Six key findings from engaging communities through messaging apps
Unlocking possibilities with staff satisfaction, future risks and iterative growth.
One year into a pilot to engage with communities through WhatsApp, the Innovation Service’s Digital Inclusion programme reflects on the pilot’s progress to date, and on a recent internal assessment that was undertaken, its various recommendations and the path forward for the pilot.
It was only ten years ago that ‘social media’ and ‘humanitarian aid’ started to be used in the same sentence. The CDAC Network — a network of UN agencies, NGOs, Media Development agencies and others dedicated to communicating effectively with crisis-affected communities — set up a seminar capturing responders’ experiences of the 2011 Tohoku earthquake in Japan. In the seminar they highlighted the power of social media and messaging apps in humanitarian response — something new for many attendees. Yet ten years on, and in spite of mounting evidence highlighting the importance of engaging with communities through social media and messaging apps in humanitarian action, organizations pursuing this journey are still just getting started. Systematic and strategic approaches to digital channels of engagement are, regrettably, still the exception rather than the norm.
COVID-19 has been an impetus for change. Physical distancing measures ushered an accelerated transition towards digital communication channels. A couple of months ago, UNHCR’s West Africa bureau released a new study highlighting the importance of digital technology and connectivity in people’s lives, specifically that — once again — WhatsApp is a very common channel of communication amongst those connected. The report recommended that digital channels should not be considered peripheral for organizations to meaningfully engage communities, but rather they encouraged colleagues to take strategic and systematic approaches that are integrated into broader engagement efforts.
It was to this end that UNHCR’s Innovation Service conceptualised a pilot to formulate a better way of doing business for the organization when using messaging apps, specifically WhatsApp. Of course, when we set out on this journey, colleagues across the sector had questions as to why we were doing this. UNHCR’s Innovation Service has long emphasised digital risk and privacy issues. We’ve worked closely with other organizations like the ICRC on reports that highlight risks with using messaging platforms, undertaken research ourselves like ‘Connecting with Confidence’ that unpacks real and perceived digital risk to communities, and created guidance on how to engage with communities through social media. It might have been that they’d expected us to use Signal, frequently touted as a preferred option for the privacy-conscious. However, UNHCR also has a commitment to community-based approaches and ensuring that our approaches to community engagement are built of local capacities. We know this means working with communities’ preferred and trusted communications channels, meeting people ‘where they are’ in online and offline spaces.
We saw this as an opportunity to strike the right balance between engaging with communities where they are online, and the need to put robust measures in place to emphasise consent, users’ privacy, and data rights. Were we going to get everything right on the first attempt? Definitely not. But noting the shortcomings of some of the existing practices in the sector, we prefered eyebrows raised to red flags, which may have been the case without concerted efforts to improve practice.
We set up the project alongside Praekelt.org — a social impact organization that uses mobile technology to improve people’s lives — and their Turn.io product, which helps organizations engage with communities through WhatsApp at scale. In our previous blog ‘Meeting communities where they are’, we outlined how seriously we were taking issues of digital risk and data protection in the design of the service.
And as part of this pilot phase, we committed to undertake an external assessment of the pilot — looking specifically at digital risk issues to complement internal data protection impact assessments that were being undertaken. We saw such a review as a vital opportunity to get an expert perspective on how we’d been looking at these issues, find space for improvement and advise on a way forward for the pilot.
Fast forward to today, we’re over a year into the project, and at the end of the first phase of our pilot. As planned we proceeded with the review. We brought onboard Nathaniel Raymond, Lecturer at the Jackson Institute for Global Affairs, who has an extensive background and knowledge relating to digital risk issues in humanitarian settings, to lead the review.
When Raymond provided us with his assessment report, we wanted to publicly reflect on some of the recommendations and share a related memo looking at ‘Conceptualizing Digital Risks to UNHCR Persons of Concern in the WhatsApp Era’. We felt these would be relevant and useful for other humanitarian organizations working to use social media and messaging apps to engage with crisis-affected people.
Download the memo here.
