Misleading Advertising – Data Analytics in Greenwashing Litigation 

Greenwashing is a form of misrepresentation where the at-issue claim refers to the environmental characteristics of an organisation and/or its products or services. These are claims that are false, misleading, vague, unverifiable, unqualified, or exaggerated. Greenwashing has attracted attention in recent years as both consumers and organisations have become more environmentally conscious. While much has been written about greenwashing, this article¹ examines greenwashing litigation through a data analytics lens. Specifically, it focuses on how analytical methods, ranging from traditional surveys to Artificial Intelligence (AI), can be deployed to address the core issues of consumer perception and decision making. 

Purchasing Process Framework 

A useful starting point for addressing the role of data analytics in greenwashing claims is to view the consumer purchasing decision as a process. While there are different frameworks for thinking about the consumer purchasing decision, many include the following steps: information gathering, information consideration, purchasing decision, and post purchase evaluation. Some steps may be more important for some types of products than for others (e.g., the purchase of a car will likely trigger more information gathering and consideration than the purchase of salt). Similarly, consumers may differ with respect to their familiarity of a product (or product category) and how they perceive information. Many companies also actively engage with their customers and capture information at different stages of the purchasing process. These initial insights provide clues for what types of data may be available and what types of data may be pertinent to the misrepresentation claims. 

Claim Meaning 

Data analytics can help assess the meaning and accuracy of an environmental claim. These claims can be explicit or implicit in nature. An explicit environmental claim is a statement that can be tested. These types of claims can be verified using a variety of data sources and methods. For example, consider a claim that Company X has a ‘net zero emissions target’. Internal information such as strategic plans, capital expenditure budgets, supplier contracts, policies and procedures, and internal communications can be used to assess the timing or credibility of the claim. Publicly available data such as production plans and lobbying activities can also be examined to authenticate or refute the claim. Identifying inconsistencies using publicly available data has become more efficient due to advances in AI and Natural Language Processing (NLP), a sub-field of AI, that can obtain and analyse data rapidly. Finally, forensic and statistical tools can be applied to technical explicit claims or claims that have accompanying scientific evidence. 

An implicit environmental claim is one where the organisation’s message could be interpreted in different ways. Indeed, there can be large disparities in what consumers understand about a claim. The implicit claim may stem from words, or combinations of words, that do not have a widely accepted or single meaning, or a combination of pictures that convey a certain environmental image. Meaning can be revealed using a variety of data analytics techniques. For example, consumers can be surveyed to determine the range of perceptions associated with a product’s advertising. Meaning can also be revealed by analysing the design, images, and positioning of the at-issue product or service relative to other reference products or services (e.g., those that market themselves as ‘natural’). Content analysis can be used to identify a word’s association with other words, and algorithms can assess the similarity between the at-issue product’s image and the image of other reference products. 

Claim Importance 

Data analytics can help reveal the importance of the claim to consumers. This involves gauging consumers’ awareness of the claim as well as consumers’ perceptions about the claim. Traditional tools such as surveys can be designed to measure the importance of particular product or service characteristics. For example, conjoint analysis is a type of survey design in which participants make a series of hypothetical product selections that can reveal the relative importance of product or service characteristics. In recent years, the emergence of consumer relationship platforms and social media have opened up new approaches for gauging consumer perceptions. For example, consumers’ social media posts can be analysed using NLP to identify the importance of product or service topics as well as to assess the reactions to specific claims made by an organisation. In addition, online marketing metrics that capture information on traffic, click through, sessions, impressions, and social engagement (among others) can provide valuable insights into what information consumers considered throughout the purchasing process. 

Claim Causality 

Several data analytics approaches can help infer the purchasing effects of the claim. These include indirect and direct approaches. Indirect approaches target the elimination of alternative (i.e., non-claim) drivers of purchasing decisions. For example, consider a consumer’s brand loyalty as an alternative explanation for why a purchase was made. Consumers who are loyal may be less likely to alter their purchasing decision based on a specific advertising message. Moreover, brand loyal consumers may be more likely to recommend the brand to a friend (who, having received the recommendation, may be less likely to participate in an extended purchasing process). By analysing consumer transaction data for recency, frequency, and value, a brand loyalty score can be created for each consumer and for the overall customer base of the organisation (e.g., a consumer who purchased only yesterday, purchases on average five times per month, and purchases high value items could be considered a relatively loyal customer). Direct approaches include combining claim specific marketing information with statistical methods to infer a causal link between the claim and the consumer purchasing decision. Temporal approaches include relating historical sales transaction data to the timestamps of advertising events or paid social media posts. Another temporal approach, which focuses on consumer engagement, involves analysing the change in counts of consumer-generated content before, during, and after the claim. 

Sentiment based approaches focus on analysing consumer posts and product reviews including what information was specifically discussed about a product or service. In some situations, it may also be possible to obtain or infer the attributes of the consumers who posted the reviews, which may indicate how the claims may have influenced different groups of consumers. More advanced techniques can be considered in some situations where the appropriate data is available. For example, market response models examine the relationship between performance measures such as sales or market share and controllable variables such as price and promotional activities. In addition, some organisations’ marketing systems and marketing campaigns may lend themselves to causal inference methods. For instance, where organisations launch different advertising campaigns across different regions or engage in personalised advertising campaigns, these differences may be exploited using econometric techniques to infer the impact of a claim. 

Conclusion 

Data analytics techniques can address the accuracy, meaning, and effects of advertising messages in greenwashing litigation. While traditional methods such as consumer surveys remain an important analytical tool in misleading advertising claims, granular and contextual consumer data sources combined with statistical techniques have created new sources of potential evidence in these matters.

1 This article addresses ‘greenwashing’ claims that are conveyed through an organisation’s advertising messages. It focuses on liability issues and does not address the role of data analytics in estimating damages (that will be the subject of a future article). 

Disclaimer: The content of this article is general in nature and is presented for informative purposes. It is not intended to constitute tax, financial or legal advice, whether general or personal nor is it intended to imply any recommendation or opinion about a financial or legal product. It does not take into consideration your personal situation and may not be relevant to circumstances. Before taking any action, consider your own particular circumstances and seek professional advice. This content is protected by copyright laws and various other intellectual property laws. It is not to be modified, reproduced or republished without prior written consent.