The study of the adoption of complex social behaviors has largely been studied using cognitive behavior theories (Rosenstock, 1974; Ajzen, 1991). The assumption underpinning such models is social marketers can “think consumers” into action by controlling cognitive factors. Recent evidence suggests being better informed does not necessarily mean consumers will “do the right thing” (Claudy et al., 2013). There are factors outside of those that are “controllable”, such as emotions (positive and negative) and cultural and social forces, that influence human desire and aspirations and also combine to influence consumers’ social behaviors (Parkinson et al., ۲۰۱۶). This wider perspective of social behavior change is the one we undertake to study complex social behavior. We extend behavior change research by exploring the role of other factors, such as emotion and personal experience, to consider how experience with a complex behavior may vary between consumers. Specifically, we use an extended version of the theory of planned behavior (TPB) which incorporates emotions and past behaviors: the model of goaldirected behavior (MGB) (Perugini and Bagozzi, 2001). Social marketing has been heavily influenced by psychosocial models of planned behavior which attempt to predict behavior change (Previte et al., 2015). Since the 1980s, TPB (Ajzen, 1991) has dominated social marketers’ exploration and explication of health and social behaviors (Luca and Suggs, 2013). This approach privileges intentions as the proximal cause of behavior and therefore implies any changes in behavior can be achieved by changing intentions. This has led to a substantial body of social marketing research where the dependent variable is intentions, as evidenced in research on alcohol use (RundleThiele et al., 2015) and exercise (Courneya et al., 2000). The problem with this approach is the conflicting evidence reported on the relationship between intentions and behavior. Some studies show a strong positive relationship (Guo et al., 2015) and others show little or no relationship (Holdershaw et al., 2011). A meta-analytical review of 47 experimental tests showed medium–large changes in intentions lead to a small–medium change in behavior (Webb and Sheeran, 2006). Taken together, this indicates assuming intentions influence behaviors can be risky and in some cases incorrect.