COVID-Washing in US Food and Beverage Marketing on Twitter: Content Analysis

Background Food companies have increased digital and social media ad expenditures during the COVID-19 pandemic, capitalizing on the coinciding increase in social media use during the pandemic. The extent of pandemic-related social media advertising and marketing tactics have been previously reported. No studies, however, have evaluated how food and beverage companies used COVID-washing on social media posts in the United States or analyzed the nutritional content of advertised food and beverage products. This study was designed to address these gaps by evaluating how food and beverage companies capitalize on the COVID-19 pandemic to promote unhealthy foods and sugary beverages. Objective We aimed to document the types and frequencies of COVID-19–related themes in US food and beverage companies’ Twitter posts during the first wave of the pandemic in the United States, and assess the nutritional quality of food and beverage products featured in these tweets. Methods Research assistants visited the Twitter accounts of the most-marketed food and beverage brands, and screen-captured all tweets posted between March 1 and May 31, 2020. Researchers documented the date of the tweet; the number of likes, views, comments, and “retweets”; and the type of food and beverage products. We coded tweets for the following 10 COVID-19 themes: (1) social distancing, staying home, or working remotely; (2) contactless delivery or pick-up; (3) handwashing or sanitizing; (4) masks; (5) safety or protection; (6) staying connected with others; (7) staying active; (8) frontline or essential workers; (9) monetary relief, donations, or unemployment; and (10) pandemic, unprecedented, or difficult times. Researchers calculated the nutrient profile index scores for featured foods and sorted beverages into categories based on sugar content. Results Our final sample included 874 COVID-19–themed tweets from 52 food and beverage brands. Social distancing themes appeared most frequently (n=367, 42%), followed by pandemic, unprecedented, or difficult times (n=246, 28.2%), and contactless delivery (n=237, 27.1%). The majority of tweets (n=682, 78%) promoted foods and beverages. Among those tweets featuring foods and beverages, 89.6% (n=611) promoted unhealthy products, whereas 17.2% (n=117) promoted healthy products. Conclusions Our findings point to a concerning marketing tactic in which major food and beverage companies promote unhealthy foods and sugary beverages during the COVID-19 pandemic. Given that nutrition-related diseases such as obesity and diabetes are risk factors for COVID-19 morbidity and mortality, food and beverage companies should reduce the promotion of unhealthy products to help decrease the prevalence of health conditions that place people at higher risk for severe illness and death due to COVID-19.


Introduction
Health is determined not only by biology and individual choices but also by corporate practices and their influence on the social environments in which people live and work [1]. These "corporate determinants of health" were most recently defined as "strategies and approaches used by the private sector to promote products and choices that are detrimental to health" [2,3]. One example of a corporate determinant of poor health is unhealthy food and beverage marketing. Exposure to food advertising increases calorie consumption [4] and is particularly concerning during the current COVID-19 pandemic, given that obesity and diabetes are risk factors for COVID-19 morbidity and mortality [5][6][7][8].
Food marketing exerts a powerful influence on children's and adults' diets. To date, the majority of food advertising promotes products that are energy dense and nutrient poor [9]. Many research studies substantiate that such marketing of unhealthy foods affects young consumers' preferences and purchasing behaviors, and increases their consumption of those marketed foods [4,[9][10][11][12][13][14].
Social media use increased during the COVID-19 pandemic [15][16][17][18][19], coinciding with food and beverage companies' growth in marketing efforts on social media. Companies that shifted their marketing dollars to digital media-centric platforms saw significant profit [20]: ad revenue from companies like YouTube, Facebook, and Twitter grew by up to 49% during the first quarter of 2021 [21]. The extent of pandemic-related social media advertising and marketing tactics have been previously reported [22][23][24][25]. Despite companies' efforts to align with public health initiatives, COVID-washing may contribute to obesogenic environments and increase health risk. A content analysis of Australian-based social media posts from Australian food and beverage companies during the pandemic showed that 100% of unhealthy food and beverage parent companies included "COVID-washing" marketing techniques-the use of COVID-19-related themes (eg, social distancing) to market products [25]. The use of COVID-19-related marketing by food and beverage companies has also been documented in Latin America. A content analysis of Facebook posts of 5 fast food chains in Argentina, Bolivia, Guatemala, and Peru also documented the use of COVID-19-related marketing strategies during the pandemic [22], and a study in Brazil documented a high prevalence of unhealthy food advertising on the largest online food delivery platform in Latin America during the fourth month of the COVID-19 pandemic [23]. No studies, however, have evaluated how food and beverage companies used COVID-washing on social media posts in the United States or analyzed the nutritional content of advertised food and beverage products. Most of the world's largest food companies are from the United States (eg, 5 of the top 10 largest food companies are from the United States) [26], and those US-based companies are increasingly targeting other countries with their unhealthy products [27], making it urgent to understand their domestic and international marketing practices. Conversely, Latin American brands do not have a large advertising presence in the United States-the most-marketed brands in the United States largely include US-based brands along with a few brands from outside of the United States (eg, Nestle based in Switzerland and Danone based in France) [26][27][28][29][30].
This study was designed to address these gaps by evaluating how food and beverage companies capitalize on the COVID-19 pandemic to promote unhealthy foods and sugary beverages. We aimed to document the types and frequencies of COVID-19-related themes in US food and beverage companies' Twitter posts (ie, "tweets"), and assess the nutritional quality of featured products in these tweets.

