Having developed an understanding of the communication landscape, including messages and resources currently available to the campaign’s audience, our next step was to develop and test concepts for the campaign. Each campaign concept represented a unifying description of how the different elements of the campaign (e.g., audience, messaging, visuals, tone) would potentially work together to achieve the campaign’s communication goals or desired behaviors. As a starting place in developing campaign concepts, our team developed a creative brief outlining the campaign’s audiences, objectives, including desired protective behaviors, key messaging requirements, and potential communication channels and materials, as well as key findings identified in the environmental scan.

With the parameters of the creative brief in mind, our team brainstormed approximately 20 concepts for the campaign In this initial stage, the concepts were little more than outlines, each identifying an intended audience (e.g., pregnant women, partners, community members), a desired emotion, sample message statement, and an image to represent the potential feeling associated with each message. Development of these initial concepts also involved the explicit incorporation of one or more constructs from health behavior theory, such as audience motivation (Andreasen, 2002); attitude, beliefs, and norms toward prevention behaviors (Ajzen, 2002); self-efficacy, and perceived behavioral control (Bandura, 1998; Maibach, Flora, & Nass, 1991); response efficacy (Beck & Lund, 1981); and social support (Kaplan, Cassel, & Gore, 1977).

To reduce the initial set of 20 concepts to a number reasonable for audience testing, we worked with members of CDC’s Zika communication team to combine, refine, and reduce the draft concepts. Key to this process was considering the existing formative research with pregnant women in Puerto Rico on preventing Zika and what it suggested about motivators and barriers to pregnant women adopting Zika preventive behaviors, as well as what had been learned about the Zika situation Puerto Rico from the environmental scan. For example, social media posts and editorials questioning the seriousness of the Zika virus threat in Puerto Rico suggested the need to test a concept that aimed to increase pregnant women’s perceived risk about the threat of Zika virus. We theorized the following conditions as important to motivating preventive behaviors:

Through this process, we reduced the number of concepts for testing to four. Among this set of four concepts, two of the final concepts shared a theoretical driver and similar copy, although they featured different headlines, taglines, and image. The inclusion of these two similar concepts allowed testing particular nuances of one messaging approach; we would alternate showing one or the other of these concepts in testing. Table 2a-d provides an overview of the four concepts developed for testing.

To test the concepts with audiences, we placed each of the four concepts on a poster-sized board, with each board displaying a concept title, supporting copy, tagline, and a representative photo or image. The presentation of concepts in this simplified form was in line with recommended practices (Lefebvre, 2013) and serves to keep the conversation at a more conceptual level. In addition to the concept boards, we also created smaller displays containing additional images and taglines to explore in relation to each concept.

We conducted the concept testing interviews at six WIC clinics in northeastern Puerto Rico, in and around the city of San Juan, and completed 82 interviews over 4 days in April 2016 with pregnant women (n = 47), partners (n = 10), and family members (n = 25). Participants (pregnant women and their family and friends) were a convenience sample of individuals presenting for services on the day the research team visited the WIC clinic. All clinic entrants were given the opportunity to participate in the research. Each interview lasted 30 to 40 minutes and was conducted in Spanish. Participants received $15 for taking part in the interviews. RTFs Institutional Review Board reviewed all research procedures before research activities began.

In testing the concepts, we briefed participants on the purpose of the data collection activities, explored participants’ awareness of the Zika virus, knowledge of protective behaviors, and sources of information on the Zika virus. Rotating the order between interviews, we then showed three concepts to each participant, eliciting feedback after they viewed each one, and asking their preferred concept at the end. We also asked about their experiences with different Zika prevention behaviors. Table 3 provides an overview of the interview process.

All of the interviews were recorded and transcribed. However, with the speed of the campaign development, in analyzing he interviews, we ended up relaying primarily on the notes taken in each interview by a dedicated notetaker. Using the notes, we used a matrix approach, as advocated by Miles and Huberman (1994), to analyze the data from the concept testing interviews, whereby we created an interview question (columns) by respondent (rows) matrix and input the participant’s response in the corresponding cells. This process allowed us to review and compare responses to individual questions, as well as sort by type of participant. In reviewing responses to individual questions, we also followed a constant comparative approach to first look for similarities or themes in the responses to identify outliers or differing views. We also noted substantive quotes that provide examples of views or common ideas shared in the participant’s responses. Simple descriptive statistics were used to characterize the participants (see Table 4).

