Extolling the Value of Big Data
By Bertrand Chen, Data Scientist, Asia Miles
It all started on a flight in 1953.
Blair Smith, an executive from IBM, sat next to C.R. Smith, then CEO of American Airlines. Besides sharing the same last name, they were both trying to solve a business challenge. The airline industry needed a reservation system to cope with the explosive growth of passengers’ volume. Thus SABRE reservation system was born. This was the first real time application for mainframe computers. It is hard to imagine now but airlines were once a hot bed for digital innovations.
We live in a world dominated by Google and Facebook. Users’ attention is sliced into micro-moments sold to the highest bidders on ad exchanges. We are in the business of courting customers’ attention. Your future is determined by the data you collect and your proficiency at maximizing its value. If Blair Smith were on that flight now, he would be extolling the value of Big Data. But there are three pitfalls that C.R. Smith would need to be aware of.
Firstly there is usually an unrealistically high expectation from data. Given the large amount invested, one is expecting a panacea. In building customer loyalty for sustainable business growth, developing a modern data warehouse is merely the first step. Without having specific objectives, you could just be building another highway to nowhere. Extracting a single view of customers is not enough. One needs a customer lifetime value to make business decisions.
Secondly there is an archaic separation of roles. Traditionally, Analytics was pegged as a support function. Business units asked questions and analytics provided answers. This however is not the case anymore. Analytics now works with marketing to decide the most effective and efficient communication channel as well as the timing of the communication. Together both functions leverage the data infrastructure and truly personalize the customer experience to optimize user engagement and thus maximize long term profit.
Lastly there is a difficulty in hiring the Talent. There has been much hype surrounding the field of Data Science but nobody has a clear idea on who its practitioners are. Are they software engineers, statisticians, machine learning experts? The only agreement is that they need to have domain expertise and be able to explain complex concepts.
Such Data Scientists are rare and probably would never step outside of Silicon Valley or Alley.
The only solution is then to build a diverse team. Even if no one possesses all the requisite skills, the collective knowledge of a team would cover all the requirements. One could also include in the team someone with business knowledge and deep domain expertise who can guide the team towards projects that would yield the biggest dividends.
If C.R. Smith were to grasp those three pitfalls, he would see the tremendous opportunity in front of him.
By harnessing the technological changes unleashed by Google and Facebook, the Loyalty business could dominate the marketplace.
This is just the beginning.