What's the Science Behind Marketing?

Earl Taylor: MSI's corporate sponsors represent a cross-section of the U.S. business community. We continuously solicit input from our trustees and others who represent our corporate members about their most pressing marketing challenges.

Every two years, we ask our corporate trustees and leading marketing academics to prioritize these topics to guide our funding of academic research and the focus of our events. Results are summarized in our recently released 2016-2018 Research Priorities, which can be viewed and downloaded free from our website at http://www.msi.org/research/2016-2018-research-priorities//.

PB: How do you use big data to measure brand performance?
ET: Marketing academics and MSI corporate sponsors are using big data in a variety of ways to assess marketing effectiveness and brand performance. Increasingly, datasets that link exposure to advertising and other marketing activities with outcomes such as sales and profitability allows managers to determine the exact effects of each element of the overall marketing mix and to allocate resources more efficiently and effectively.

Academic research supported by MSI has demonstrated that properly interpreted and weighted data from social media can be closely correlated with traditional brand health tracking metrics. In fact, social media can yield leading indicators of brand health, allowing managers to anticipate and respond to emerging trends (positive or negative).

MSI helped found and continues to support the Marketing Accountability Standards Board (www.themasb.org), which is dedicated to vetting and promoting metrics agreed to by both marketing and finance that can reliably demonstrate the value created by marketing and branding.

PB: How can data and analytics help tell a marketing story?
ET: Most traditional quantitative market research relies on theories of consumer behavior that yield hypotheses that can be tested against empirical findings. The advantage of this approach is that results can be incorporated into a coherent framework for interpretation and application, yielding a cumulative body of knowledge over time.

With the advent of big data analytics using machine learning and other techniques, we can now efficiently discover patterns in data that we might not otherwise have noticed, but which can be interpreted theoretically and applied, thus advancing marketing science and practice.

Regardless of how they are obtained, insights are best shared the way humans have always communicated' through stories, personas and the like that allow managers to understand, assimilate and extrapolate from them as new situations arise.

PB: What will people gain from attending your conference presentation?
ET: When data does not readily fit existing quantitative formats and analytics, it is often referred to as 'unstructured." In fact, datasets taken from social media, online review sites and the like are highly structured! Whether in real-time or asynchronously, exchanges in social media are variants of the structures that inform ordinary conversation where sequencing and context largely determine what a given contribution means to others engaged in the dialogue.

While certain insights can be derived from techniques that extract words or phrases and re-assemble them as word clouds, in many cases preserving sequential structure is critical to understanding what consumers are saying and why.

Drawing on sociological research, I will make the case that conversational analysis offers a distinct alternative to purely inductive big data analyses of social media. Importantly, findings on how information is conveyed in stories, jokes and other forms of ordinary conversation can also help us better communicate insights from all forms of quantitative and qualitative analyses.

Want to hear more from Earl Taylor? Join us at the Marketing Analytics & Data Science Conference. Learn, network and share best practices with the most influential leaders in data science and analytics. Stay connected at #MADSCONF.