Aim: A great deal of enthusiasm and interest exists in using social media for public health communications, but few research studies have examined its success in promoting and adopting protective health behaviours. To begin to understand how best to develop effective online social marketing campaigns, this paper provides a summary of success factors and key lessons learnt from selected social media campaign case studies.
Study type: Case study review
Methods: A selection of case studies was reviewed for lessons in campaign development, delivery and evaluation from both the corporate and public health sectors. Information about the objective of the campaign, the tactics used and the lessons learnt was extracted from each case study. Lessons learnt from across the case studies were then sorted according to themes.
Results: Lessons from the nine case studies selected were categorised into eight themes: planning, use of social media tools, community, content, personal benefits, promotion, costs and challenges. Outcome evaluation data were lacking in the case studies.
Conclusions: Overall, the nine case studies show that social media hold promise in changing user behaviours and that social media are highly effective in recruiting participants and motivating them to take small, concrete actions. The case studies also demonstrate that there is room in social media for targeted, inexpensive, small-scale projects, as well as large, well-funded, mass-reach marketing blitzes. Social media campaign process and impact evaluation measures are readily available. Outcome evaluation models and measures are needed to better assess the effectiveness of social media campaigns in changing health behaviours.
There is a great deal of enthusiasm for, and interest in, using social media for public health communications, but there is little understanding of the connection between online engagement and behaviour change. Few research studies have examined the success of social media in influencing health promoting behaviours.1 Internationally, public health organisations are keenly experimenting with using social media as tools for both information sharing and behaviour change. What remains to be seen is how best to reach out to social media users and what types of messages cut through the online clutter.
Strong evidence indicates that public health social marketing campaigns conducted through mainstream media can have a direct and positive effect on behaviour.2 Building the evidence base for conducting health promoting campaigns through social media will require both applying what is known to work in traditional media channels and developing new methods that incorporate the unique features of social media. For example, a great advantage of online media is that experimenting with communication techniques and messages could be much less resource intensive and potentially less risky than investing money in an expensive new television campaign. Small-scale experiments assessing the most appropriate message content, format, delivery and online social networking site could be potential starting points for developing campaigns.3
Despite the great potential to deliver innovative and cost-effective public health campaigns online, a limited body of research is available to help program planners develop and evaluate their campaigns. To begin to understand how best to develop effective online social marketing campaigns, this paper provides a summary of success factors and key lessons learnt from selected social media campaign case studies.
A selection of case studies was reviewed from the public health literature, as well as those from the corporate sector that were well outside the public health sphere. A case study review approach was used because of the very limited availability of quality published evaluations of social media campaigns. Additionally, given the disparate topics and goals of social media campaigns, a case study approach allowed for richer and more detailed understanding of what contributed to the success of each of the campaigns examined. A traditional, systematic keyword search within academic databases for relevant papers was not conducted, as there were exceptionally few published data of direct interest. Case studies were primarily identified through searching the online grey literature of public health agencies and marketing organisations.
Case studies were selected based on the following criteria: 1) publicly available information on objectives, strategies and results; 2) inclusion of campaign evaluation; 3) delivery through a popular social media channel; and 4) integration of social media content with other promotional activities. Analysis of the case studies included extracting information about the objective of the campaign, the tactics used and the lessons learnt. Lessons learnt were then sorted into eight key themes that emerged following case study review.
Given the commercial sensitivity of marketing information, little information was available in the public domain on budget or results, despite there being many high-profile examples of corporate social media campaigns. The limited published results also suggest the corporate sector may face the same struggles as not-for-profits and government agencies in determining how to best evaluate a digital campaign.4 When selecting case studies to include from the public health literature, it was essential to choose examples that had included some form of evaluation. As with the corporate case studies, examples of well-executed, comprehensively evaluated and publicly available reporting of digital campaigns in the health domain were limited. However, by pooling together the lessons learnt from the selection of case studies, it was possible to get a sense of the opportunities and challenges of implementing digital health campaigns. Table 1 shows a summary of the nine case studies included.
