Find Out Big Elephants From Big Data?

When I was a child, I lived in the countryside. It was a time when people just had very little understanding about computer and the operating system was DOS. My aunt had a grocery store and I wandered there everyday. One thing impressed me was that my aunt seemed always to know what her customers looked for.

One day, a young lady came in the shop and my aunt talked to her "Ms. Lin, I think you are coming to buy rice." The lady smiled, "yes, what do you have?" "Yesterday we just imported some fresh rice from Ping-dong, I believe you will like it," said my aunt. "Good, please weigh me a kilogram." The other day, a man rushed to the shop and said, "Ms. Chen, my wife is cooking and she runs out of soy sauce and asks me to buy it, but I don't know which brand she buys." "No problem, one moment." After 30 seconds, my aunt brought a bottle of soy sauce to the man, "she always picked this one." I was curious and asked my aunt why she always knew what customers needed, she just said "it's magic" and then showed me one thing. That's an account book. She recorded every transaction in the book, so she knew what people usually bought and she could expect when they would probably buy it again.

After 25 years, "big data" becomes a trending idea for marketers. Marketers use it for the same purpose as my aunt's, predicting customers' behavior by past records. Of course I know an account book never equals to big data. However, what I want to say is that they both go the same way. My aunt dealt with hundreds of residents in the small town and today's markets interact with millions of customers in the market. They both try to figure out the "personalized behavior patterns" to improve their business. Just we can say, looking for customers' personalized patterns becomes more important than ever because people can shop online other than shopping at stores. Besides, in the past decades, marketers persuaded customers to buy their products by standard messages and customers were thought as "mass" in the context. However, they access abundant information today so they are not persuaded easily anymore. Big data can help tailor personalized messages according to their purchase history.

However, big data is useful but not almighty. In the following article, it discusses that big data can provide us validated information but cannot tell us what to do or where to go. Some other problems such as privacy, is also an obstacle for marketers.
http://www.imediaconnection.com/content/34461.asp#singleview
Big data is an effective tool, but I think marketers' analyzing ability is more important than the tool. Not only being equipped with firearms but also having smart strategies can help us win in every battle.  

In addition, I also want to address another limitation. Customer's expected reactions are based on their past behavior. How can we do if we have nothing about them? How to find out the leads? How can we know what the potential customers are thinking? It is always painful to spend a lot of money to accumulate enough data until we can analyze their preference and provide them corresponded offers. Marketers take the process of "acquiring new customers" for granted. But the results often become reliable only after repeated tests. In digital world, customers are savvy. What they behave in a campaign cannot represent their profiles completely. Is it worthy it to spend a lot to build databases by repeated tests when we have nothing in the beginning? We may need to think about the balance between finding out big elephants and keeping company profits.