Data Visualization

We have covered Business Intelligence, Business Analytics, Data Modelling and Data Cardinality, it's the right time to discuss Data Visualization.

Data Visualization is what is seen by End Users. The Success or Failure of the project is determined based on it. As the phrase goes, 'If you can't draw it, you don't Know it.' And it’s very true. One can spend millions of dollars, can implement a perfect Data Strategy, can have perfect EDW or Data Lake etc. etc. but after all the hard work if the End User can't see what is implementation, if the End User is unable to make use of the data for DSS, if End User will have to go back to their Finance or Technical departments for manual data extraction then all the efforts put-in is nothing more than Waste of Time and Dollars.

Now, there are rules to have good Data Visualization which is relevant to any visualization tools in the market like Tableau, Qlik, Power BI, OBIEE, Cognos and many more.

• Rule#1: First identify all the relevant KPI(s) to be presented on the same page.
• Rule#2: Don't have more than 5 KPI(s) on one page.
• Rule#3: User must know the Cardinality of all the tables which he/ she is going to join to eliminate the chance of Cartesian data [this is very important].
• Rule#4: Use 1 tabular section on one page. Max can go up to 2.
• Rule#5: Never enrich or prepare data in a Data Visualization tool as doing so will take a lot of memory. Better to let the memory be used by slice & dice or drill in/out tasks. If you need to prepare data like to have column#3 by multiplying column#1 to column#2, do it at the datastore level. Yes, you will end up using additional storage but don't forget we are targeting faster Data Visualization rather aiming for saving storage.

The above are not hard rules and can be optimized based on the infrastructure in place. Plus, these are a few of the many rules to create an effective Dashboard or Report.

Another big mistake that organizations are making is to have their Data Analyst do BI Specialist roles as well. These are different roles and need different kinds of Data Visualization tools.

As there are many visualization tools in the market, it is very confusing which is for Data Analysts, and which is for BI Specialists. Again, this is just my assessment of which one can agree or disagree. To name a few, for me Tableau, Qlik and Power BI are data analysis tools for Data Analysts whereas Cognos, OBIEE and SSRS are for BI Specialists. Having said that there are no doubts, data analysis tools are enhancing the features to be called a BI tool but still in my review there is a long way to go.