Pregnant women, the campaign’s primary audience, said they were most motivated to prevent Zika virus by Zika Can Take a Bite Out of Your Life, and Zika Can Bring You Down. They liked that the messaging spoke to the health of the pregnant woman and the baby. For Zika Can Take a Bite Out, pregnant women liked that the concept focused on the health of their baby and they identified with the image of the worried pregnant woman. While women shared generally positive feedback on messages in Zika Can Bring You Down, they were less comfortable with its more somber tone.

Partners, family, and friends who saw the Zika Can Take a Bite Out of Your Life found it to be the most personally motivating, while those who saw the Zika Can Bring You Down concept instead preferred You’ve Got the Power to Combat Zika. Participant liked the message of unity and strength in numbers in You’ve Got the Power to Combat Zika; all participants agreed that it would motivate the broader community the most because cleaning the environment of potential breeding grounds for mosquitoes requires a group effort.

More generally, participants recommended including images of pregnant women, babies, and the community to motivate the community to take action together, and so that pregnant women would not feel like they were facing the threat of Zika virus alone. Participants also recommended including a clear call to action in the messaging, as well as concrete information about prevention measures.

Other key information we learned from the concept testing, which echoed information shared already in CDC’s formative research with pregnant women on barriers to Zika prevention, was that women were concerned about having all the burden of Zika virus prevention put on them, especially during the already stressful time of pregnancy. They suggested needing support from family and friends and their community. Also, the individuals we interviewed were interested in knowing more about how Zika virus was different than other mosquito-borne diseases in Puerto Rico, especially given emerging information at that time indicating that Zika virus could also be sexually transmitted (Oster et al., 2016).

To apply the findings from the concept testing in developing campaign materials, we next updated the creative brief with the findings and developed a single campaign concept that would be used to guide draft messages and materials. The brief also included a brand description and draft headlines, taglines, and call to action, as well as a description of a desired visual tone and elements.

Following the findings from the concept testing, the brief emphasized a pregnant woman, with the need to protect her and her baby from the Zika virus, as the central theme in text and imagery, while also incorporating images of community and the collective aspect of Zika prevention activities. Our draft headline became, “Así Prevenimos el Zika,” or “This is how we prevent Zika,” while the tagline was “Hazlo por ellos” (“Do it for them”), and the call to action was “Únete” (“Join Us”).

Working with a local creative firm in Puerto Rico, we developed draft campaign messages for testing, three variations of a single print ad, each featuring variations on the draft headline text, a campaign logo, with the statement “Detén El Zika” (“Stop Zika”) and tagline “Hazlo por los Niños” (“Do it for the babies”), and script of a video public service announcement (PSA) to test with the campaign audiences (see Figure 3).

To gauge receptivity to the draft materials among our campaign audience, we conducted three focus groups in San Juan with adults over the age of 18 years: a group of pregnant women (n = 11), a group of male partners (n = 9), and a group with members of the community (n = 11) (See Table 4 for participant characteristics). Participants in the focus groups were a convenience sample recruited using a marketing research firm and screened for key campaign audience characteristics (e.g. add screener info).

In all three groups, we asked participants about their basic knowledge of Zika virus prevention (e.g., What have you heard about what you can do around your home or in your community to prevent Zika?) and then explored their opinions on the campaign concept, messages, and prototype campaign materials (one PSA script, three posters, and two logos) to ensure the message and creative executions resonated with the target audience groups (e.g., What do you think is the main idea or message behind this logo? Does the ad make you think or feel differently about protecting yourself from Zika?). Bilingual staff conducted the sessions in Spanish. Focus groups lasted between 60 to 90 minutes and participants received an incentive ($60) for traveling to and participating in a longer interview.

To analyze the data from the focus groups, we summarized reactions to the three materials tested using the same matrix method for concept testing, noting differences according to the audience. Simple descriptive statistics were used to describe participants’ characteristics. We developed recommendations from the focus group findings to revise the final campaign materials.