What’s next: the past becomes the future

As my last post indicated, I left Blue Apron in October 2018.

When I was leaving the company, several people asked me if I was taking time off to reflect and consider my next steps. The simple answer was, “I’m not.” While I’m envious of the world travels of friends and former colleagues as they unfold on Instagram, self-knowledge is key. “Time off” isn’t really my thing — and, with a characteristically strong sense of urgency, I was far too excited to bring a 7-year old passion project to life.

So what’s next? A renewed focus on my dream to build technology that makes it easier for people to find products that fit their values and lifestyle. To help people better themselves and the world around them, one purchasing decision at a time. To provide consumers with a simple way to shop smart — no matter where they are, and what they care about.

I know, I know. It’s a big idea, and a lot for a first-time founder to take on. Where did the idea come from, and why am I so passionate about this path forward?

To answer that question, you’d have to go all the way back to my dissertation research at Penn, which began in spring 2011. In the first 3 years of my PhD program, I’d focused on how the rise of partisan news impacted different generations. The more I learned, the more depressing the topic became: political trust, engagement, and participation were plummeting for young people, and partisan news was making things worse. In fact, there were only two bright spots in the data about Millennials: increased rates of volunteerism, and greater confidence that consumer decisions could positively shape the world around us. I chose to focus on the latter, and spent the next two years studying responses to corporate social responsibility (CSR) initiatives, and the emerging trend of “socially conscious” consumer behavior.

Initially, I was skeptical. I believed the stereotype that cause-related products only appealed to affluent liberals, and suspected that the primary motivation for buying such products was “conspicuous consumption” rather than genuinely altruistic behavior.

Study by study, these assumptions were disproven.

First, I reviewed the findings of ~300 quantitative studies in the academic literature that were either tangentially or directly related to the topic. (No one said a PhD was fun. This part certainly was not.) I was surprised to find that, across disciplines as diverse as marketing and environmental science, the vast majority of studies found that consumers were willing to buy cause-related products — even when these products were more expensive. I’ll never forgot one study in particular: an experiment in rural Guatemala which found that impoverished farmers were willing to pay more for legal firewood when given the opportunity. These findings contradicted the assumption that only the developed world and the wealthy could afford to take into consideration the sustainability, supply chain, and origins of products. It impacted me on a profound level.

But although these studies showed demand for cause-related products across a wide variety of causes and categories, there was mixed evidence about whether demographics could predict willingness-to-pay. This inspired my first experiment, which borrowed from the well-documented finding in political psychology that people only care about a handful of issues (thus demographics are too coarse to predict the behavior).

My dissertation experiments confirmed this hypothesis: the reason why demographics were a spotty indicator of interest in cause-related products was because political attitudes vary within demographic groups, and most people only cared about handful of issues. Targeting is key. People who buy Fair Trade don’t necessarily care about cruelty-free products. People who buy cruelty-free products aren’t necessarily interested in organic. People who buy organic aren’t necessarily interested in Made in the USA — and so on, and so forth. Logically, if you find out what people care about and then target them with products that check that box, consumers are willing to pay more. But if you don’t segment consumers appropriately, the effect gets washed out.

Having established that people were willing-to-pay for cause-related products that fit their values and lifestyle, my next study focused on how to persuade people to contribute to the social good by paying more for such products. Because a handful of studies in psychology suggested that social identity plays an important role in “pro-social consumer behavior” (translation: “doing good”), I wanted to test whether highlighting support for a cause — “hey you, the environmentalist, buy this product!” vs. “hey you! buy this product!” — could close a 15% price gap for consumer products. It worked.

In retrospect, my learnings seem obvious: it took two years to prove that targeted marketing for cause-related products works. But it’s the level of rigor required to get past a PhD dissertation committee that underpins my decision to double-down on the idea for a start-up: there is no doubt in my mind that demand for this information about where products come from and how they are made is broad, and growing.

Next up: how the lessons I learned along the way at Google, YouTube, Blue Apron, and Curalate will inform the product — and business — I’m building to address this consumer need. Coming soon: stay tuned.

For those who are interested, I’ve included the conclusion to my 163-page dissertation below. What’s striking is how little has changed since I wrote these paragraphs in 2013. That’s why I’m taking the leap.

Conclusions

The findings of this dissertation have a number of implications for researchers, marketers, and even citizens themselves. While nothing could seem more mundane than standing in line at the checkout counter, my dissertation has shown that at least some of these decisions are deeply political. Particularly as rates of participation in traditional forms of politics (such as voting or contacting a government official) decline, it is important to note that new behaviors have sprung up. In contrast to common stereotypes, my research shows that political participation via socially conscious consumerism is not constrained to wealthy, liberal, educated, and/or young individuals. Although demographic factors such as those I have listed may influence the issues that a person cares about (as well as the degree to which they care), they by no means preclude engaging in this alternative venue for political participation. Rather, my research demonstrates that socially conscious consumers are heterogeneous rather than uniform in both their attitudes and behavior, reflecting the diverse (and sometimes oppositional) nature of political life in the United States.

