Teaching Analytics and Digital Media to Advertising Students

I frequently have students ask me one question, “Can I really get a job in analytics with an advertising/public relations/communication degree?” My answer is always emphatically, “Yes!” Those in the advertising industry are consistently looking for employees with strong analytical reasoning skills, but more importantly, the industry needs effective communicators who can uncover a dataset’s meaning and utilize it to properly optimize campaigns. In 2019, more money was spent on digital advertising than in traditional channels, globally, and with more money being spent in a quantified environment, Walter (2020) noted in Ad Age that brands “need to lean into analytics and data science insights” (para. 2). Strong communication skills are essential to obtain value from data science insights. In a recent report published by the ANA Educational Foundation, Hilary DeCamp, the Chief Research Officer at Lieberman Research Worldwide (LRW) stated, “If you can’t communicate your analysis as if you are having a cocktail party conversation, it doesn’t matter how good the analysis is” (Lum, 2020, p. 6). Simply stated, data without easily explained, actionable insights are useless, and effective communicators are crucial in unlocking a dataset’s meaning. In fact, Forbes highlighted data storytelling as a valuable skill that is heavily desired by corporations in various industries, noting the necessity of being able to transform “insights into actions” (Dykes, 2016).

This burden to transform data into actionable insights does not rest solely on analytics teams, alone (Dykes, 2020). Specifically in the field of advertising, proper data analysis is an integral skill for more than just analysts; rather, media planners, account managers, and even creative teams must possess a foundational understanding of the digital landscape and how to evaluate and understand campaign effectiveness. As such, the development of analytical reasoning and data storytelling skills must be prioritized in advertising classrooms to prepare students for a job market that maintains data literacy as a required asset (Marr, 2019).

In this article, I discuss the approaches I have taken to teaching analytics courses within a top U.S. advertising program over the past 4 years. After receiving my PhD, I worked as a strategic data analyst in the digital advertising field developing analytics and testing solutions for a variety of global retail and business-to-business brands. Yet, despite my professional experience, transitioning to the classroom still presented many challenges, including the time-consuming tasks of locating datasets, identifying helpful instructional software, and brainstorming pedagogical teaching methods that prepared students for the critical, creative problem solving that is required for a career in advertising. After reading Li (2019)’s helpful article outlining his programmatic advertising course, I felt outlining my digital metrics course might also be helpful to other professors who find themselves tasked with introducing analytics topics into their curriculum by pointing them toward resources that have proven valuable for me and my students. Furthermore, after receiving a grant through a university-wide Experiential Learning Initiative, ¹ I have transformed some of my teaching approaches, and I hope to highlight how active and experiential learning (utilizing Kolb’s experiential learning cycle) are key pedagogical approaches to amply preparing students for careers in digital advertising analytics.

Course Learning Objectives: Blending Hard Skills and Theory

Having received feedback from students regarding the types of skills they wish to obtain in my courses, I have noticed that both undergraduate and Master’s students communicate their desire to grow their resumes, often with a focus on adding hard skills and additional certifications. As such, I approach my classroom with a firm understanding that not all students are afforded the privileges that make all internships attainable (especially internships that are offered in larger cities or those that are unpaid), so I feel it is my job to do the following: (1) help students learn how to position themselves for the job market, (2) equip them with hard skills that will make their resumes competitive, (3) utilize those hard skills as a conduit to teach higher-level strategies that persist regardless of how technology evolves within the field. The School in which I teach operates with a theory X practice model of instruction that grounds my approach to serve the various experience levels of my students.

Central to my pedagogical approach is a combination of both active learning activities and experiential learning opportunities that follow Kolb’s (1984) experiential learning cycle. This four-stage process prompts students to engage in a cyclical process of concrete experiences (having an experience), reflective observations (contemplating the experience), abstract conceptualization (learning from the experience), and active experimentation (acting on learnings) (McLeod, 2017). Each learning module in my course includes a series of in-class activities and homework assignments meant to start the experiential learning cycle by providing students with concrete experiences that apply key learnings from course content. These are low-stakes assignments from a grading perspective. Removing the structural barriers that accompany grades with these types of activities is crucial, especially in the beginning of a course, as students need safe spaces to “fail” so that they can engage in beneficial reflection. Also, most in-class activities and lab assignments are able to be completed in pairs if a student chooses to do so. It is my firm belief that analytics is a “team sport,” and when I have allowed students to work together, they often will more effectively troubleshoot any issues that arise when completing an assignment on their own. When a student is able to critically examine their mistakes and brainstorm solutions, they move into the abstract conceptualization and active experimentation sections of Kolb’s (1984) experiential learning cycle. Working in pairs/teams also requires students to discuss analytics approaches, deck creation, data visualization, and so on, with one another, and that peer feedback will help improve their skills as an analyst and data storyteller.

My digital metrics course features five primary learning modules, which I detail below:

Big Data Foundations

The first module requires students to fully understand the concept of “big data” and the issues that it presents to our society. Students discuss privacy legislation, how the advertising industry has focused on personalization of messaging, how individuals are identified online, and the different types of data that are available. Students master terminology and learn the digital landscape, including discussions around demand side platforms, supply side platforms, real-time bidding, first/second/third party data, and so on. This initial module is meant to create a solid, foundational understanding of the industry in which they hope to work. This module introduces students to various free tools during in-class activities such as Google Data Studio, ² Measure of America, ³ and Sentiment viz ⁴ by allowing them to explore the data included in each tool, identify the types of data that are present, evaluate the visualizations, and decipher the “story” each tool can present. In this module, students also complete a lab assignment that requires them to define the target audience for a brand of their choosing. This assignment is meant to reinforce the importance of the proper audience targeting and the customer purchase funnel in digital advertising. Students are required to outline how their target audience experiences each stage of the purchase funnel and brainstorm what digital tactics they would use to “speak” to them at each phase through various digital channels.

Data Analysis and Visualization

Next, I introduce a key software tool that is used throughout the semester: Tableau. Tableau provides free access to students and educators to be used for educational purposes. ⁵ This software is an analytics and data visualization tool that is widely used in the advertising industry and is a tool that I utilize throughout the semester to teach a variety of concepts. After teaching this course for several semesters, I learned to teach this tool early in the semester, as students will utilize it for a variety of in-class and homework assignments throughout the course. It is important to note that my students already possess a pre-requisite understanding of Microsoft Excel. If your students are not required to be proficient in Microsoft Excel prior to entering your course, it is a necessary skill that should be included in your course curriculum.

My goal with teaching Tableau to my students is to provide them with a basic level of proficiency and have them expand their levels of comfort with the software throughout the semester. For the Tableau assignments, we first utilize the sample Superstore dataset that is standard within the software, itself. This dataset is great for teaching students how to properly visualize and analyze retail data. I walk students through a tutorial of how to build popular data visualizations that is similar to the ones offered on the Tableau website. ⁶ It is important that the instructor spend time exploring the tool, themselves, before utilizing it in the classroom to ensure that they can confidently walk students through the tutorials. We, then, utilize free datasets that can be obtained from AirBNB ⁷ to complete both in-class and homework assignments. These assignments use active learning techniques, as students are required to take an independent, hands-on approach with the software and answer pre-defined questions using proper data analysis and visualization techniques. For example, one question requires students to use AirBNB data from New York City and answer the following question: “Create a “highlight table” that can help you answer the following question: What neighborhood has the cheapest average weekly price for a (a) boat, (b) bed and breakfast, and (c) Loft?”

In this module, students are also introduced to the differences between reporting and analysis. Here, students learn that reporting is simply a restating of what happened in a campaign; whereas, analysis is going a step beyond to inquire why it happened and what actions can be taken in response. True analysis offers meaningful, actionable insights. This lesson establishes the foundation of the data storytelling technique used in my classroom throughout the semester which requires students to answer three questions in each presentation: (1) What?—Students must identify metrics that state what happened in a campaign; (2) So What?—Students must extract the meaning of that result and describe its importance; and 3) Now what?—Students must identify an action that can be taken as a result of their analysis.

Digital Advertising Strategies and Measurement

The next module focuses on mastering how digital strategies are planned, implemented, and tested in the following spaces: social, display, and search. We discuss how to properly target audiences across platforms, different targeting strategies, what metrics indicate success, and key industry trends that they need to be aware of to be successful in the industry. At this point in the semester, students begin their participation in the Stukent Mimic Social simulation, ⁸ which provides them with an experiential learning activity. Experiential learning moves beyond my traditional active learning approaches by allowing students to explore datasets in a more open, unstructured manner, culminating in students reflecting on what they learned through the process. By engaging in tasks that simulate their eventual careers, students are able to better understand and grasp the role of a strategic analyst by experiencing the analytics process for themselves. This teaching technique is imperative for my analytics instruction to inspire my students to become more comfortable working autonomously, motivated by their own analytic curiosity rather than a complex assignment sheet or grading rubric. I believe students experience growth when we allow them an opportunity to face some degree of uncertainty and give them the space to simply explore and learn as they move through the stages of Kolb’s experiential learning cycle. This principle is key to why I selected to utilize a simulation in my course over a client partnership. In previous semesters, I have allowed students to run search campaigns for clients; however, the simulation instantly returns their data after running each round, which provides them with a low-stakes opportunity to reflect on their decisions and brainstorm how data can be used for optimization purposes. This approach also minimizes the potential for external distractions that clients bring into the course.

Each week, students must complete a “round” of the simulation, where they write their own social posts, manage a budget while considering the costs associated with content creation and promoted posts, and analyze the resulting data. A small portion of this assignment is graded competitively to simulate the “competitive” nature of social media marketing, and is a component of Stukent’s recommended grading structure. Each round, students are required to download their data from the simulation portal, upload the data into Tableau, and analyze it to determine how they can optimize their campaign to improve their performance in the next round. They create short PowerPoint decks that allow them to reflect on what worked well, what did not work well, and what they plan to do the next week. This reflection assignment simulates a weekly check-in deck that would be provided to a client and allows them to improve their data analysis and data storytelling skills. The client deck comprises the majority of the assignment’s point value (see Online Appendix).

Web Analytics

Students also obtain certifications in Google Analytics, ⁹ Google Ads, ¹⁰ and a third certification of their choice in an area of interest by completing either a free Linked-In learning course or completing another professor-approved program (Twitter Flight School, ¹¹ Facebook for Business, ¹² Tableau, ¹³ Code Academy, ¹⁴ etc.). Initially, obtaining certifications in Google Analytics and Google Ads was an optional requirement in my digital metrics course; however, students were quick to state that they wanted these crucial certifications to be a course requirement to force them to actually complete them. Now, students are not only required to become certified in industry tools such as Google Ad Words, Google Analytics, and Tableau, they are also required to utilize these tools to critically analyze data, to optimize campaigns, and report key metrics. Several programs require students to become certified in these programs; however, students also need to learn how to actually utilize those tools and gain a thorough understanding of the broader applications of each tool. Tools come and go in the digital space, so it is imperative that instructors use these tools as a conduit to teach the broader, higher-level analytics strategies of uncovering the story the data provide.

For instance, in this module, students are required to utilize Google Analytics’ demo account ¹⁵ to complete a homework assignment answering a series of questions. I provide them with a required, standard date range to use (to ensure that they will all obtain similar answers for grading purposes) and ask questions such as, “When examining the “contact us” event, did more new or returning users contact by email? Why would this information be useful?” Questions should require students to segment the data found in various reports in order to find the correct answer, while also requiring them to communicate why that metric matters. This assignment provides students with hands-on Google Analytics experience in a manner that simply completing the certification exams does not. Former students have noted that their bosses or internship coordinators have been surprised by how well they can navigate Google Analytics immediately upon being hired.

Digital Testing

Finally, students are required to complete a module on properly designing and analyzing tests in the digital space (A/B testing, multivariate testing, randomized test/control experiments; hold-out testing). One example of an active learning activity allowed students to analyze and visualize 500,000 rows of data from a display advertising vendor to measure and report on the effectiveness of an advertising campaign. Students must use the metrics included in the dataset to complete a structured assignment and approximate an incremental return on ad spend for a digital campaign. While the assignment can be found via Harvard Business Review cases (Katona & Bell, 2017), I require students to, instead, answer my own questions and require them to submit a Tableau file showing their resulting analyses and visualizations.

For this module, I also administer an additional, experiential in-class learning activity, which involves me providing students with a dataset and simply instructing them to work in pairs to analyze an A/B test and create a dashboard with the result. That is their sole instruction—no additional hints are provided. The dataset is one that I found online; however, I altered the data to introduce bias in the test design (dataset can be provided upon request by emailing the corresponding author). Students maintain the autonomy to fully explore the dataset, analyze the result, check for any issues with the test design, and determine how to report their findings. In order to set students up for success, I introduce this more autonomous, experiential learning technique later in the semester when they are more comfortable with me, my classroom environment, and the analytics software they will use to analyze data. Admittedly, at the beginning of this assignment, students are uncomfortable by the lack of direction. Yet, the primary objective of the assignment is to encourage them to utilize the skills they have gained during the semester to work through that discomfort on their own. Once they start discussing and working on the assignment with their partner, nearly all groups come up with solid, accurate analyses. After the activity, students are instructed to reflect on the assignment by discussing the following with their partner: what makes a person a good analyst, what they learned about themselves as an analyst and as a communicator this semester, and how they feel their current analytic approach relates to “real-world” situations. I’ve found that student reflection is a key component to experiential learning (citation blinded). It is through solid reflection on these activities where my students truly understand what becoming a strategic analyst in advertising entails. After all, I did not tell them; rather, I let them discover it for themselves through curated activities and assignments throughout the semester. This single activity tends to be the “lightbulb moment” where students realize that I was right at the beginning of the semester: they can work as a strategic analyst with an advertising degree.

Conclusion

Overall, teaching analytics to advertising students is necessary to properly prepare them for a career in the advertising industry. Li (2019) noted that “for our students to have a successful career in the ever-changing business of advertising, our curriculum must keep abreast of industry advances and new research findings and explore the next wave of innovation in brand communications” (p. 107). Analytical thinking is key to ensuring that our students can make well-informed decisions regarding advertising campaigns, can utilize creative problem solving, and can ask the right questions in meetings. Through proper understanding of measurement and analytics in the advertising classroom, students will graduate understanding how data can help them improve campaign performance, troubleshoot campaign issues, and maintain strategic partnerships with their clients.

Supplemental Material