New possibilities and reflections on digital risks in humanitarian innovation initiatives
For the review, Raymond was provided with relevant documentation on the project, applicable UNHCR policies which was complemented by informant interviews with critical UNHCR team members involved in implementation. The following is simply a snapshot of the report contents but what we deem the most relevant findings that will help us to pivot the direction of the pilot:
- It’s a great opportunity for staff to be closer to communities, and this has positive effects
“How UNHCR does or does not use messaging apps as part of communication with Persons of Concern will likely affect staff retention and wellbeing — in both the short and long term. Human resources should study (including through anonymous surveys) the role that UNHCR’s attitudes towards mainstreaming and managing Persons of Concern Relationship Management activities with messaging apps has on staff satisfaction and burnout.”
The first and surprising finding we’d like to prioritise was how important such work has been for the UNHCR staff in the country operations. Maintaining proximity through digital channels had a positive impact on staff satisfaction in their functions. This was also directly linked to management support for such work i.e. where management created space for greater engagement with communities, that had a positive impact on staff wellbeing. The Innovation Service runs a programme directly aimed at senior managers in creating more enabling space and conducive environments for innovation and these findings underpin the importance of investing in buidling the capacity of managers in this area, with that additional insight linked to staff satisfaction. We had honestly not thought about this dimension at the outset of the project, but we want to examine it further, providing insights to HR to take forward.
2. Risk is more than data protection
“Misinformation/disinformation/hate speech and online fraud/exploitation are the two most urgent threats identified by this report.”
One of the findings, that came as something of a relief, was a broad affirmation that the approach we’d taken with regards to data protection within the system was the right one. We worked closely with UNHCR’s data protection team to address this, and we discussed it in an earlier blog. There are a number of risks such as a breach of the personal data of refugees through insecure systems, or unauthorised third party access. Such breaches can result in targeting from third party actors, fraud, identity theft, and more.
Still, this doesn’t mean the approach we’d taken was perfect.
The assessment found that the most important of these risks is linked to misinformation, disinformation, and hate speech. Beyond a couple of areas of improvement, the review highlighted the need for better digital risk assessment frameworks for the Innovation Service, the Agency and the sector as a whole. While data protection impact assessments were undertaken, issues such as misinformation and disinformation are not adequately addressed by data protection frameworks alone.
Moving forward, we’ll aim to maintain a strong approach to data protection in the pilot, but also seek to build on the recommendations of the review and develop more advanced risk assessment frameworks that can also be used beyond this project. Findings from the Turn.io pilot will feed into UNHCR’s action plan on Hate Speech developed in support of the UN strategy and Plan of Action on Hate Speech. Through these frameworks, we aim to move beyond protecting UNHCR systems towards more people-centric frameworks that address the real and perceived digital risks communities face in their information and communication ecosystems.
3. While there have been no critical incidents, we need to do more to plan for them
“A clear chain of command should be established for…critical incident management and prevention in the case of a messaging app/CWC modality related emergency.”
The Innovation Service has consistently emphasised proactive activity, scanning horizons for future challenges that may impact a project’s outcomes. The same ethos applies when looking at critical incidents. We can’t wait for them to happen. The assessment found that while there were safeguards in place, no systems are perfect and that critical incident planning needs to be advanced when it comes to the communications ecosystem. This is not limited to issues like data breaches for instance, but also covers broader issues including account hacking, impersonation, actors negatively targeting specific groups, or releasing personal information on vulnerable community members. The review highlighted that a key risk was a third party impersonating UNHCR through WhatsApp with a fake account — easily possible if accounts used do not have the ‘green tick’ verification (all Turn.io lines do have this verification). Such activity could risk massive levels of misinformation, erode trust in UNHCR and lead to potentially life-threatening risks to community members. UNHCR needs to develop frameworks to address such events with clear protocols and processes.
