Skip to Content Skip to Footer
More Information or More Emotion? Suggestions on Ad Content Design

More Information or More Emotion? Suggestions on Ad Content Design

Soogand Alavi and Lam An

woman embracing car

Journal of Marketing Research Scholarly Insights are produced in partnership with the AMA Doctoral Students SIG – a shared interest network for Marketing PhD students across the world.

Despite the rise of digital advertising, television remains one of the most preferred advertising platforms. As advertising spending forecasts rise to $172.9 billion worldwide in 2021, understanding the effectiveness of ad content for different communication objectives has become a critical question for managers. Even though existing research has looked extensively at the relationship between ad spending and consumer outcomes, the role of ad content has been largely overlooked. Should managers focus more on increasing the informational or emotional content of their ads? Recent research by Ivan A. Guitart and Stefan Stremersch posits that the answer to this question depends on the communication objectives of the ad as well as the product positioning.

Using the U.S. automotive industry context, the authors compiled 2,317 television ads representing $11.3 billion in ad spending for 144 car models during a period of 3.5 years. They also collected new car registrations, volume of online search, advertising and content, and product quality ratings and attributes during the same period. Their main findings indicate that advertisements with a high level of emotional content generate higher online search volume than those with lower levels of emotional content, regardless of product positioning. Both informational and emotional content positively influence sales. However, increases in informational content lead to more incremental sales for low-price and low-quality cars than for high-price and high-quality ones. In turn, increases in emotional content generate more incremental sales for high-price cars than for low-price cars. Furthermore, the conversion rate from online search to sales is higher for low-price and low-quality cars than for high-price and high-quality ones. 


The findings of this research offer important guidance for managers. Managers of high-end cars should prioritize emotional rather than informational content in ads, while managers of low-end cars should emphasize emotional content if their objective is to increase online search and emphasize informational content if their objective is to increase sales.

Given the relevance of the findings and important implications for managers, we reached out to the authors to learn more about their research.

Q: What was your motivation behind this research, and why were you interested in pursuing this topic? 

A: Advertising is central to the marketing profession. Marketing academics have frequently studied the effect of ad spend on behavioral outcomes such as search and sales. From such studies, we know the average effect of advertising spending is small, but this effect is very different across campaigns: while some firms succeed in setting up campaigns that shape consumers’ search and buying behavior, others have no effect at all. Clearly, the specific ad content of a campaign chosen by a specific firm plays a role in its effects on search and buying behavior—at least the entire ad agency world seems to believe this. However, ad content is surprisingly rarely studied in academic ad spending studies. We wanted to address this and close this gap in the advertising literature, recognizing that ad content may also have differential effects on search and sales and providing initial estimates for what such effects may be in an industry that advertises a lot, namely the automotive industry.

Q: What are the main challenges facing the industry about this topic in the time period of the data set? Have these industry-based challenges changed with the improvement of digital advertising over the years? 

A: Determining what type of content is most effective has always been challenging. Advertisers have traditionally relied on copy testing and surveys to understand which ads work better. However, these pre- and post-roll tests suffer from the well-known attitude–behavior gap and restrict researchers’ ability to single out the effectiveness of individual content types.

Digital advertising (and the technologies associated with it) provide opportunities to overcome this challenge because it allows us to easily measure the effects of ads on online behaviors such as visits to the website or purchases. Moreover, companies can increase the impact of their television ads by pretesting them digitally. For instance, companies could release different video ads in a test market using YouTube and then choose the best performing copy for the national rollout in television. Moreover, if a large pool of ads and their content are available, companies have a relatively straightforward way to assess the impact of individual ad content on behavior.

Q: What practical implications do you suggest for managers of low-price and low-quality cars?

A: Managers should use informational advertising unless they wish to increase online search. Online search allows consumers to increase their knowledge about products. Managers might want to increase online search when they think that this search will reveal information that is not delivered in the ad but might still be important for consumers. Online search might also lead to other benefits such as brand building and awareness of other products of the brand, although we do not quantify these in our article.

Q: For the academic community, what are the main challenges for researchers interested in continuing this topic?

A: Access to high-quality data on advertising content. For our research, we bought the ad videos and then asked research assistants to code content. This was time consuming and expensive. We literally had a dozen assistants sitting in a room watching hundreds of ads and filling in surveys with hundreds of questions for many, many weeks. A coding experience as long as this was quite taxing not only for the coders, but also for us; we had to frequently supervise the lab sessions and run quality checks on the data. Another important cost associated with the coding was the use of physical infrastructure. We were lucky that IESE Business School agreed to block a computer lab for us for such a long period of time.

Nowadays, some firms offer the ad content to researchers, although the data only provide information on a limited set. Another alternative is to buy the videos and use machine learning to extract the content. In any case, access to data continues to be challenging because it is limited and/or expensive.

Q: In your opinion, how will your article’s findings hold in the current era of digital advertising? What are some factors that have changed in digital advertising that might alter your main results? How can future research address this?

A: We believe our results should hold today despite the rise of digital advertising. First, television is still an important advertising channel and receives about one-third of companies’ total ad spend. Second, despite the increase in the time that consumers spend online, the consumption of television has remained stable over time. For instance, Statista indicates that an average American watched 2.73 hours of television in 2010, and this number was 2.81 hours in 2019. Third, although the online presence of consumers has increased over time, during the period of our study the predominance of consumers’ online search was already high (77% of new vehicle buyers were using the internet in the shopping process back in 2010). If anything, the increase in consumers’ online presence (e.g., using mobile devices) has strengthened the connection among television advertising, online search, and sales, so our results should hold in the current environment as well.

Digital advertising allows companies to target audiences, but targeting is still somewhat limited in television advertising. It is unclear how our results change across audiences, so this adds another layer of complexity to the problem we address in our article. For instance, cars that are at the same price and quality levels could be designed to satisfy the needs of individuals or families. We don’t know if these segments react equally to informational and emotional content, yet this is important information for managers. Future research could explore this issue, for instance, by conducting field experiments in which the audiences receiving ads are experimentally manipulated.

Q: Are there any influential platform-based factors that may impact the results of your study? Specifically, how will these findings hold in other mainstream advertising platforms?

A: Connected with the previous point, different platforms host different audiences, and these audiences could be quite different from television audiences. For instance, the average Facebook user tends to be younger than the average television viewer. Thus, it is hard to assess whether our results would hold across platforms.

Additionally, different media platforms allow different advertising formats. We believe that, if we control for audience characteristics, our results about ad content effectiveness should hold across platforms when advertisers use video ads. However, we do not know if our results would hold, for instance, for interactive stories in Instagram or search ads on Google.

Finally, another important factor is that digital ads generally contain direct links to the website, where consumers can find additional information about products. Thus, information acquisition might be more focused on the advertiser’s website compared with the search that occurs after television advertising. This provides an opportunity for brands in that they can have more control over the information consumers will find online about their products.

Read the full article: 

Guitart, Ivan A., and Stefan Stremersch (2021), “The Impact of Informational and Emotional Television Ad Content on Online Search and Sales,” Journal of Marketing Research, 58 (2), 299-320.

Soogand Alavi is a doctoral student at University of Texas–Dallas.

Lam An is a doctoral student at University of Central Florida.