Tuesday, November 08, 2011

Seven Billionth Baby and Corporate Dashboards!

Few days ago, Oct 31st to be precise, the news of seven billionth baby was creating lot of buzz in the media.

While I was not keen to understand the census, I was really curious to know the math behind this count. I checked with my ex-colleague and Datawarehousing/Business Intelligence Specialist Paramesh  on how this could have been possible? There is every possibility that this number could be inaccuarate because of the sheer size of the canvas. In this case, the canvas is the whole world.

We all know the practical challenges in aggregating data within a company?. And How do they do it across the globe?

Thanks to Paramesh, Today he reverted with an exact article that talks about the same challenge. It was a surprising coincidence! The article talks about - How did they count? How did they comeup with that number - 7 Billion - on that specific day?

The fact of the matter is United Nations don't know. The announcement was meant to be symbolic than to be precise. UN has confirmed there is no way it can predict at which minute of the day - 31st Oct - the 7 billionth baby was born in the world.

Howere, it is not a random number dropped just out of the hat. There is a process/method that U.N. has followed to comeup with that projection. United Nations collect country wise population data every five years and based on the system generated models, the date has been picked up at Oct 31. There is a 2% margin of error expected in the global tally, which in this case, could be huge in terms of census and the time period!

I see the corporate dashboards(BI/Reports) that take data from individual line-of-businesses within a company can take a lesson or two from this news. I have tried to list few of them:

1. There is a process and system in which the sub-system (countries) provide data to the main system (United Nations). So, Its not just about technology. We will need a process and method to coordinate the data aggregation. The systems/applications and runtime infrastructure have to implement the processes and contracts to ensure aggregation.
2. People are aware of the percentage of error and at peace with Good Enough BI. In lighter vein, What you sow is what you reap!. :-) Awareness on Data Quality and Incremental expectations are the key to successful BI implementations!
3. The focus is more on the ways in which data can be leveraged for analysis/insights and less on the precision. Often People debate about the trees that they forget to see the forests!. I have seen executive dashboards often spark controversies around numbers that the real issues get pushed to back seat!.
4. There is a model in place to predict the growth. Yes, finally BI solutions should not endup as concrete/one-off implementations. The definition of formal models that abstract the concepts in respective verticals hold the key for extensibility of the solutions.

Finally, As any BI implementation would invite disputes, Here is somebody in Asia pacific disputing that the whole seven billionth baby announcement from UN was stage managed!. :-). Again, its not just about numbers and technology! :-)

No comments: