Messaging Platforms: Best Practices, Costs, Security, and Privacy

Session by: Maurice Sayinzoga (Digital Impact Alliance), Boris Maguire (Echo Mobile), Christoph Pimmer (Learning Across Frontiers), and Charles Copley

Written by: Cathy Richards

“By 2019, an estimated 3.9 billion people will be using messaging apps (Activate & WSJ Tech). NGOs and the development community have already begun to embrace the opportunity afforded by these platforms to reach more people and track progress and success of their programs.” Led by Maurice Sayinzoga (Digital Impact Alliance), Boris Maguire (Echo Mobile), Christoph Pimmer (Learning Across Frontiers), and Charles Copley (Praekelt.Org) this panel session provided an overview of the current landscape of messaging platforms, described how these platforms are being leveraged for development and relief work, and discussed the related opportunities and challenges of implementation.

Dr. Christoph Pimmer spoke about his study of nurses and nursing students who participated in WhatsApp-moderated professional groups during placements and school-to-work transitions. He found that these groups generally enhanced participants’ knowledge and resilience and that they reduced professional isolation and stress. The project involved not only formal training and education but also informal learning and problem solving — such as knowledge transfers. The challenges encountered by Dr. Pimmer mainly dealt with the unregulated nature of these groups such that there were risks around patients’ privacy breaches, blurred boundaries compared to the traditional healthcare provider/client relationship, increased proliferation of misinformation, and on occasional inappropriate use of the chat groups at the bedside. His recommendations included leveraging pre-existing social capital by encouraging local leaders to initiate the groups themselves and act as moderators. Similarly, these local leaders can develop ground rules on the scope and accepted behavior of these groups. 

Maurize Sayinzoga listed affordability as the top concern around the use of messaging platforms. His recommendations include:

1. Go where people’s attention already lies: Try to not only understand regional communication preferences (do people prefer SMS over WhatsApp?) but also the community’s communication behaviors and preferences among different demographics. 
2. Prioritize user needs over implementer needs: Conduct user research on the various platforms and select based on user appeal not necessarily ease of integration. Understand the costs that users will pay to use your system along with their willingness to pay. Remember that SMS is still an option.
3. Partner for scale and technical expertise: Make sure you have enough resources to grow. Governments can help overcome the challenges to scale while third party developers can fill technical gaps. Similarly, messaging app providers can help overcome limitations of features and policies while partners can help provide content. 
4. Prioritize content and personnel: Systems are only as good as the content they provide. Make sure to develop maintain sectoral expertise and make plans to handle user feedback and inquiries. 
5. When possible, engage with more users through the use of multiple channels: Make sure to assess who and how many people can access each messaging platform, the cost savings for users when with provided multiple channels, as well as the costs and potential complications that multiple channels can add to a project. Prepare to manage these parallel systems when possible.
6. Take into account the gender gap: The gap in internet use increased in Africa between 2013 and today. Women in low and middle-income countries are 10% less likely to own a mobile phone.

Charles Copley (above) spoke about how his organization was partnering with WhatsApp to programmatically deliver messages in development contexts. They are conducting research studies around the use of this tool to improve outcomes using methods and tools such as 2x2x2 factorials, sequential multiphase adaptive randomized trials, experiments as markets and natural language processing — specifically around the maternal health context with the MomConnect Connect project and related Turn application. Copley’s tips include:

1. Balance individual privacy with overall good: Aim for a consent-driven model. While certain data would be useful for developing improved health systems, this data still has to be used responsibly.
2. Consider anonymous groups: Develop the capacity to host anonymous chat groups in which individuals do not know each other’s names or numbers. 
3. Websites are still foreign in certain contexts. It is much more common to interact with a Facebook page or WhatsApp contact than to visit an actual website. Additionally, mobile surveys can be data-heavy.