Just as the ways in which Americans engage in political life are changing, so has the context in which consumer decisions are made. In the past decades, researchers and marketers have become much more savvy about how to use technology to target consumers and/or citizens. Some, such as Turow (2012), have raised concerns about the normative implications and potential inequalities of targeting content based on consumers’ background characteristics. According to Turow (2012), the “rhetoric of consumer power” has been replaced by the “rhetoric of esoteric technological and statistical knowledge that supports the practice of social discrimination through profiling” (p. 3). One could argue that my dissertation research makes a clear case for the latter rather than the former: that researchers and marketers will maximize their efforts by targeting consumers with prosocial products based on their existing attitudes. This approach is quiet likely to maximize the short-term probability that consumers will purchase products that are consistent with their values and support causes they believe in.

However, it bears consideration that the experiments employed in my studies (as well as the statistical models evaluating their results) evaluate consumer behavior at a single moment in time. My research is not longitudinal, and cannot evaluate the long-term effects of this sort of targeting. There is no reason to believe that socially conscious consumption and the attitudes motivating it are stable, and/or stagnant — for example, that caring about one issue does not, over time, open to the door to caring about other issues. This is supported by the qualitative data collected at the end of my studies. For example, one respondent noted: “I hope that I do not unknowingly purchase products that have been animal tested or produced in sweat shops. I feel that there should be more labels showing that fact on products like the ones you showed here so that the public can make an informed choice.”

In other words, my findings should not be interpreted too literally, because that neglects the possible long-term benefits of exposing people to different kinds of products. While it is fascinating to note that normative appeals emphasizing a person’s sociopolitical identity can close the gap created by higher cost, it is equally important to remember that even individuals who say that they don’t care about an issue have more than a 70% probability of choosing a socially conscious product if it is of equal cost as a generic alternative. Over time, such small choices (such as seeing a new kind of socially conscious product, or buying it when it is on sale) may serve as a gateway to attitude change that in turn expands the universe of issues a person supports. This inference is supported by two recent studies. One indicates that the first experience of buying an organic product serves as an “icebreaker” for buying other socially conscious products, such as locally grown products or other organic products (Gottschalk & Leistner, 2013). The other indicates that the behaviors captured in experiment settings fail to capture the dynamics pattern of market data over time — — namely, that adaptation and “learning experiences” shape a person’s long-term proclivity to buy socially conscious products (Araña & León, 2013).

Just as it important to remember that my findings depict choices made at a single point in time, it is also important to be mindful that my research agenda does not go so far as to evaluate the veracity of such claims. For instance, examples of “green-washing” — the use of deceptive marketing to promote the perception that a product or company is environmentally friendly — date back to the mid 1980’s. In fact, throughout the course of my research, a number of respondents raised similar concerns in the open-ended comments collected after each study. Echoing this sentiment, one respondent wrote: “The best products are those which clearly state their company philosophies and practices. Too many large corporations are trying to jump on the ‘Green Bandwagon’ and advertise their products as being ‘natural’ or ‘green’ and yet their corporate practices are horrible and still focused on only the bottom line.” Indeed, although using hypothetical products in my research eliminated the potential confound of branding, a byproduct of this approach is that participants were presented with non-verifiable claims. As Hassan, Shaw, and Shiu (2013) demonstrate through a series of interviews and focus groups, consumers are not unaware of this potential wrinkle in socially conscious consumer behavior. In fact, subjects indicated that the complexity, ambiguity, and lack of credibility of some marketing claims causes them to delay their purchase decisions. By this account, current estimates of people’s proclivity to engage in socially conscious consumption may be too low.

Although few companies actually go so far as to make deceptive marketing claims, many brands have raised the point that it is difficult of keeping track of supply chains in a globalized economy. After a factory fire killed 1,127 workers in a Bangladesh factory in April 2013, public outcry led to a number of consumer petitions demanding companies to take a more proactive role in ensuring safe working conditions for factory workers in third world countries (Greenhouse, 2013). However, large retailers such as Gap, Target, J.C. Penney, and Wal-Mart declined to participate due to the threat of litigation by labor groups, stating that “supply chain matters” are “appropriately left to retailers, suppliers and government” (Wal-Mart, 2013). This point is not dissimilar to the statements made by Apple following a scandal regarding the working conditions at a major iPhone supplier in China, Foxconn — and perhaps the reason one respondent noted, “I cannot bring myself to believe the ‘non-sweatshop conditions’ statement” in Study 1.

