• Should I do some kind of calculation during that process? Which kind of? Business rules or technical rules?

The load is the last step of that process and highly related to the setup of the amount of data. It’s important to defined the right size of packages, restrictions of loading, storage methods, etc. That step will answer questions as:

• Where that data will be stored? A cube, a simples table?
• Will that data be separated into different objects?
• When and how frequently will that be extracted?
• Do I have to filter my data?
• Should I load it completely or just the delta changes?
• Do the existing past data have to be reloaded in the future due to possible changes?