Do the words “Big Data” scare you? Do you panic when you think about analyzing your marketing efforts? As a blogger and an avid social media user I never thought I would need to know and understand marketing analytics, but I definitely need to. Luckily, I came across a book that is great for anyone not involved in marketing but who needs to understand marketing strategy, analytics and who would like to apply actionable techniques to their business and see results. It is called Marketing Analytics: A Practical Guide To Real Marketing Science by Mike Grigsby. This book provides great real-world examples and shows you different marketing solutions as well as how they add value to companies’ strategic endeavors.
I connected with Mike on social media after reading his book and the following is a guest post from him.
Surviving The Big Data Panic!
“Omigod! New data–different data–big data!”
“Quick, all hands on deck, we have a different data source. Throw out all we know about analytics and consumer behavior and marketing strategy and start from scratch.”
“There are new data sources! Obviously the old ways need be eliminated and we desperately need to design new algorithms and new strategies.”
Sound familiar? How many articles and books and blogs and forums and posts and seminars and emails announcing digital and non-traditional and big data have you seen? Every day! How many meetings have you attended where everyone shook their heads in FEAR AND PANIC? No one knows how to deal with these new sources of (unstructured) data. Divert all attention, stop everything now, because there are additional sources of consumer behavior.
It’s big data! It’s everywhere, you cannot escape it. Big data has become the Kim Kardashian of analytics.
OK, take a deep breath. I’ll make a few confessions and the first one is that I’ve been around nearly (gulp!) thirty years doing marketing analytics. (I have put my horse and buggy whip away for now.)
There have always been waves of data and it will continue. The 1970s introduced relational databases, storing data in hierarchical formats. Then in the 1980s came the emergence of business intelligence. That’s when I started, doing analytics when we first merged POS with marcomm responses. This is small data. We thought RFM was sooooo analytic! I saw the first panic then. People reached for things like the Taguchi method, which was about measuring inanimate objects from the manufacturing industry! It was mis-guided and inappropriate but it looked and sounded very hip. New data sources required a new approach. Then we tested it in the field and it provided nothing but confusion.
The 1990s saw the introduction of WWW and the internet. Medium data! Clickstream data arrives, a new source of consumer behavior. Of course we thought we needed a new algorithm and new strategies. Somehow we forgot it is still marketing, it is still consumer behavior. Neural networks became vogue, at least until Jurassic Park came out and Jeff Goldblum uttered those mesmerizing words…Chaos Theory! Wow, did THAT hit a nerve! For the next ten years I heard all about unsupervised techniques and voodoo / black-box things geared more for mystic rather than insights. Enter SAS with Enterprise Minor! Fortunately David Shepherd (of Direct Marketing Association) put a bounty on his website for any proof that unsupervised techniques out-performed traditional econometrics in the field. No one ever took that bounty.
The above is not to say that digital data IS NOT very different than traditional data. I LOVE clickstream data that shows just what page a consumer views, for how long and in what order. That is an amazing tracking of consumer behavior. And the new social media is bringing about a paradigm shift from outbound marketing to inbound marketing. It’s different kinds of data but why would it require new statistical techniques? Is it not still about quantifying causality?
Consumers are still behaving, shopping, choosing and buying. Right? That’s why I wrote a book (MARKETING ANALYTICS, Kogan Page, 2015) to look at this in detail. I advocate a practical application of traditional techniques applied to different kinds of (non-traditional and otherwise) data.
I’m not against new algorithms when needed. I typically do not think they are needed. I am also philosophically opposed to much of the conceptions that seem to be behind these new techniques, in that they try to remove the analyst from the analysis. Maybe I’m old fashioned and just wrong headed. The Global Chief Strategy Officer of a large prominent agency told me that the things I knew and the things I believed in are no longer valid. He has a new job now, with a much much smaller agency.
Look, additional, different data does at least one good thing: gives us new and deeper insights into consumer behavior. For marketing analysts that is always a good idea.
Additional complexity, as appropriate, is the right dimension to pursue. Over simplification is wrong. Remember the three-dimensional globe? When we simplify it into two-dimensions and flatten it out on a wall, Greenland looks to be the size of Africa. Clearly wrong.
Thus, added dimensions of complexity is a valuable input. We do not need to search for exotic algorithms or knee-jerk into wildly different strategies. We need to embrace the layers of information we have about consumer behavior and take that all into account. We have analytic techniques (and have had for decades) for doing that very thing: simultaneous equations, structural equations, Vector Auto Regression, etc. Yes these are more complex and that is where our attention should be: learning to perfect modeling that incorporates explaining additional complexity in consumer behavior. After all, marketing is, and has always been about, understanding and incenting and changing consumer behavior. That will be no different when the next wave of data hits.
Marketing Analytics introduces concepts relating to statistics, marketing strategy, and consumer behavior and then works through a series of problems by providing various data modeling options as solutions. By using this format of presenting a problem and multiple ways to solve it, this book both makes marketing science accessible to beginners and aids the more experienced practitioner in understanding the more complex aspects of data analytics to refine their skills and compete more effectively in the workplace. Get your copy today on Amazon.
About The Author
Mike Grigsby has been involved in marketing science for over 25 years. He was marketing research director at Millward Brown and has held leadership positions at Dell, Sprint, Hewlett-Packard and the Gap. With a wealth of practitioner experience at the forefront of marketing science and data analytics, he now heads up the strategic retail analysis practice at Targetbase. He is also known for his academic work, having written articles for academic and trade journals, and currently teaches marketing analytics at the University of Texas at Dallas. He is a regular speaker at trade conventions and seminars. You can follow Mike on Twitter @m666grigsby and check out his blog: marketingscience.biz.
Photo Credit: BigStock
