Wednesday, February 25, 2015

Why Telcos need to care about Big Data Analytics?




Read more: BigData in Finance/Banking, BigData in Retail

Over the years telecom providers have primarily focused on delivering quality, cost effective medium for voice & data. This strategy to squeeze the most bandwidth from existing infrastructure helped profitability initially. However, competition has meant that each player has had to keep lowering the price bar, severely impacting the margins over the years. These are the telcos that provide the very medium for all of the data that is ushering in the 'data revolution'. Yet, with their 19th century business strategies on monetizing purely based on 'service usage' or 'voice services' has put significant pressure on margins and growth. Largely the market has gotten saturated as every player have almost similar services, offers and pricing. The Indian telecom landscape is no different, a decade ago you probably had just 2-3 players each one enjoying a niche region and double digit y-o-y growth, the story is very different now.

The chart is a good reflection of what has changed (and what has remained largely stagnant) over the last several years. The dark line at the bottom of the chart is infact the 'voice' volume over the years when compared to data. Although the voice portion has predominantly remained steady, the data shared over the medium has grown exponentially.


Being 'voice focused' Telcos have largely been reduced to the role of 'dumb pipes of communication' while the data services & businesses ()    have leveraged on the data and monetized it too.

Telcos are now awakening to new possibilities with data, towards new revenue streams and the opportunity of taking the customer engagement & experience to a whole new level.


What are these new opportunities for telcos through big-data analytics?

With hadoop based big data technologies maturing over the years, today we have tools such has IBM BigInsights that now makes it possible to analyze large amounts of un-structured data and develop better insights.
1) Churn management through Predictive customer Intelligence:
This is about developing a customer profile around 'churn propensity' and taking proactive measures in wooing customers through offers and freebies and potentially preventing churn.

2) Targetted 'smart' customer campaigns & marketing:
With the privacy norms being tightened, this is preventing operators from repeatedly reaching out to customers with marketing calls. There is added pressure in driving very specific targetted marketing campaigns based on the customers purchasing preference and potential. All this is now possible based on customer profiling and historical data analysis.


3) Location specific services:
The telcos are well placed in being able to determine the location of customers as they travel inter/intra city. This is invaluable information as telcos now have the power to engage with location specific vendors and woo customers with real-time, contextual offers pertaining to shops, retail, restaurants in the particular location which the customer is likely to pursue.

4) Fraud detection/prevention:
Being the carrier of data, it is the best placed to analyse patterns and prevent common telecom fraud scenarios including identity theft, subscription theft, etc.


5) Network optimization
With data on network usage goes a long way towards helping in better managing and planning for capacity requirements to maintain/improve QoS. Telcos shall also be able to better optimize network investments to maximize impact for most lucrative customer segments

Read more: BigData in Finance/Banking, BigData in Retail

Wednesday, February 4, 2015

Can IBM Watson gain wisdom?


IBM Watson is a cognitive computing system with natural language based analytics capabilities. Watson first shot to fame when it defeated the reigning champions in a game a jeapordy on live television. Jeapordy is a form of quiz which contestants are presented with general knowledge clues in the form of answers, and must phrase their responses in question form.



The game showcased how Watson is able to handle not just a breadth of complex questions but also understand and respond to metaphors and puns in the questions.
 
 What makes Watson different?

1) Can understand natural language: Can decipher data from natural language text sources including wikis, web pages, tweets...

2) Can learn from past results and improve: Gets better and able to learn from past mistakes.

3) Is domain agnostic: Not built for any specific area of business/domain.

It reminded me of  the DIKW pyramid and the definition of 'Wisdom' in the context of computer systems:
Reference: http://en.wikipedia.org/wiki/DIKW_Pyramid

Wisdom is defined as 'the quality of having experience, knowledge, and good judgement; the quality of being wise' and it is assumed that 'computers do not have, and will never have the ability to posses wisdom'. With Watson's ability to not just decode data but also make sense out of it and 'learn' from past experience, it looks certain the above rule on 'Wisdom' might soon cease to hold good.

Watson is already helping with:

 * Medical research:
           - Mayo clinic

 * Financial Services
           - USAA

 * Reinvent cooking
           - Bonappetit

 * Personal shopping expereince
           - Fluid Inc.

and lot more.

We might be very close to the day when Watson would be termed as 'wiser' than humans! These are very exciting times ... and a privilege to be part of an era where we are witnessing a new generation of computing - IBM Watson.




Predictive Analytics ....... what next?

I have often pondered on this question, wondering what could possibly be the next big quantum leap in the real of data and data centric de...