Thus far, 2020 has been a challenging year, to say the least. The COVID-19 pandemic, economic collapse across countries, autonomy protests in Hong Kong, protests of police brutality and racism in the United States, and a border dispute between India and China are a few of the major events and conflicts that have dominated public consciousness of late. Given that we are living at the intersection of so many momentous and uncertain developments, it may come as no surprise that in addition to being turbulent times, these are emotional times.
Data from Twitter can help illuminate the collective mood. We examined a sample of 1 million tweets per month between January and May of this year using the Linguistic Inquiry and Word Count (LIWC) program to identify words conveying either positive or negative emotions. We observed a roughly 8 percent decrease in positive emotionality over that time.
The percentage of words in the sample identified as conveying positive emotions declined over the first five months of the year, from 3.42 percent in January to 3.15 percent by May.
Chintagunta et al., 2020
The highly charged and increasingly negative emotions surrounding current events make this a difficult environment for marketers and other corporate communicators. Historically, many companies have taken apolitical stances on controversial events; most have preferred to be neutral observers when it came to sociopolitical topics. However, customers today, especially millennial customers, expect the companies to whom they are loyal to be more vocal about such issues. They expect them to take a position that aligns with their own beliefs. This expectation is reflected in millennial purchasing behavior: a Boston Consulting Group study finds that millennials are more likely to purchase products that are associated with a particular cause than are nonmillennials.
How they responded
Airbnb ▸
Response 1: Twitter statement and $500,000 donation announcement
Amazon ▸
Apple ▸
Netflix ▸
Pepsi ▸
This new customer expectation has resulted in more companies issuing corporate responses to a wide variety of issues. To better understand the reputational implications of these statements, we analyzed public response (on Twitter) to corporate statements issued by five companies— Airbnb, Amazon, Apple, Netflix, and Pepsi—on the recent events around police brutality and the Black Lives Matter movement. We picked these brands for the analysis because each had a substantial volume of conversations that we could analyze before and after its response.
We observed that most of these brands issued corporate responses that affected public sentiment positively (i.e., they increased positive emotions). We did so by measuring the change in the percentage of words reflecting positive emotions in tweets about a company between the week prior to the statement and the week after the statement. We measured this change relative to a control sample of 100,000 random tweets per day during the same duration to make sure the results reflected the impact of the corporate responses and not a change in overall sentiment on Twitter.
On Twitter, corporate responses ultimately resulted in a higher concentration of words associated with positive emotions.
| Company | Change in concentration of positive words |
| Amazon | +18% |
| Apple | +27% |
| Netflix | +25% |
| Pepsi | +/- 0% |
| Airbnb | +780% |
Chintagunta et al., 2020
However, just issuing a statement was not enough. The positive impacts noted earlier came after statements in which companies promised concrete actions and investments for the cause. For most of the companies we analyzed, such a concrete response was the second or third statement they issued. Most started by issuing a “words only” statement of support for the cause; these were usually not well received by customers, and many of them led to an increase in negative sentiment. Only when the companies pledged actual investments did the negative sentiment reduce.
Furthermore, the experience of Netflix suggests that just pledging any amount of money is unlikely to be enough in the eyes of the public, or at least in the eyes of Twitter users. The first two statements issued by Netflix were words only and led to a decrease in both positive and negative sentiment—that is, they did not evoke much emotion. However, in its third statement, Netflix pledged $5 million, and this led to a 61 percent increase in negative sentiment (and a 31 percent increase in positive sentiment). Only when Netflix issued a fourth statement pledging $120 million did the negative sentiment go back to normal (while the positive sentiment remained at the high level).
Airbnb was the only company in our analysis that pledged money in its first statement, and thus saw a positive impact right away. A disclaimer here is that these are short-term impacts; longer-term impacts might vary depending on a number of other factors.
Responses from each company are arranged from left to right in chronological order. The impact of each response is measured relative to the concentration of negative words observed prior to the company’s first response.
Chintagunta et al., 2020
The companies in our analysis are far from alone in having had to issue additional statements, publicly apologize, or change their stance after customers felt their response was inadequate or inappropriate. One example is Starbucks, which had to reverse its policy prohibiting employees from wearing Black Lives Matter–related clothing in stores after a public outcry. Not only did it change its policy, but the coffee chain also said it would make 250,000 Black Lives Matter shirts for baristas who wanted them.
Crafting the right corporate response, one that includes not just sentiment but also concrete actions and initiatives, is a balancing act. The delicateness of this task is further exacerbated by increasingly divergent political views. For example, a Pew Research Center study conducted in late April finds that 82 percent of Democrats viewed the coronavirus outbreak as a threat to US health, but just 43 percent of Republicans viewed it as such.
Given the perils of getting the response wrong, and the difficulty in getting it right, it would be wise for companies to develop a nuanced strategy and an associated playbook to guide them in creating thoughtful responses to social issues. Naturally, the right strategy won’t be universal: it will incorporate the company culture and ethos, the values and beliefs of the company’s customers and employees, the perceived seriousness of the topic, and the magnitude of the response that customers are likely to expect. The nature of the business will also have implications, as retail and consumer-facing businesses are more likely to be subject to heightened scrutiny by the public, whereas other businesses would need to be concerned primarily about employees and other internal stakeholders.
The specifics of an appropriate response to social crises may vary considerably across businesses, but the prudence of planning for such responses does not. Customer (and employee) expectations of corporate purpose and social response will likely only increase. And as 2020 has shown us, the next crisis is always around the corner.
Pradeep K. Chintagunta is the Joseph T. and Bernice S. Lewis Distinguished Service Professor of Marketing at Chicago Booth.
Yogesh Kansal is a consultant at Boston Consulting Group and a recent graduate of Chicago Booth’s MBA Program.
Pradeep Pachigolla is a research assistant at Chicago Booth and an incoming PhD student at Cornell.
Chicago Booth’s Gregory D. Bunch explains how business leaders can cope with VUCA conditions.
How to Navigate a Volatile World
Consumers may be wary of autorenewing promo deals.
The Equation: Why Some Subscription Offers Can Turn Off Customers
A natural-beauty movement on social media has not necessarily discouraged cosmetic use.
‘Natural Beauty’ Isn’t Effortless (or Free)Your Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.