A review of campaign partnerships between Data for Good at Meta, UNICEF, the Yale Institute for Global Health, and the Public Good Projects
As we enter the third year of the COVID-19 pandemic, vaccines remain one of our most effective tools along with other public health measures, including mask use, handwashing, and testing, to reduce cases and deaths, as well as to help children return to school, people return to work, and communities return to normal. Although more than 5 billion people have received at least one dose of the COVID-19 vaccine worldwide, countries with weaker health systems and multiple competing health challenges have limited capacity to roll out widespread COVID-19 vaccination. These countries often have overstretched health workforces, limited investment in risk communication and community engagement, and challenges with vaccine hesitancy in some communities.
Over the last year, Data for Good at Meta has been collaborating with UNICEF, the Yale Institute for Global Health, and the Public Good Projects (PGP) to develop campaigns that drive COVID-19 vaccine acceptance in Kenya, India, Pakistan, and Ukraine, and that have reached more than 150 million people. In celebration of World Immunization week, we analyzed the results of these campaigns and are sharing lessons we learned about the most effective messages for public health practitioners to learn from.
In spring 2021, the Data for Good team at Meta worked to analyze public Facebook posts about the COVID-19 vaccine to understand drivers behind vaccine hesitancy being discussed on Facebook, as well as the main reasons for hesitancy from Facebook’s COVID Trends and Impact Survey (CTIS). UNICEF, the Yale Institute for Global Health, and PGP assessed these insights and subsequently produced relevant content for each country.
After launching the campaigns via the relevant UNICEF country office page, we measured the effectiveness of different messages by comparing post-campaign survey responses of people who saw the ads with those of people who did not. The survey asked respondents whether they remembered seeing the ad recently, as well as up to four other questions relating to awareness and attitudes about the COVID-19 vaccine.
Analysis of public Facebook posts revealed that people had basic questions about COVID-19 vaccine safety, efficacy, availability, and eligibility, with the specifics of these concerns varying between countries. Survey data revealed that in Kenya and Ukraine, concerns about potential side effects were the main reason for hesitancy; however, in India and Pakistan, most people indicated that they wanted to wait and potentially get a vaccine at a later date. In Ukraine, public posts also suggested that some people were interested in getting the vaccine but might need additional nudging to do so, and that many people had lingering questions about safety and efficacy.
For each country, we worked to ensure that the campaign content reflected the community’s main concerns and questions. In Ukraine, for example, the insights led to the development of two campaigns — one that provided information on eligibility and how to register for the vaccine, and a second that focused on vaccine safety and efficacy. Similarly in Kenya, one campaign focused on providing practical information to young adults stating that they were eligible and that vaccines were freely available. Apart from the information in the ads themselves, all campaigns included links to government or UNICEF resource pages where people could click to read more about the vaccine or register for an appointment.
In Ukraine, the campaign that provided information on eligibility and registration succeeded in improving the likelihood that people would advise their friends to get vaccinated, and the safety and efficacy campaign increased perceptions of vaccine importance. Similarly, in Kenya, the practical campaign was effective at improving perceptions of vaccine safety.
In India and Pakistan, content in local languages, rather than the national language of English, also led to higher percentages of people clicking on the links to vaccine resource pages provided in the ads than the English content did. This finding is helpful for settings where the assumption might be that online populations will be more likely to engage in English content despite there being large populations both online and offline that may be more attracted to content featured in languages such as Hindi and Urdu.
Existing behavioral science research has found a significant association between vaccine hesitancy and the upholding of certain moral values such as personal liberty and a belief in purity. Meanwhile, prior research suggests that leveraging local social norms such as collectivism and equity can be powerful tools in influencing people’s likelihood to vaccinate.
In our public post analysis, Data for Good at Meta found discussion of moral values and social norms were common across countries. For example, when posting about vaccines in Ukraine, many people emphasized the right to choose whether to vaccinate, suggesting that liberty is a core value of many Ukranians. In India, however, many posts expressed pride that India had produced its own vaccine and was distributing it to other countries at the time. Meanwhile, in Pakistan, public posts about the vaccine discussed concerns about inequity in vaccine access. Such values were leveraged in the messaging content and creative design.
As a result of our pre-campaign research, UNICEF Ukraine ran messages emphasizing that vaccines are a personal choice that can lead to positive outcomes for individuals and their families, which worked well for improving perceptions that it was better to get vaccinated than to risk getting COVID-19. Leveraging local insights in India, the team developed a campaign that underscored collective action and used India-focused imagery, which performed especially well at increasing the likelihood that the Hindi-speaking audience would advise a close friend or relative to get the COVID-19 vaccine. In Pakistan, insights informed the development of a campaign focused on the values of fairness and equity that achieved the strong results across several outcomes tested in Pakistan and was particularly successful in improving perceptions of vaccine importance.
Existing public health research has found approaches such as parental stories can be effective in getting people to seek more information about vaccines, and UNICEF and Yale’s Vaccine Messaging Guide outlines that using narrative techniques can be a compelling messaging strategy. In our own precampaign research, we found that many people in India shared on Facebook their personal stories of being vaccinated and encouraged others to do the same. In India, a high proportion of public posts also showed appreciation for health-care workers. Meanwhile in Pakistan, public posts revealed that many people were trying to register their parents and older relatives for a vaccine appointment, which offered an opportunity to tell locally relevant stories about young people working to protect elderly members of their families.
In India, our findings prompted the development of testimonial-style illustrative content featuring health-care workers, and in Pakistan, the resulting campaign featured everyday people expressing their eagerness to do their part and get vaccinated, as well as to register loved ones, including older relatives, to do the same. In India, the storytelling campaign was particularly effective for improving perceptions of social approval of the vaccine and the likelihood that people would recommend the vaccine to their friends. Similarly, in Pakistan, the testimonial campaign was effective for improving the likelihood that people would recommend getting the vaccine to close friends and relatives.
Across four countries and over 150 million people reached, results from our campaigns on the COVID-19 vaccine suggest that filling gaps in basic information, developing messages that speak to relevant values and social norms, and leveraging techniques like testimonials and storytelling with effective messengers can all be powerful tools in combating vaccine hesitancy.
Although these findings represent a substantial step forward in understanding how to increase vaccine acceptance through digital outreach, this research and message testing represents the first phase of this collaboration between UNICEF, the Yale Institute for Global Health, PGP, and Meta. Currently, we are conducting randomized control trials in India, Pakistan, and Ukraine to test whether these digital campaigns lead to improved vaccine confidence and higher immunization rates overall, which will further illuminate the role that digital campaigns can play in advancing global public health goals.
Special thanks to Surangani Abeyesekera, Social and Behavior Change Specialist, Demand for Immunization at UNICEF HQ, as well as Dr. Saad Omer, Sarah Christie, Amyn Malik, and Scott Bokemper from Yale Institute for Global Health for their contributions to this research.