|IKEA||Engage consumers in the opening of a new IKEA store in Sweden||
|Cadbury||Generate interest in and sales of Cadbury Creme Eggs||
|Smirnoff||Revive brand image of Smirnoff||
|Burnet Institute||Trial the delivery of sexual health promotion via social networking sites to two key at-risk groups: young people aged 16−29 years and men who have sex with men||
|World Lung Foundation||Make measurable change to tobacco control policies in low- and middle-income countries6||
|New Zealand Health Sponsorship Council||Educate 12–24-year-old New Zealanders on the harms associated with tobacco use||
|Smokefree South West UK||Encourage smokers to seek National Health Service support for quit smoking attempts||
|Association of Healthy Communities, Queensland||Reinstate sexual health bus shelter ads after they were removed due to public complaints8||
|New York Alliance for Donation, Inc.||Promote organ donor registration to 18–24-year-old students solely through online media||
The effectiveness of a social media campaign can be positively affected by the use of traditional media; the two need not be viewed as separate entities.10,11 Social media campaigns that realise a viral effect often greatly benefit from news media coverage through both mainstream and online channels.8
Integration with broader campaigns seems to be useful, but is not a necessary driver of success. Stand-alone digital campaigns can achieve results. For example, when a social media campaign is highly novel or creative, it can generate a tremendous amount of earned news media (as opposed to paid advertising) coverage that further boosts the campaign.6
Campaign developers should use functions that social media users are already familiar with, such as photo tagging and retweeting. Complex third-party tools or the need for participants to register or provide personal details can severely dissuade participation and engagement.12
Participant demographic and campaign traffic data can be automatically collected through social media analytics tools. High bounce rates (the percentage of users who immediately leave a website after opening it) to external websites9 suggest that click-through ads on social media sites may work better if users are directed within the social media site pages, as opposed to being forced offsite to an external website. Including a variety of social media channels can increase participant involvement in a campaign, with the proviso that running a campaign well across a smaller number of popular social media channels makes more sense than trying to include networks that do not reach many additional participants.
When starting out in social media, partnering with organisations and people who already have a social media presence is useful for building an online community. Highly connected people and agencies that support and promote the campaign can influence others to both join and further share the campaign. Building social media communities needs to be a continual focus, even during periods of campaign inactivity, as it may be possible to capitalise on that strong community support for any future campaigns.8 Online campaign communications should routinely and explicitly ask all members and followers to help build the community, as it leads not only to more followers, but more highly engaged followers.6
Online campaigns seem to work best when there is a clear and achievable call to action, such as in the IKEA case study, where users were asked to tag themselves in photos on Facebook.12 For example, in health advocacy campaigns, supporters are asked to contact their local politician or add their name to a list of supporters. This could be because these types of campaigns are also the easiest to evaluate and show cost-effectiveness. Campaigns that seek to change social norms and have a population-level effect on behaviour have been shown to be effective in mainstream media, so it stands to reason that social media campaigns that are executed and evaluated correctly could be as effective.
Generating a large number of shares or having a campaign ‘go viral’ cannot be seen as the primary, most important outcome of a social media campaign. Developing campaigns that are shareable and that create change may mean taking calculated risks with campaign content that is suitable to the target audience.13 This could include the use of language and imagery that may be offensive to some users but may then greatly increase campaign appeal to other users. An example is the description of the New Zealand Health Sponsorship Council anti-tobacco viral game, Kiss Off, which is targeted at youth: “Prove yourself as the best kisser in the world and then send your kissing profile to your friends. Find the hotty to kiss by avoiding all the disgusting, annoying smokers. Don’t get caught with one of them or you’ll lose points with a difficult challenge”.13
Successful social media campaigns help users to feel like members of a community and establish identity by allowing participants to express part of themselves to others.14,15 Participants enjoy interacting on social media because they can be both anonymous and personal.16 Providing an opportunity to win a prize that is both relevant and desirable – not the standard iPad of many online marketing contests – increases meaningful participation.12 Offering rewards for participating and spreading marketing messages may also increase participation.17,18 These rewards can be simple and often cost-effective, such as publicly thanking people for sharing content or profiling people on a campaign social media page.
It cannot be assumed that once a social media campaign is unveiled, viral growth is guaranteed. An active, ongoing communications strategy is necessary to ensure a strong launch, maintain promotion of the campaign and drive traffic to the site.6 Recruiting large numbers of motivated volunteer ‘seeders’ who can leverage their personal social media connections with others appears to be more likely to generate action than impersonalised advertisements.9, 19
Users need to be able to easily receive and share valuable information with other people. Ideally, campaigns should spread from friend to friend through the automated sharing process of existing social network page feeds, and not require onerous further action of participants.10
Digital advertising on social media has been shown to be a very cost-effective tool for program recruitment compared with traditional mass media. For example, the digital arm of the campaign Smokefree South West UK was extremely cost-effective compared with other media in achieving responses from smokers interested in quitting.7 Although developing a successful social media campaign may be lower cost than a mass media campaign, it can be time and human resource–intensive because it may require ongoing content generation and interaction with users.5 Reducing complexity and running simple, low-tech and low-cost campaigns can be highly effective when conducted through the appropriate channels.12
Although it is possible to largely contract out social media campaign services, inhouse expertise can also be built through training. Staff responsible for managing social media campaigns require training on the breadth of social media platforms, collecting and evaluating social media analytics and developing shareable content.5
The sheer speed with which social media move is challenging for all organisations, but even more so in disciplines like public health, where government agencies make up a large proportion of stakeholders. Multi-tiered approval processes for all public communications can impede the ability to respond to and engage with social media users in a timely fashion.8 This is further complicated when considering that an organisation may ultimately be legally responsible for all content posted on its social media pages, including all user-generated comments and content. This means that diligent and timely moderation and monitoring of social media pages is necessary. Again, this emphasises the need for adequate human resources to manage campaigns.20
Internet-based programs have the potential to reach many people who are not part of the intended target audience, including those who live in other countries. This can be particularly problematic when it comes to being accountable for government funding sources that require target audiences to be local citizens.7
Collectively, these case studies reveal that social media can be highly effective in raising brand awareness, recruiting participants and motivating them to take small, concrete actions. The case studies also demonstrate that there is room in social media for both targeted, inexpensive, small-scale projects and large, well-funded, mass-reach marketing blitzes.
