Given the increasing interest in sustainable food consumption and production, this study aims to understand how consumers perceive the value proposition of vegan food. Over 120,000 tweets relating to veganism were extracted from Twitter, which were then analysed using the text analytics tool Leximancer to ascertain the predominant themes of conversation taking place around vegan food. Our results show that, in light of the three main drivers for vegan food choice—ethical, personal health, and environmental—surprisingly, we see a limited number of environmental or sustainability motivated tweets. This is a significant finding, as, while vegan food consumption is reported to be sustainable, this is not a preferred topic of conversation for consumers. Value propositions communicated with respect to personal health attributes (e.g., dairy free, gluten free, and nutrition), and consumption benefits (e.g., tasty, delicious) are more likely to resonate with consumers and motivate increased consumption while concurrently delivering environmental benefits as a positive side-effect. Furthermore, the polarity of the attitudes and conversations taking place between vegans and non-vegans on Twitter underscores that a single value proposition is unlikely to reach both groups simultaneously and that different value propositions are likely to be required to reach these respective groups.

Trust is fragile. The 2018 Facebook and Cambridge Analytica debacles highlighted how data harvested from social media platforms can be used not only for commercial purposes but also for political manipulation. This incident and the widespread discussion around it further demonstrated the following issues: unethical data collection enabled by a platform; unethical use of data for corporate and political interest; and unethical data sharing by an academic. Research needs to be credible to maintain social license. Data is the lifeblood of research. For research to remain credible, research needs to remain fundamentally ethical and research methods comprising data collection and data analysis need to be robust, transparent, repeatable, and auditable. Such methods alone cannot create credibility, but research data infrastructure design and implementation can provide a foundation for credibility by addressing these fundamental processes. Social science research has traditionally relied on data collection methods such as surveys, interviews, and ethnographic observations. However, an increasing proportion of human life is being mediated by online platforms, with approximately 2.3 billion active users on Facebook and 326 million active users on Twitter (Statista 2019). Social media data collection and analysis have become imperative for researchers interested in various phenomena playing out in these new media. This paper discusses the current state and issues of social media data collection and describes the Digital Observatory’s approach to establishing a credible and trusted research data infrastructure.