In the wake of such controversies, brands and consumer groups have adopted a number of approaches to attempt to provide greater transparency regarding global business practices. The Sustainable Apparel Coalition (which includes brands such as Nike, Walmart, Gap, J. C. Penney and Target) is reportedly incorporating social and labor measurements into an existing measure of environmental sustainability, the Higg Index. However, this metric would initially be used for internal business purposes, and it is unclear when or if would be made available (and interpretable) to the public (Clifford, 2013). In a similar effort, clothiers such as Everlane have increased transparency surrounding their labor practices. Perhaps the most novel attempt to track the supply chain is by a smartphone app called “Buycott” that allows consumers to scan product bar codes to determine if a product comes from objectionable origins. However, due to complexity of supply chains in a global economy, such automated approaches to “tracing a product’s ownership back to its top parent company” cannot guarantee that ethical standards of production have been met. At best, automated approaches to evaluating the socially conscious characteristics of products can be seen as making probabilistic forecasts that are dependent on the availability of information as well as the quality of that information. In other words, although software can enhance human judgment when it comes to differentiating between products, the possibility of false positives (products that falsely appear to meet ethical standards) and false negatives (products that actually meet ethical standards but are not classified as such) remains.

In their current state, efforts such as the “Buycott” app add an additional burden to consumers with already limited time and monetary resources. Software which attempts to simplify the “matching” process between issue importance and product availability by requiring users to scan each item that they consider purchasing (after specifying each cause they support or oppose) takes additional time and effort that shoppers may not possess. For example, one of the participants in Study 1 noted: “I would like to buy all American, but usually I am in such a hurry to get through the store that I do not have time to read the labels that well. Companies should advertise on their commercials that they are made in the USA.” Unsurprisingly, multiple respondents to the survey also pointed to income as a prohibitive factor, noting that they “sadly lack ‘the coinage of convictions’” or are “financially strapped.” Another respondent identifying herself as a single mother poignantly wrote: “I do 50-cent surveys on Amazon just so I can afford toilet paper. Choosing to buy more expensive products is a luxury for the rich. I cannot imagine anyone taking your surveys has a lot of extra discretionary income.” This latter comment — as well as the corroborating statistic showing that roughly 37% of each sample earned less than $25,000 per annum in each of my studies — suggests that my studies offer a strong test of IIPM’s influence on consumer behavior. It also bears consideration that my studies investigate the politics of relatively inexpensive products — in other words, my research may elucidate the politics of the checkout counter, but it does not necessarily translate into the politics of the parking lot, high-end handbags, or household appliances.

Comments such as those above highlight the difficulties still facing socially conscious consumption, but the results from my dissertation studies offer evidence of its potential. My results indicate not only that the topic is ripe for further inquiry by researchers, but a promising course of action for practitioners — for better or for worse. I will leave it to the political philosophers to debate whether or not “politics at the checkout counter” enriches or detracts from civic life, and whether such behavior is sufficient replacement (or complement) to more traditional forms of political participation. Rather, I argue that the increasing prevalence of socially conscious consumption, coupled with citizens’ tendency to see such behavior as a political act, merits serious attention from scholars of political communication, political behavior, and public opinion. Notably, only a handful of national public opinion surveys in the United States and Europe have bothered to ask respondents about their views on the political significance of their purchasing behavior — and, as my review of the literature has shown, findings from analyses using these overly generic measures are not always consistent with the larger body of research. If scholars are to take socially conscious consumption as seriously as the general public does, it is thus imperative for us ask more questions — and better questions — about this emergent form of non-traditional political behavior in national public opinion surveys.

Finally, my dissertation suggests that it is time for scholars, practitioners, and the public to take a long look at our affection for the American tradition of boycotting, and question whether or not there may be a more fruitful avenue for social change. While in some cases, boycotting successfully polices corporate behavior, it does little to improve the processes of global consumerism entwined with modern daily life. Yet as my data show, consumers are more than willing to offset the potentially higher costs of production for products that satisfy their moral druthers — and in fact, that they are even more likely to do so if you first remind them that such druthers exist. Although it is beyond the scope of this dissertation to study the response of institutions to consumer behavior such as that demonstrated in this dissertation, my studies have clearly established that consumers are both “active” (independently motivated to buy socially conscious products) as well as “reactive” (responsive to normative appeals deployed by institutions). In other words, my data support the argument that socially conscious consumption is simultaneously “bottom-up” and “top-down.” Favoring the consumer carrot over the stick may not only feel better for consumers– it may also do more for our global society.