Abstract
Media analysts are in increasing demand as the media industry is reshaped by the evolution of technology (e.g., artificial intelligence being the latest example). As a result, undergraduate and graduate programs on media analytics have sprung up across the United States to equip students with the necessary skills and knowledge to navigate this rapidly changing media landscape and the associated job field.
In this book, C. Ann Hollifield—professor emeritus in the Grady College of Journalism & Mass Communication at the University of Georgia—and Amy Jo Coffey—the new director for the School of Journalism and Media at the University of Tennessee—offer a rich, comprehensive overview of media analytics for different sectors of the media industry, such as news, audio (e.g., radio, podcasting), video (e.g., television, streaming), social media, and entertainment media. Special attention is paid to dissecting the unique business models of each sector and how media analytics need to be tailored to meet the specific demands of each sector. For anyone who aspires to shape the future of the media industry, this book serves as a solid foundation, enabling readers to develop a comprehensive understanding of the knowledge and skills necessary to drive data-informed decision-making and optimize engagement across various media platforms.
While there is general consensus on the importance of high-quality media analytics, the discussion regarding the urgent need for a set of criteria to ensure the quality of media analytical work across all media sectors remains limited. As the authors of the book aptly point out, “media analytics is a garbage-in/garbage-out process” (p. 68), the explosion of media consumer data available online has made black-box “data crunching” unprecedentedly feasible. For example, in discussing the opportunities and challenges associated with Big Data, the book references to a case of media analytics by the Weather Company’s IBM Watson platform such that “whenever it was windy outside, juice sales spiked (Fernando, 2016)” (p. 182). Upon further digging into the source of the finding (Fernando, 2016), more intriguing findings are revealed:
The Weather Channel’s data-driven analyses show that juice sales increase during periods of high wind and low temp. Soda sales go up during snowy or rainy winters. And detergent sales peak during average precipitation and below-average temperatures.
How intriguing these findings are! But we know that finding a relationship does not necessarily they represent the real world—particularly in Big Data research. This is an area that could have been investigated a bit further in the book. Responsible data analysts need to remain critical of one’s own work and that of others by constantly asking more questions about the nature of why we find what we find. For any media analyst who is dedicated to producing quality, actionable insights, this is where the quality-check process begins, whether on one’s own work or that of others. It is essential to critically examine the data, methodologies, and assumptions that underlie these findings to ensure their validity and reliability. Responsible and experienced media analysts must ask: Could these relationships be an artifact of third variables or inflated Type II errors as a result of Big Data? If not, then what are the plausible explanations, and how can we uncover them? As much as the job of media analysts involves producing findings, it is equally important to pay attention to what not to produce. As the authors argue in this book, “one of the major challenges media analysts face is simply digging through the avalanche of data that digital analytics tools provide . . . A media analyst can easily waste time and money generating analyses that produce useless information” (p. 358)
Transcending the idiosyncrasies inherent to each media sector, how can we collaboratively establish a set of consistent and universally applicable qualities that define an effective media analyst, based on the insights presented in this book?
Solid domain knowledge of a specific media sector. As the authors meticulously elaborate throughout this book, every media sector possesses a distinct history, target audience, and business model. Hence, each sector cares about different sets of key performance metrics (e.g., circulation for newspapers and magazines versus ratings for radio and television broadcasts). A proficient media analyst must, therefore, possess a deep comprehension of the critical measures pertinent to their respective sector, as well as those of marginal relevance.
A comprehensive understanding of and well-rounded training in both traditional and computational methodologies. One prevalent misconception surrounding media analytics is the assumption that it is synonymous with Big Data analytics. For many media sectors, understanding audiences through traditional methods like surveys and focus groups (aka. sample-based data) is still as much relevant as census-based audience Big Data (p. 244). As the authors point out in this book, linear television or cash-paying movie viewers remain relevant to certain media sectors, it is, therefore, crucial for media analysts to possess a balanced skillset that allows them to effectively leverage both classic techniques and emerging computational approaches to gather and analyze audience insights from diverse data sources.
Commitment to research ethics and transparency. As highlighted in this book, the proliferation of syndicated research firms and data brokers in response to the need for media analytics has made transparency and ethics paramount to ensuring the quality of media analytics. On the one hand, the abundance of Big Data can easily mask errors or even intentional manipulations of data. How is a particular dataset being collected, structured, organized and cleaned? Verifying this has become increasingly challenging now that many data brokers explicitly prohibit the sharing of their raw data for proprietary reasons. Compounding the issue, many syndicated research firms and research vendors, on the other hand, “employ different designs, sampling techniques, and methodologies—even when measuring the same things. That can result in wildly different numbers that have real financial consequences for media companies” (p. 70). Hence, it is imperative for media analysts to emphasize the validity and reliability of their research while striving to be as transparent as possible in reporting the procedures associated with the research findings (e.g., algorithms, statistical models, assumptions).
Curiosity and adaptability. Perhaps what makes media analytics so interesting (or dangerous, depending on your angle) is the fact that no one can proclaim that they know everything about it. The media landscape is heterogenous and constantly changing and so is media analytics. As the authors point out, “the key performance indicators (KPIs) that senior executives are using today will be different from the ones they asked for yesterday and from the ones they’ll be chasing tomorrow.” (p. 276) In addition, “data can be context-specific; analysts need to know the context in which they are answering their research questions, whether the data fit the context, and what the data ‘expiration date’ might be within that context.” (p. 304) As such, it is crucial for media analysts to maintain a mind-set of curiosity and adaptability to effectively navigate the rapidly evolving media analytics domain and its ever-shifting requirements, data contexts, and performance metrics.
The transformative advancement of artificial intelligence over the past 3 years has redefined and continues to reshape the meaning, requirements, and expectations of media analysts. With the advent of smarter AI-powered tools, the ability to effectively leverage AI and make sense of various audience data will be a distinguishing factor for talents in the business of media analytics. However, this will require forward-thinking that transcends conventional boundaries and a new approach to how we prepare future media analysts through education. This book serves as a meaningful starting point for continuing the much-needed dialog about what it takes to produce quality media analytics and cultivate proficient media analysts in this rapidly evolving landscape.