When it comes to persuading others—whether about voting, pandemic-related safety measures, or other significant behaviors—we frequently focus on the wrong folks.
That’s the conclusion of recent research led by Zakary Tormala, a professor of marketing at Stanford Graduate School of Business, and his doctoral student Christopher Bechler.
“How do people choose their persuasion targets?” Tormala asks. “That’s the basic question we started with. And then, as a secondary question, how well do they choose?”
“There’s a large body of research on attitudes and attitude change,” Bechler says. “But we have virtually zero insight into how people perceive changes in other people’s opinions and how they choose their targets of influence.”
Bechler and Tormala theorize that people naturally aim to effect “qualitative” attitude change in others, or to shift someone’s opinion from positive to negative (or vice versa), rather than moving someone toward a stronger version of an opinion they already hold—such as voting for a particular presidential candidate or wearing a mask in public to combat COVID-19. So we target people on the other side of the fence instead of those who are already on our side but weaker in their conviction.
In separate studies focused on voting and COVID-related behaviors, Bechler and Tormala found that people do indeed target those with attitudes opposite theirs, but that the most receptive to influence tend to be people already leaning toward the persuader’s opinion and who have room to shift to an even more supportive stance. The findings have strong implications for allocation of resources during political and health campaigns.
How We Influence
To get at patterns of influence, the researchers conducted a series of experiments.
For example, during the Democratic presidential primaries, Bechler and Tormala studied people with a wide range of attitudes toward Joe Biden as the potential nominee. They started by asking participants to report their attitude toward Biden on a scale ranging from 1 (extremely against) to 9 (extremely in favor), with 5 labeled “neutral.” They then designated those expressing strong support for Biden (an 8 or 9 on the scale) as “persuaders,” and asked them to choose persuasion targets among the other participants, including those who were strongly against Biden (2 on the scale), slightly against him (4), and slightly for him (6).
“Persuaders overwhelmingly chose the targets who were slightly against Biden,” Bechler says. In other words, people generally aimed for targets with opinions that were moderately opposed to their own. They largely ignored targets with extreme opposing opinions—presumably recognizing them as lost causes—but also tended to overlook targets who already were slightly positive. “In theory, this strategy could make sense,” Tormala notes. “If you knew you could flip somebody from negative to positive, that could create meaningful behavior change.”
The problem is that targets on the other side of the fence tend to be less receptive to persuasion compared to people already leaning your way. For example, when Bechler and Tormala shared a positive news article about Biden with the different participant groups, they found that the ones who indicated the biggest shift in their own behavior (likelihood to vote for Biden in the election) were the participants slightly in favor of Biden already. The news article served to intensify their positive opinion.
The researchers followed a similar protocol when studying influence related to COVID-19 safety measures, including the wearing of face masks and the willingness to sign up for vaccine trials. The findings were similar: Persuaders tended to target those slightly across the divide, but influence worked best for those already leaning in the same direction as the persuader.
“In short, we found that people are more receptive to messages that are congruent with the way they’re already thinking,” Bechler writes in summary. “But that’s not reflected in persuaders’ targeting decisions. Persuaders are more likely to target people with views that diverge from their own.”
How We Should Influence
The research has clear practical implications, especially for the upcoming presidential election and the ongoing COVID-19 challenge.
“Lots of people are participating in postcard and letter-writing campaigns to try to have an impact on others’ voting decisions,” Tormala says. “Our data suggest if you’re trying to help a candidate, targeting voters who are already leaning your way but might need a nudge—for example, to actually vote—could be a good way to go. That audience is likely to be more receptive to your message and may increase their likelihood of voting. Targeting people leaning the other way may end up being a worse use of resources—whether it’s about how they vote or getting them to wear a mask.”
“People tend to dichotomize attitudes and behavior,” Bechler says. “They think you’re either for or against something. They forget there’s a third group who’s not voting at all, or who is leaning your way but failing to take action,” and those people are often the most target-worthy.
“It’s not necessarily the case that someone always wears a mask or doesn’t, or that someone supports or opposes everything about a candidate or nothing,” he says. “Many factors might keep someone in favor of wearing masks from always doing that, such as their peers’ behavior. Often there’s still room to increase the likelihood of a desired behavior like mask-wearing among people who are already leaning toward it.”
Beyond politics and public health, the findings apply to business, as well. “Marketers frequently overspend on customer acquisition as opposed to retention,” Bechler says. “They’re trying to get people to convert, or flip, when resources could sometimes be more efficiently allocated toward retention and increasing affinity toward your brand.”