Machine learning and customer attitudes

A recent issue of the McKinsey Quarterly touches upon a couple of oft-repeated but still little understood topics: combining big data to get bigger marketing and  bringing artificial intelligence to better understand customers. The former points out how "technology (alone) is not enough;  what’s needed is a practical approach for creating a workable partnership" between the CMOs and CIOs and, in particular, for marketing departments to know what to look for. 

Per the Quarterly, fusing the two areas is perhaps best achieved by creating areas of excellence that span both departments with an emphasis on  transparency with "scorecards" on progress. The establishment of "translators" who can bridge potential gaps in understanding between somewhat divergent specialties. That may call for retaining a "business-information officer" whose role includes translating business strategy on an enterprise-wide basis resulting in a more "effective use of big data and other technologies" thereby setting apart winners from losers.

The other article in the Quarterly's  goes on to herald the coming of the "second" machine age when machines "overcome the limitation of mental power" as opposed to the first machine age when the limitations of muscle power was overcome.  Pointing out that "machine-learning algorithms are actually as good as or better than humans" at many things that are thought of as uniquely human capabilities, the article notes the exponential rather than geometric growth that it brings about as data and computational capability grow exponentially. Further, the results of "previous machine-learning exercises can be fed back into the algorithms resulting in each layer becoming a foundation for the next layer of machine learning with the whole thing scaling in a multiplicative way".

The foregoing when applied to customer behavior data can result in clear pointers as retailer Pier 1 Imports did when they  used advanced predictive data analytics and machine learning to fullfill their goal of   understanding their customers and serve them better with a more personalized experience across all interactions and touch points within their brand.

It is not clear that Gallup  used machine learning to gauge hotel customers' needs  but the polling company earlier this week came up with the results of a survey across all segments of the hotel industry and found that what customers want depends to a considerable degree on the segment with a few elements like value price, reputation, room quality and location spanning all segments. A notable if somewhat obvious finding was that "engaged" customers provide a "financial premium".

Unfortunately, a mere 20% of hotel guests were "fully" engaged with those belonging to higher-priced hotels' being more engaged.  The ability for hotels on the lower end of the price scale to engage customers may not necessarily be limited as more often than not it merely requires offering services such as "nicer" television for which they are prepared to pay. That seems like an opportunity to improve profitablity at more than one level.

Published by

Vijay Dandapani

Co-founder and president of a New York based hotel company for 24 years. Grew the firm to five hotels in Manhattan and also developed a greenfield project at MacArthur airport, New York. Speaker at numerous prestigious forums including Economy Hotels World Asia, Lodging Conference, NYU, Columbia University Real Estate Roundtable, Baruch College's Zicklin School and ALIS. President and ceo of New York City Hotel Association since January 2017.