Selection of Food and Beverage Brands
We conducted a content analysis of food and beverage brands' US Twitter accounts during the first wave of the COVID-19 pandemic. We chose Twitter as the social media platform to analyze in our study because companies spend US $2.9 billion of their advertising budget on Twitter [31]. Additionally, our previous research shows that food and beverage brands use diverse and powerful advertising strategies on Twitter, including adopting personalities that drive consumer engagement and interaction on the platform [32]. To determine our sample of food and beverage brands, we selected a random subset of 29 of the most-marketed brands from the following categories, using the Rudd Center for Food Policy and Obesity's Food Advertising to Children and Teens Scores (FACTS) reports: fast food [30], sugary drinks [33], cereals [34], and snacks [28]. Five research assistants then visited Twitter and searched for each brand's official Twitter account page. On each account page, Twitter suggests three similar accounts in a "You might like" list. We also included these suggested accounts in our sample. Twitter uses an algorithm to populate these "You might like" lists, suggesting additional brands that are likely relevant and of interest to users who follow and interact with the initial brand account they are viewing. Our choice to include brands from these lists was because users who "follow" a particular brand account on Twitter are likely to also follow the suggested accounts in the "you might like" list on their account page. By also including these suggested brands, we aimed to produce a robust sample of popular food brands on Twitter. This search process yielded a sample of 52 brands-29 brands from the FACTS reports and 23 brands from Twitter's "You might like" lists.

Collection of Tweets From Brands' Twitter Accounts
The lead author divided the 52 brands between seven researchers, who then visited their assigned brands' Twitter accounts and recorded the number of "followers" each brand had. To collect our sample of tweets, the researchers then scrolled through each brand's reverse-chronological feed of posted tweets, until they reached a tweet dated May 31, 2020. They then screen-captured each tweet posted between March 1 and May 31, 2020, and documented the following data for each captured tweet in an Excel (Microsoft Corporation) spreadsheet: date of posted tweet, number of likes, number of views, number of comments, and number of "retweets." We only collected tweets and retweets generated by the brands, and did not collect any user-generated content. Finally, the researcher identified and documented each food and beverage product featured in the tweet. Products were included in the data collection if they were either referenced in the text of the tweet or depicted with an image of the product.

