Brand crises in the digital age: The short- and long-term effects of social media firestorms on consumers and brands

Highlights

• Social media firestorms as a digital kind of brand crisis are relevant phenomena.

• Negative effects of firestorms can occur in both the short- and long-run.

• Their consequences for brands and consumers vary strongly with contingencies.

• Firestorms that spark many tweets and those caused by product failures are impactful.

• Vivid triggers of social media firestorms relate to more negative effects.

Abstract

Social media firestorms imply the sudden occurrence of many, predominantly negative social media expressions against a brand. Do such firestorms leave a mark on consumers and their brand judgments—in the short term but also over time—to a degree that deserves managerial attention? What kind of firestorms have the strongest destructive potential? This manuscript treats firestorms as a digital form of brand crisis and proposes a conceptual framework to identify which firestorms harm short- and long-term brand perceptions and become part of consumers' long-term memory. A unique data set combines secondary data about 78 real-life firestorms with daily brand perceptions obtained from the YouGov panel and survey data from 997 consumers. The results indicate that of all affected brands, 58% suffer from a decrease in short-term brand perceptions, and 40% suffer long-term negative effects, suggesting that social media firestorms can indeed harm businesses but also show that strong variations exist. Contingency analyses of the conceptual framework with regressions and generalized estimating equations indicate that social media firestorms are most impactful in terms of negative brand association changes and/or memory effects when they are initiated by a vivid trigger (e.g., video in the first firestorm tweet), linked to a product/service or social failure, characterized by a large volume of social media messages, and when they last longer.

Introduction

Brands may spend years building their reputations, but they can go down in seconds, especially with the pace of digital media.

—Marketing News Weekly (2017)

When New York's taxi drivers struck for an hour at Kennedy Airport, in protest of Donald Trump's so-called “Muslim travel ban,” the ride-hailing company Uber announced on Twitter, “Surge pricing has been turned off at #JFK Airport” (Cresci, 2017). Within seconds, consumers voiced their criticisms of Uber's strikebreaking behavior and apparent support for the ban, using messages such as “Congrats to @Uber_NYC on breaking a strike to profit off of refugees being consigned to Hell. Eat shit and die” (O'Sullivan, 2017). Soon the hashtag #DeleteUber spread, transforming into a vast protest marked by thousands of angry, emotional tweets (Collins, 2017; Cresci, 2017); it was the top trending topic on Twitter that night and in the days that followed. Major news media reports about the firestorm further fanned the flames throughout social media.

Uber is by far not the only company to have suffered such a “social media firestorm.” United Airlines experienced a global firestorm following a shocking cellphone video of a security officer violently removing a passenger from an overbooked flight (Victor & Stevens, 2017), and Abercrombie & Fitch found itself subjected to waves of angry, emotional comments on Twitter after CEO Mike Jeffries said in an interview that he only wanted “cool” and “popular” kids to wear the chain's clothes (The New York Times, 2014).

Such social media firestorms have become standard elements of the modern marketing environment, constituting a digital form of brand crisis. Whereas in the pre-digital age, consumers were only able to express their dissatisfaction with a brand individually, either by contacting the firm or talking to a very limited number of others, without much opportunity to bundle their protests (Rauschnabel, Kammerlander, & Ivens, 2016), the rise of social media has shifted power from organizations to consumers (Hennig-Thurau et al., 2010; Labrecque, vor dem Esche, Mathwick, Novak, & Hofacker, 2013). Social media enable individual consumers to inform large numbers of other consumers in a very short time frame by co-creating and spreading brand-related content through their digital networks (Labrecque et al., 2013). This environment of social media networks sets new, digital forms of word of mouth apart from their analog predecessors, by enabling, for instance, real-time access to large numbers of friends and followers (Hennig-Thurau, Wiertz, & Feldhaus, 2015).

As a digital kind of brand crisis, social media firestorms exhibit their own characteristics, reflecting the influence of the social media environment. Whereas part of what we know about “traditional” brand crises likely applies to firestorms, other factors are likely affected by the specificities of this digital surrounding, which require new understanding. For example, social media are characterized by extreme dynamics, meaning social media firestorms can experience exponential growth in a very short time frame. As a result, social media firestorms can constitute an important threat to firms' brand assets, in line with many corporate leaders' sense that “reputation risk” is a grave threat (Deloitte, 2015). But this judgment is not undisputed; some commentators suggest these threats have been exaggerated, with little actual economic meaning for firms (e.g., McKinsey, 2012), especially in the long run (i.e., “memory fades,” Knowledge@Wharton, 2015).

We aim to contribute to this debate by investigating the specific negative effects of social media firestorms on consumers' perceptions of a brand. This approach anticipates that the power of social media firestorms will not be “strong” or “weak” in general but rather will vary, according to certain contingencies. We derive contingencies from general research on brand crises (e.g., the reason for the firestorm), but also from the digital characteristics of social media firestorms (e.g., amount of social media word of mouth). With the empirical insights gained from this study, we seek to help managers identify threatening social media firestorms early, so that they can allocate sufficient and appropriate resources to prevent their escalation. With firestorms having become a regular occurrence for large firms, marked by enormous and radical dynamics, we consider this insight essential for establishing effective corporate responses. In addition, our findings suggest ways to determine appropriate metrics to track the development of firestorms.

To achieve these contributions, we build on brand crisis and social media research to craft a theoretical framework of firestorm characteristics that prompt negative changes in brand perceptions in the short and long run, as well as those that cause the firestorm events to become part of consumers' long-term memory. We link firestorm factors to the elaboration likelihood model of consumer behavior (ELM; Petty & Cacioppo, 1986), which we apply to the social media environment. Leveraging the ELM, we derive factors that might influence how consumers process and remember firestorm information (i.e., based on their motivation, ability, and opportunity). The resulting framework distinguishes trigger characteristics of social media firestorms that are initially detectable (e.g., vividness of the trigger) from evolving firestorm characteristics (e.g., its length). This distinction offers brand managers both early warning indicators and important metrics to track throughout the duration of a social media firestorm.

We then test our conceptual model with two studies. First, we check the brand-related hypotheses using YouGov data about German brands affected by firestorms between January 2012 and December 2014, applying an event study approach. We analyze the effect of our ELM-based trigger and firestorm characteristics on negative abnormal changes in brand perceptions in the short- and long-term. Second, we test the consumer memory-related hypotheses by combining our data set with a survey that we used to measure consumers' memory of the respective firestorms, two years after their occurrence. With this step, we aim to assess whether changes in brand perceptions stem from a particular firestorm. Here, we use generalized estimating equations (GEE) to account for the hierarchical structure of the data.

Results show that both trigger and firestorm characteristics drive short- and long-term effects of social media firestorms on consumers and brands. Strong social media firestorms, with a high volume of social media messages (i.e., tweets), relate to negative changes of consumers' short-term brand perceptions, as well as two years later. Firestorms caused by a defective product or service have stronger consequences than those triggered by inappropriate communication strategies by a company. The length and vividness of a firestorm additionally influence consumer memory, as do firestorms caused by a social failure. We discuss the managerial implications of these findings to conclude this article.