Health care organizations and businesses, which spend billions of dollars each year on advertising, should adapt their message to their specific audiences, according to a study conducted by Tuck School of Business professor Punam Keller and visiting Tuck professor Donald Lehmann. The study, "Designing Health Communications: A Meta-Analysis," was recently published in the Journal of Public Policy and Marketing.
The study found that health advertisements targeted to a specific audience -- based on factors like race, age and gender -- and then tailored to that segment of the population were most effective. For example, Keller found that younger audiences are more affected by advertisements that show "social consequences," while older audiences are more influenced by communications depicting "physical consequences."
"When you're younger, you're much more affected by peers and driven by wanting to be part of a group than when you're older," Keller said. "Physical consequences are much more salient to the elderly population because they are more likely to have a lot of physical problems."
The study also considered the effectiveness of different advertising tactics, including the use of fear and hiring an actor of a particular gender.
Using pictures and stories in advertisements was not found to be effective for health communications, Keller said, although it can be very effective for other kinds of advertisements.
"We found that it arouses such a high level of fear that people engage in what we call 'seizing and freezing,' and then they don't want to look at or do anything else," she said.
The study took three years to complete, including one year spent building a database of previous research data from psychology, communication, marketing and health.
Based on the compiled research, Keller and her team developed a model to help health organizations shape their future advertisements.
"Once you have the database, then you develop the model and finally you test the model," she said.
The model can be applied two ways, Keller said.
"If you [already] have [an advertising] campaign, the model tells you what target audience it will be most effective for," she said. "But you can also go the other way around. That is, if you have a target audience, the model will help you design an effective campaign for that specific group."
Dartmouth organizations have already contacted Keller about how they can use the results of her study to benefit students. Brian Bowden, coordinator of the Alcohol and Other Drug Education Program at the College, recently asked Keller for advice about how to promote the College's Good Samaritan policy more effectively, Keller said.
"The goal is for more students to make use of the program and call when their friends are in trouble," Keller said. "We want to look into ways to revamp the program's communication strategies."
Keller also presented her study to a governmental health organization, which subsequently requested her assistance in assessing their campaign to encourage young children to exercise more often.
"I coded their ads and predicted which ones would be more or less effective for certain target audiences," Keller said. "[The organization] measured kids' exercise activities before and after the ad campaigns, which helped me get additional data to validate the model. It turns out we did a really good job predicting which [commercials] work."
Keller said she plans to present some of the findings from her analysis of the advertising campaign to the organization in a few weeks.
"What I'm hoping for the future is that they'll give me data on another problem, such as [HIV prevention campaigns], so we can see how robust the model is," she said. "The second thing I'm hoping for is to make the model more user friendly for health practitioners."
Keller said she decided to study health communication because she wanted to use her expertise in business and marketing to make a social impact.
"The thing that drove me to write this paper is the idea of accountability and return on investment -- the idea that you don't just educate people or make them aware of something, but you have to see whether your techniques are actually changing people's intentions and actions," she said. "If you're spending money on an initiative, it makes sense to see whether or not it's having some sort of effect."
Tuck professor Koen Pauwels said in an e-mail that Keller's research should have a significant impact on health advertising.
"Keller and Lehmann did a great job in distilling several concrete rules from the wealth of academic research on the topic," Pauwels wrote.
Pauwels added that he hopes future research will examine how health communication and advertising can influence long-term behavioral changes.
"For instance, to get kids to exercise more for the rest of their lives, how much and which advertising messages are needed and should they be targeted also at parents, peers and school communities?," he wrote.
Lehmann could not be reached for comment by press time.
