Google and Microsoft took what initially appeared to be an innovative path to decrease residential energy consumption. Believing that customer’s decisions to conserve electricity were impeded solely by the lack of easily-understood, real-time consumption information unobtainable via paper utility bills, both created web-based analytics platforms, displaying consumption costs in real-time to help make conservation decisions. And both companies canceled the projects. Their tools had absolutely no effect on behavior.
According to a recent article in the Wall Street Journal, oPower programs yielded $234 million in energy savings last year, removing 1900 gWh from the electrical energy grid, enough energy to power 190,000 homes for a year. How did oPower succeed when goliaths failed?
The overall goal is to decrease energy usage. Google & Microsoft solved a data problem with their belief that people naturally want to conserve energy but are prevented from doing so by a lack of insight. Their failure disproves their hypothesis. oPower solved both a context and motivation problem using both descriptive and injunctive norms. oPower partnered with public utility companies to present via monthly bills a comparison of recent historical energy consumption with nearest two or three neighbors. Because individuals measure their own behavior against their perception of peer norms, consumption data in context with neighbors or peers can change behaviors. This is an example of a descriptive social norm. Those consuming greater energy than their neighbors began to conserve energy. But as is common with descriptive norms, it had a boomerang effect on those consuming less energy that subsequently relaxed conservation efforts and quickly climbed to the average. oPower added an injunctive norm wherein they added a smiley (☺) or frowney (☹) face to the descriptive comparisons, representing approval or disapproval of their positioning relative to their neighbors. The addition of this judgment is a form of operant conditioning, which is a powerful driver of behavior.
Google, Microsoft, and oPower all provided insight into energy consumption. Their assumptions were vastly different. Google and Microsoft incorrectly believed that people inherently desired to conserve energy and insight would enable behavior change. oPower correctly assumed that social pressure was a far more effective means of shaping customer behavior. Research has shown repeatedly that people desire to conform to social norms. Other research has shown they overestimate the prevalence of undesirable behavior and use these perceptions as standards for comparison. Energy customers overestimate their neighbor’s energy consumption and conserve when faced with descriptive norms showing otherwise coupled with assessments of desirability of behavior.
How can you use descriptive norms to change behaviors? One chief customer officer (CCO) showed a B2B customer how often they were calling customer support in comparison with other customers and as a result were decidedly unprofitable. The number of support requests tapered dramatically thereafter. Another company uses peer mediation in their customer communities to assess behavior in comparison with other players and assess penalties in an effort to root out toxic behavior.
What other applications of these principles have you found?