The Curious Case of Advanced Analytics
The 2008 David Fincher movie ‘The Curious Case of Benjamin Button’, had a unique plot wherein a man born mature, ages backwards only to become an infant and leave the world. Interestingly, I can somehow relate this story to that of analytics, the way it came into the world, its relevance now and how it might become inconsequential soon (unless of course there’s a change in the way we look at it).
Undeniably, ‘Analytics’ is the most used term ever since it was first spoken about a few years ago. Every industry and every sector was impressed with what analytics could do for them, every vendor was selling some kind of analytics and it would not be wrong to say that it was a born hero! However, if you look at what value analytics provided to organizations a few years ago, it is evident that it has diminished substantially, and will continue to do so if the approach remains the same. This diminishing value of analytics with increasing maturity is what drives me to equate its story with that of Benjamin Button.
Let us say, 7 years ago, you had 2 TB of data that was generated each day on an average. When someone sold analytics to you, their value proposition would be to enable you to filter out irrelevant data and guide your focus towards a small portion of the overall data. This small portion of the data that you had to work on would be of much smaller size, perhaps say 10 MB.
Over the years, many things in this equation have changed, but sadly, not everything has. The amount of data generated has grown exponentially, giving rise to terms like ‘Data Tsunami’, and the trend continues as we read this. However, if you examine the other part of the equation, the fundamental value proposition of analytics solutions has not changed as drastically. So, your organization’s scenario of looking at 10MB of analytics data obtained from 2TB of accrued data has now changed to looking at 2TB of analytics data obtained from 2 Petabytes of accrued data! In other words, the amount of filtered information you have today is larger than the amount of total data you had a few years ago.
So, does this mean that analytics is now useless? Does this mean that we humans should give up on data-driven decision making, and rely on intuition because there’s so much data to look at? Or should we deploy robots that decide for us? Perhaps, neither of the above. What we need is a shift in paradigm, and a drastic one.
Analytics today gives you numbers. Numbers that by themselves do not really lead you anywhere unless influenced by an external factor – ‘Business Context’. Yes, business context is what adds value and meaning to numbers. Consider a scenario where a telecom operator is looking at details of call traffic between different regions. Analytics will typically summarize and tell you the average duration of calls between destinations, most busy terminating destinations, the revenues you are earning from various circles, etc. While all of these are useful information, they do not lead to any direct action. When you read it, you have to work a lot further to figure out what your course of action is. This is because, the solution that gave you this analytics information itself does not know what the use of it could be.
Now, imagine the same system telling you what rate plans to offer to which circle and also predicting the percentage of revenue increase if you do so. Would be useful, right? Well, of course yes, but how can this be done? The trick is to infuse a sense of business context into analytics with the help of domain experts. By doing so, you’ll no longer be looking at numbers, charts, graphs and figures, but rather be ready to action upon the intelligence that you have obtained.
If only there was a similar solution for Benjamin Button, which would help him overcome the mysterious problem and live on normally! But then again, if that was the case, who knows how many people would have liked the story, and whether or not it would have been able to make over USD 300 million at the B.O. :)