Codebook Development and Qualitative Coding of Tweets
To create a codebook, the lead author and two senior researchers reviewed a subset of tweets in our sample to identify featured COVID-19-related themes. These identified themes were then documented and defined in a codebook to be used in a content analysis of the full sample of tweets. The lead author then trained the seven researchers on how to use the codebook to perform the qualitative coding. If throughout the coding process, a researcher identified an additional COVID-19-related theme, they notified the lead author who then updated the codebook. The final codebook included 10 distinct COVID-19-related themes: (1) social distancing, staying home, or working remotely; (2) contactless delivery or pick-up; (3) handwashing or sanitizing; (4) masks; (5) safety/protection; (6) staying connected with others; (7) staying active; (8) frontline or essential workers; (9) monetary relief, donations, or unemployment; and (10) pandemic, unprecedented, or difficult times (Table S1 in Multimedia Appendix 1).
Each tweet was then reviewed and coded for the 10 established COVID-19-related themes by one of the seven researchers. If there was a lack of clarity on how a tweet should be coded, the researcher raised the question with the research team, and through a team discussion, a consensus was reached. To account for coding consistency, the lead author reviewed 20% of the sample from each of the seven researchers. If a coding discrepancy was identified, the lead author discussed the discrepancy with the initial coder, and they reached a consensus. Researchers coded tweets for each theme that was featured, so some tweets were coded with more than one theme.

Nutritional Analysis of Food and Beverages
We calculated nutrition scores for each featured food product using the validated Nutrient Profile Model (NPM) [35], with higher scores representing less healthy products and lower scores representing healthier products. We then converted the NPM scores to a more interpretable nutrient profile index (NPI) [34] score using the following formula: NPI score = −2 × NPM score + 70. NPI scores ranged from 1 to 100, with 1 being the worst nutrition score and 100 being the best score. We classified food items with an NPI score ≥64 as "healthy," as this is the threshold for products that can be advertised to children in the United Kingdom [36]. The NPM is limited in its use with beverages and codes some sugary sweetened beverages as "healthy," so we sorted featured beverages into the categories based on the sugar content outlined in the Rudd Center's Sugary Drink FACTS Report [33]. Food and beverage products that were referenced in the text of the tweet or depicted with an image of the product were included in the nutritional analysis. In instances where a featured food or beverage product was not easily identifiable (eg, image of a brand's take-out bag), we calculated and averaged the NPI scores of the brand's top five marketed products, as described by Bragg et al [37].

Principal Findings
We identified a wide range of COVID-19-related themes. Brands used social distancing themes most frequently, often with a play-on-words using the brand name or slogan (eg, "Maintain Social TWIXstance" or "Lunchables from home"). During the first wave of the COVID-19 pandemic, between March and May 2020, we found that 78% (n=682) of our sampled tweets with COVID-19-related themes promoted food and beverage products, of which the majority were unhealthy.
Our findings are consistent with previous studies that have analyzed food and beverage marketing during the COVID-19 pandemic, showing that major food and beverage brands worldwide frequently posted on social media using COVID-19-related themes [25,38,39]. In our sample of tweets, companies most frequently referenced social distancing themes (eg, "OH____6 ft____YEAH!" [Kool-Aid], "No shame in the work-from-home leftover game" [Domino's Pizza], "having your breaks at a distance" [KitKat], or "Quarantine Cuisine" [Popchips]). This finding is consistent with previous studies that examined other social media platforms like Facebook and Instagram where the majority of posts from major food and beverage brands referenced "isolation activities" (eg, #stayhome) [25]-activities that are included under our "social distancing/staying home/working remotely" theme. Discussing social distancing and isolation activities may be useful for food and beverage brands. With millions of families remaining in their homes, people have turned to the internet to order food and beverages. The four major US food delivery apps-Grubhub, DoorDash, Uber Eats, and Postmates-reported a collective US $3 billion increase in revenues in the second and third quarters of 2020 due to the pandemic's shelter-in-place restrictions [40]. Particularly concerning is the increase in online ordering of fast food that continued to increase through 2021. A recent report by UpMenu, an online food ordering software service provider, reported that Chick-fil-A experienced a 590% increase in online ordering during the pandemic, followed closely behind by Burger King and Little Caesars with increases of 407% and 279%, respectively [40,41]. Our findings that Little Caesars and Chick-fil-A were in the top 4 most active brands posting COVID-19-themed content in our sample, suggests that COVID-washing may have played a role in this increase in online ordering during the pandemic.
Using the NPM, this study is the first to also examine the nutritional quality of products promoted in companies' COVID-19-related tweets. The analysis revealed that most products were classified as unhealthy, which is consistent with other analyses of food and beverage advertisements [42][43][44][45][46][47]. However, COVID-19 raises new ethical considerations for food and beverage companies' marketing practices because diet-related diseases like obesity and diabetes are major risk factors for severe illness and death due to COVID-19.
These findings point to a concerning yet typical marketing tactic of food and beverage companies. Corporate marketers taking advantage of trending social or health issues for profit is a textbook strategy from the corporate marketing playbook. However, using the uncertainty and disruptions of the COVID-19 pandemic for promotion is particularly more egregious than say seasonal or holiday food marketing campaigns because overconsuming unhealthy food and drink can indirectly increase the risk of more severe COVID-19-related outcomes.
Posting on Twitter and other social media platforms has enabled companies to market themselves to a wide audience, as most of the US population during our study period followed stay-at-home orders and spent more time online than ever before. Scholars refer to this kind of corporate activity as a type of cause marketing or corporate social responsibility [48] strategy, whereby companies align themselves with certain social or health issues to enhance their own image. COVID-washing portrays food and beverage companies as empathetic and responsive to the pandemic. In reality, however, this is a classic corporate strategy that may contribute to poor diet and exacerbate poor outcomes related to COVID-19.

