OBIEE Bubble chart

First of all from a visual design stand point they are horible! Why? You are comparing 5 sources of data on a 2D surface. Most people already have problems interpreting a 3D graph. Anyway let's give it a try:

Let's start with a basic report:

It gives two dimensions (Brand & Market), two measures (Revenue and Revenue Yago), there comparision (Revenue Yago / Revenu)

The OBIEE buble chat has 5 elements:

: X-axis

: Y-axis

: Dimension 1

: Dimension 2 (optional)

: Comparison value

This would give us:

Resulting in this chart:

On a personal note I find this a horrible chart....

If your really need this much info at once try splitting it into two graphs:

Till Next Time

OBIEE Dashboard Design 001

As promised in a previous article (http://obiee101.blogspot.com/2009/04/obiee-final-skills-creating-great.html) I will show you some tips and tricks on how the implement the findings of Stephen Few in OBIEE.

Let's start with a simple top 7 list of product revenue:

If we go to the the default chart we get:

Step 1 is to convert this to a horizontal barchart:

Step 2 convert to 2d rectangle effect:

Step 3: Remove the grid lines:

Step 4: Add a title and data labels:

Step 5: Set the axis title and labels:

Step 6: Remove the "chart box" by setting the Major Gridline colour the same as the background.

Step 7: Set the bar colour:

Let's compare both:

Still you have to consider if a simple list:

  doesn't give the user enough information, charts tend to take up more dashboard space and space is 'expensive'!.

Till Next Time

OBIEE The final skills, Creating great dashboards.

Most project we spend 70% on getting the data (ETL), 20% on creating the repository and 10% percent on creating the dashboards. Regarding step 1 and 2 most of us have enough skills do them successfully. The final step, creating the dashboard, is something we often take to light hearted.

Let's make a bold statement here: 60% percent of the dashboards we create, within six months, aren't used in the way we intended them or even worse aren't used at all.

Why? Most of them don't communicate to user in the most effective way. The user can't find what he is looking for at glance. Either we tend to overload him with data or present it in a non effective way.

If you want to master these last skills you should really read :

Information Dashboard Design: The Effective Visual Communication of Data by Stephen Few. (http://www.amazon.com/Information-Dashboard-Design-Effective-Communication/dp/0596100167)

The current version of OBIEE has the (dis)advantage that it doesn't have many charting options. This means we have put in some extra toughts in the way we put the data on the dashboards.

In comming period I will post some examples how the findings of Stephen Few can be implemented within OBIEE.

Till Next Time