4. Working with(in) Terms of Service agreements
In many humanitarian projects, there are external factors that influence outcomes beyond the immediate control of those delivering them. We have humanitarian principles to underpin our actions but often it is creative thinking and innovation that channels a pathway through, challenging assumptions, testing hypotheses and looking at every possible alternative. In order to use the WhatsApp Business API, organisations must accept the WhatsApp Terms and Conditions and this includes its privacy policy. These are not ideal for humanitarian organisations. Specifically, the WhatsApp Terms and Conditions don’t consider the mandate of UNHCR as an international organization, nor the humanitarian nature of the work, given they are essentially geared towards commercial organizations (it is a business API after all, right?). One concrete example is linked to the business models — specifically advertising frameworks — that companies like Facebook lean on to generate revenue. If vulnerable populations are part of this pool, it could link to targeting based on said vulnerability, potentially placing communities at further risk. UNHCR will continue to engage with WhatsApp and others to highlight the unique protection needs of refugees and the need for frameworks that take these into consideration.
5. Investments with both time and money pay dividends
“Protection officers in the Ecuador operation have reportedly saved one hour to an hour and a half of time a day due to reduction in time expenditure required to answer PoC info requests.”
One thing the pilot highlighted was how it was possible to save the time of UNHCR protection staff when using a system like Turn.io to support relationship management with communities, using automated functions to support the triaging of requests and messages. Automated functions could include the provision of office locations, advice on asylum procedures, access to services from partner organizations and more. This is not a ‘quick win’ in order to save time. The teams needed to invest more time and effort in not only creating an adequate service that meets the needs of users (with extensive user testing etc.) but also answering people’s questions and knowing how to refer any issues or concerns that require follow-up from specialized staff. It creates space for teams to think more strategically about interventions rather than ‘firefighting’, which can spark new and innovative approaches to staff members’ work.
Recommendations were made to ensure that teams have the confidence and capacity to adopt new communication channels. This means creating a culture of open communication and potentially adapting staffing structures to support such digital engagement. This will be part of the criteria for selecting the next round of countries in the pilot.
6. Building on the initial pilot and testing new assumptions
“Initiate a PILOT PLUS phase that continues the pilot phase in current countries and adds additional pilot countries through the end of 2022. Additional pilot countries should be intentionally selected because they have different operational contexts than the current pilot countries, which are primarily urban Persons of Concern contexts.”
There have been a lot of discussions following the first phase relating to the ‘standardisation’ of the tool within UNHCR. We knew that we weren’t really ready for this and the assessment agreed with our instinct. We’d only covered limited types of contexts and population groups (mainly refugees in urban areas) and a limited number of services had been launched at the time of assessment.
While UNHCR’s internal processes made such a ‘scaled’ pilot slightly challenging to initiate (it’s not a ‘normal’ thing to do) eventually we came to an arrangement that will indeed see the pilot continued for a phase two over the coming year. We see this as a critical step as part of the innovation process, which the Service seeks to follow across its projects and initiatives. We started with a clear challenge, developed and implemented a solution in close discussion with communities, and are now scaling iteratively, rather than jumping headfirst into corporate decision-making.
Moving to the next phase
“It would be easy to treat integrating the reality of digitally networked PoC communities into UNHCR’s ways of working and public profile as just another fad in a litany of data-driven innovations, many of them short-lived.
Don’t.”
Internal UNHCR Turn.io Assessment, Nathaniel Raymond
As we take heed of these findings and recommendations, we’ll be moving forward with the pilot in its second phase. This will give us an opportunity to strengthen further our approach to data protection and digital risk issues, while testing the approach in different contexts.
Ultimately by the end of the second phase, we’ll aim to have not only reached a significant number of people with the different services across different parts of the globe, but to have generated significant learning and understanding. We hope that this will contribute towards a growing body of literature that details clear tactics for how humanitarian organizations can move forward to engage communities through social media and messaging apps. To complement this, we will develop tools and products that can support teams on their journeys.
Stay tuned for further updates on the project and its progress. We are interested in hearing more about other efforts, as well as any suggestions you might have with regards to useful products or tools, so please don’t hesitate to reach out to us at hqconref@unhcr.org.