Limitations
This study has several limitations. One limitation is that it is possible we did not capture every food and beverage brand's Twitter account in our sample. We included a subset of the top marketed brands from the Rudd Center for Food Policy and Obesity's FACTS reports in our sample, which likely omitted smaller food and beverage brands. While we also included the additional brands suggested in Twitter's "You might like" list on each brands' Twitter account, these suggested lists were not comprehensive, as they only include three brands each. It is therefore possible that there are additional food and beverage brands that we did not capture in our sample. Further, as only Twitter was included in our sample, we only captured a subset of these food and beverage brands' COVID-19-themed marketing. It is possible that some food and beverage companies may maintain a more active marketing presence on other social media platforms such as Instagram, TikTok, and Facebook. Future studies should examine COVID-19-themed marketing across all social media platforms.
A second potential limitation is that we selected the brands in our sample from unhealthier food categories (fast food, sugary drinks, snacks, and cereal) and did not focus on brands in healthier categories. Our sample did, however, include some brands with healthy products (eg, Perrier, Poland Spring, Chobani, or Subway), and many of the unhealthy brands in our sample also offer some items that are healthy. Over one-third of the brands (n=20, 39%) in our sample featured a mix of healthy and unhealthy products in their tweets; however, the healthier options were not products that appeared as frequently in the ads in our sample. Future research could examine the prevalence of COVID-washing by healthier food and beverage brands; however, we predict this would likely comprise a tiny proportion of the overall food marketing landscape because we know from previous research that the unhealthy brands spend more on advertising than healthy brands [28,30,33] and that food companies mostly promote their unhealthy products on their social media accounts [49][50][51]. Because of this discrepancy, we were less interested in comparing healthy versus unhealthy brands and more interested in documenting the extent to which these unhealthy brands are promoting junk food. This phenomenon is problematic because unhealthy brands have the most money to spend and should not be capitalizing on the pandemic to sell products given the risk factors of obesity and diabetes related to COVID-19.
Another limitation is that we were unable to measure the actual exposure and impressions of these Twitter posts. Finally, we did not examine how these COVID-19-themed ads may directly affect the dietary and purchasing choices of consumers, which is an area of future research.

Conclusions
Our findings document the prevalence of food and beverage companies' use of COVID-19 themes in tweets that promote unhealthy foods and sugary beverages. Given that obesity and diabetes are risk factors for COVID-19 morbidity and mortality, food and beverage companies should reduce the promotion of unhealthy products to help decrease the prevalence of health conditions that place people at higher risk for severe illness and death due to COVID-19.