December 17, 2014
The past couple of years have witnessed a tussle between proponents of big and small data with advocates for the larger playing an outsize role in promoting its usefulness, real and imagined, for marketing in a variety of industries. But they aren't mutually exclusive although budgets appear to determine which companies lean towards big data as it almost always requires "big" resources and expertise.
Definitions for Big Data are fairly broad but in regular business parlance it has come to mean vast terabytes of data requiring analytical capabilities that exceed commonly used computational assets. Small data on the other hand usually can be computed if not entirely by humans, by more conventional means including a simple Excel spreadsheet. That said, in practice the dichotomy is less clear with a variant of Moore's law enabling increasingly affordable computerized applications of algorithms to identify and monitor key performance indicators and stats thereby leading to more effective marketing outcomes.
Apropos the foregoing a Harvard Business Review article earlier this week on big data and algorithms points out that there is a pressing need for an algorithm based effort that relies on automation rather than expensive data scientists that most firms cannot employ. HBR cites the example of a usually imperceptibly slow movement of a firm's consumer profile that is more likely to be detected manually even by the best data scientists. Automated algorithms on the other hand "are faster, more accurate, more scalable, and more adaptive than manually analyzed data". Underscoring the point is the fact that stock trading that is algorithmically driven routines beats its analog human version.
A recent Forbes article written by IHG's CRM head details how hotels are using big data to improve the customer experience while consequently driving revenue. It is a tilt at "small" rather than big data using the "vast amounts of data that our customers have shared with us" and is somewhat akin to "closed loop marketing" as big data usually requires analyzing macro trends that by harnessing data beyond what is gathered within a firm.
As the IHG chief points out most of their "guests book their hotel room through our direct channels such as on our IHG App, or through our IHG.com website, so we have a tremendous opportunity to further amplify this personalization (of guest experience)". However, like other large hotel companies that continue to grow their footprint and customer base they can soon aspire to be like Wal-Mart or Amazon which are totally customer-centric making them very amenable to "big-data" solutions.
A welcome corollary to big-data analysis has been predictive analytics which is increasingly being used not only to drive marketing resources but also revenue management including in hotels where an ostensibly "perfect" room rate can be had. Revenue-management start up Duetto purports to offer just that and claims to use data from "web shopping regrets and denials, the most popular days search for on booking sites, social review, air traffic into a hotel's city, weather forecasts (even forecasts for surrounding cities)" to arrive at that perfect room rate.
But perhaps, a cautionary note is in order for data big or small. Many of the former's proponents fail to note inherent biases that are not easy to overcome. Social media based data still tends to leave out a significant demographic while the correlation-is-not-causation exception adherents would have us believe that there is a link between "women who were interested in exotic travel and those who bought Kashi cereal"!