Page 118 - MYM 2015

Today even more new data sources are developing. Geographic data from mobile devices has dramatically increased the scale of big data again. Promotions and ads can be delivered based on our exact physical position.
Over 40 years data scale has grown bigger and bigger. What appeared to be “big” one year became “small” in comparison to the emerging new data sources. This exponential increase in the volume of data and associated increase in the power of marketing analytics has generated the excitement we all feel for “big data”.
What Big Data Can Do
Big data has allowed targeting of marketing efforts to the individual consumer in real time. The analytics surrounding such targeting has generated a new
set of media options and
service companies to
provide marketers targetable
advertising tools. Internet
TV and interactive TV will
make even TV ads targetable
to individuals. Integrating
purchase histories via credit
card records promises
comprehensive pro ling of
consumer behavior. The new
big data enables even more
targetable marketing by
using geographical location
for ads and promotion.
media right? A CMO might ask “Should advertising total spend be increased or decreased next year”? Big data can supply a comprehensive history on exposure (e.g. banner exposure) but often cannot link this exposure
to sales in products like autos where a purchase is not made buy a click. Big data cannot tell how exposure converts to sales in these cases. For example, BMW has over 20 million likes on its social media sites,
but how do these likes link to strengthening brand image and sales? Big data is good at exposure and involvement monitoring (e.g. clicks on Facebook BMW news feed or views YouTube video posted by car owner). But to measure deeper behavioral measures like consideration of BMW as a new purchase requires a controlled consumer sample exposure to new
media where pre/post surveys quantify brand image, consideration and purchase probability. Such controlled panel tests can measure attitude changes that are not
As well as speci c targeting,
general mining of the big
data base can nd signi cant
opportunities. New insights
and speci c behavior patterns can not only generate better targeting but improved consumer service and loyalty programs. New marketing analytic tools are appearing. For example new sales forecasting tools have been developed based on the frequency of search words used (Brynjolfson and Wu, 2013). Geographic data not only will allow geographic promotions to consumers, but enable better alignment of distribution systems. For example heat maps of cell phone concentration maps can be used to supply emergency and health services.
What Big Data Can Not Do
Big data has generated well deserved enthusiasm, but there some things big data cannot do well. For example, big data is great for targeting advertising, but it cannot tell us much about how many dollars to allocate to
TV versus new media or how much to spend in total on marketing. Is the current 40% of spending on new
118 I October 2015
Marketing is at an exciting point in its history and big data growth will drive further innovation.
recorded in big data sources.
Big data can tell us where
we are, but cannot create new innovations. It may identify opportunities, but cannot generate a new creative campaign or product. For example storytelling in theory is powerful in social media, BMW did not realize the quantitative impact of storytelling until they did
with MIT based on pre/post measurements in a panel of potential buyers with tests of consumer generated stories. After this study they created #BMW stories and improved their social media
involvement and brand ratings. Similarly, areas of new product potential may be identi ed by big data, but creating and testing a product requires custom marketing research and innovative thinking by managers.
A Balanced View of Big Data
What is needed is a balance view. Maximize the potential of big data, but use custom market research where the big data does not generate solutions to major strategic problems. Be both data and problem driven. Search big data for wisdom and guidance, but also conduct behavioral research if it is needed to understand the consumer decision process.
One exciting area of marketing analytics that integrates big data and strategy is marketing experimentation. A/B testing of alternate creative advertising with micro big data can generate a depth of understanding and sales response. Test marketing of new products with