The cobra effect occurs when an attempted solution to a problem actually makes the problem worse. I came across this term in a Freakanomics podcast by Stephen J. Dubner and Steven D. Levitt. The term is used to illustrate the causes of incorrect stimulation in economy and politics. There is also a 2001 book with the same title by Horst Siebert, a German economist and professor.
By the way, I am an avid listener and follower of Freaknomics. I am amazed at their ability to combine research, insights and the fine art of storytelling. It’s an invaluable resource for BI & analytics professionals as we help organizations build data-driven cultures.
The term cobra effect began at the time of British rule of colonial India. The British government was concerned about the number of venomous cobra snakes in Delhi. They offered a reward for every dead cobra that was brought to the government leaders. Initially this was successful as a large number of snakes were killed for the reward. Eventually, however, enterprising locals began to breed cobras for the income. When the government became aware of this, the reward program was scrapped, and the cobra breeders set all their cobras free. As a result, the wild cobra population only grew. The apparent solution for the problem made the situation worse.
The same problem has happened with rats in Hanoi, Vietnam and feral pigs in Fort Benning, Georgia. In each case, incentive seekers came up with creative ways to maximize their results – and pest populations grew. Incentives don’t always work out the way we expect them to.
So how does this relate to BI? Are there Cobras of BI?
In the world of BI, The EDW or Single Version of Truth with industry grade BI tools is treated as an incentive/solution to kill off viral and ungoverned spread mart reporting in Excel. Very often the use of Excel is associated with erroneous misreporting of metrics. However, it is essential as it affords analysts the ability to reason with data, leading to improved understanding of issues, opportunities and changes to deliver on business objectives.
We also know that a journey towards to single version of truth presents two significant challenges; time and keeping up with ever evolving business needs. More often than not, analysts continue to create spread marts, even with the Single Version of Truth in place, compromising the core objectives of building one in the first place. Over a period of time these spread marts continue to grow, giving rise to the Cobra Effect.
In this two-part blog series, my aim is to table strong argument, explanation and a strategy to address this Cobra Effect in BI with the use of PowerPivot and Power BI. Next week we will take a closer look at the Cobras of BI, the failed incentive, new breed of cobras and the incentive that finally works!
Also if you happen to be in Dallas, you can attend our Power BI luncheon event this week. At this event my colleague Chaitanya Khaladar will walk through a real-world demonstration of Microsoft’s Power BI tool. Power BI provides business users and analysts with the freedom they need to reason with data while maintaining governance to eliminate virality of spread marts. In others words, making incentives work for business users, analysts, business leaders and IT.
See you next week for Part Two – The Cobras of BI.