Off Track: How Broken Streaks Decrease the Likelihood of Continued Behavior

Marketing Department; Faculty Advisers: Deborah Small, Alixandra Barasch

We examine how streaks of consecutive behaviors (and whether they are broken or remain unbroken) affect people’s subsequent decisions to continue those behaviors in the future. In six studies, we show that a broken streak makes an individual less likely to continue a behavior, even when that broken streak is caused by external events outside of their control. This effect occurs because broken streaks disrupt the perceptions of momentum people feel when engaging in repeated behaviors, making them consider subsequent decisions less automatically and more carefully. Making streaks more salient motivates people to preserve them when they remain intact, but magnifies the negative effects of broken streaks. Consistent with the perceived momentum mechanism, these effects persist regardless of streak length. However, the negative effects of broken streaks diminish with greater temporal distance from the break, allowing consumers to restore some momentum through new behaviors. These findings provide insight into the trade-offs involved in motivating repeated consumption behaviors by increasing the salience of consumers’ streaks.

Hot Streak! Consumer Inferences and Predictions about Streaks of Virtuous Choice

Marketing Department; Faculty Advisers: Deborah Small, Alixandra Barasch

We explore when people make optimistic forecasts about goal-directed (i.e., virtuous) behavior, and when they are pessimistic that such behaviors will sustain. In four studies, we examine how predictions regarding an individual’s likelihood of choosing virtue are affected by that individual’s recent pattern of behavior. Specifically, we show that even when the overall rate of behavior is identical, a recent streak of virtuous behavior increases the predicted likelihood that the individual will persist, compared to a variety of other patterns. This effect is driven by a higher level of perceived commitment after a recent streak, and thus only exists for predicting behaviors that require a high level of a commitment, and not for choices between other activities that do not. As further evidence that a recent streak implies commitment, we find that the effect is attenuated when other signals of commitment are available (e.g., the individual has a high overall rate of choosing virtue). Understanding these predictions is important because they can affect whether consumers pursue more restrictive and effective versus more laissez-faire strategies to help themselves or others reach their goals.