Big Data. What’s all the fuss about? (Part 1)
Take a bank like State Bank of India. It has 31.5 crore Savings Accounts. That's like the entire population of USA, every man, woman and child, holding a savings account in one bank! That's how big State Bank of India is. It is easy to see that if SBI just understands and addresses its existing customers' every need, it may never have to acquire a single new customer! Now, how can SBI do this?
Given its extremely large data base of customers, any analysis of their behaviour will involve observing and making sense out of hundreds of bits of data. That's what Big Data Analysis or Big Data in short, is all about. To understand how this is done let's take an example. Say the majority of people applying for car loans are between the age of 30 and 35. Let's say that SBI decides to e-mail all people in this age group, offering them a car loan. It then discovers that there are 15 crore account holders in this age range! Sending out 15 crore mails, even if they are customised only to the extent of addressing each recipient by name, would be a humongous task costing millions! So SBI would like a more efficient way to figure out who among their customers are more likely to want a car loan. Enter data mining, a big part of Big Data.
In this step you find more parameters common to those who are likely to need a car loan than just their age. You find that most applicants are salaried with a monthly income over Rs.50,000. Most are men living in cities. Married. Now instead of just one criterion - age, you have 6 - age, male, salaried, earning over Rs. 50,000 p.m, married and living in a city. Big Data now looks into the giant data base, into each and every savings account holder and finds people who match all these 6 criteria. Suddenly the population of 'people likely to need a car loan' drops to just 3 million! Now SBI can devise an appropriately customised mail to address this audience and hopefully get a rich haul of car loan takers!
In very simple terms that's how Big Data works. It must be immediately apparent to anyone that Big Data Analytics must be quite expensive. It requires a tremendous amount of computing power, efficient data storage and querying, perfect communication between the IT engineers and marketing guys and finally, a great deal of time. Most marketing departments would find that list of requirements daunting. Is there an alternative way of doing this…even in this digital age of Big Data??
Watch this space!
To get our digital marketing techniques working for you, click here!