The lessons reported by the case studies were largely limited to campaign design, resources and process evaluations. Process evaluations make use of analytics and metrics that can be easily captured, using existing inbuilt social media tools, to assess how successfully campaigns have engaged with campaign participants and audiences. Typical social media analytics include the number of ‘views’, ‘shares’, ‘comments’ and ‘likes’. This makes sense as a starting point because of the key point of difference online media have over traditional media – it is social and interactive.
The use of social media for health promotion is suggested to be most valuable for their potential to engage with audiences.21 A common criticism of social media is that, while they increase participation, they actually lessen the level of commitment and engagement needed to be part of a campaign or initiative.22 It takes very little effort to like a campaign on Facebook, and there may be no direct link between the number of likes and the likelihood of behaviour change. Large numbers of followers or participants may actually mean very little in terms of how important or meaningful a campaign truly is.23 Given that public health agencies are most interested in affecting behaviour, it is essential to establish if there is a connection between engagement and behaviour change.
Engagement for the purposes of process evaluation has been described on three levels:
1) Low engagement – an agreement or preference for content, such as a like on Facebook
2) Medium engagement – people are involved in sharing content with the capacity to influence others, such as sharing or retweeting a campaign message on Twitter
3) High engagement – actual participation in offline interventions that results from some exposure to a social media campaign, such as making an appointment for cancer screening.21
This engagement model could be adapted to measure the level of engagement of participants in a social media campaign that aims to change personal health behaviours such as quitting smoking or eating healthy foods. However, given that changing these behaviours does not actually require participants to take part in a real world ‘event’, engagement could be better stratified in terms of those participants who had the lowest level of online engagement (liking or following content but with no further interaction) through to those who had the highest level of engagement (interacting with all online campaign activities and creating original content).
What remains to be understood or studied in depth is how, or if, increased engagement and participation increases the likelihood of action or behaviour change. There is an underlying assumption, evident in the lessons reported by the case studies, that fostering higher-quality online engagement leads to increased likelihood of action. Evaluating whether this is true would be an incredibly valuable contribution to the social marketing literature.
Conventional process measures, such as the number of likes, tell us little about a campaign’s impact on behaviour. It is essential to move beyond the exhaustive metric monitoring of process evaluations to outcome evaluation. It is possible to extend and adapt process evaluations to include more outcome-based measures. For example, a key performance indicator for a social media campaign promoting sun protection behaviour could be the number of young people who apply sunscreen before every trip to the beach. This information could not be gleaned from a collection of simple metrics, but could be achieved through multiple methods, including analysing the content of any online comments and interactions for evidence of behaviour change, and conducting surveys and interviews with campaign participants to directly ask about their behaviour.
To test the underlying assumption that highly engaged participants in a social media campaign are more likely to change their behaviour, it would be very desirable to survey a sample of campaign participants with varying degrees of engagement. Engagement could be segmented into the low, medium and high levels of engagement captured within a process evaluation. To illustrate, the membership activity of a Facebook campaign could have a possible breakdown for comparison of:
1) Those that simply ‘liked’ a Facebook page (low)
2) Those that commented on or shared Facebook page content (medium)
3) Those that did all three of these activities plus made any further original contributions or posts to the Facebook page (high).
Alternatively, contacting people who are known participants of the social media campaign at any level of engagement, compared with the general public, could also provide useful comparison data on effectiveness. A combination of these approaches would offer a comprehensive view.
If it was found that those participants deemed to be highly engaged with the social media campaign were the most likely to change their behaviour, simple metrics that measured engagement could serve as indicators that behaviour change was also likely in future campaigns. If evidence can be found to support this assumption, it would be a very useful and fast way to assess if campaigns are working from the very beginning. This approach would never replace comprehensive outcome evaluations; however, it could help program planners to adjust campaigns that failed to generate high-level engagement. Finding that the level of engagement had no actual bearing on behaviour change would be an equally worthwhile outcome.
Social media are unlikely to be the sole solution to changing behaviour, but it appears that they can play a valuable role in contributing to campaign objectives. To make the most of the opportunities that social media offer, and to minimise the risks, social marketers need to consider their objectives, invest the appropriate time and resources, and take a ‘test and learn’ approach. To truly understand the possible impact, moving beyond measuring the number of likes and developing outcome measurement frameworks is vital.
© 2015 Freeman et al. This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Licence, which allows others to redistribute, adapt and share this work non-commercially provided they attribute the work and any adapted version of it is distributed under the same Creative Commons licence terms.