A lot of civil society organisations have a secret.
Their tech projects are failing.

What does the research tell us?

All seven MAVC-funded projects using ICT to promote accountability in health systems “saw ICTs as offering immediate solutions, without carefully considering how and why these technologies might facilitate citizen engagement and voice.” Four out of seven found: “disconnects between their expectations of what the technology could do, what it actually did, and the implications for accountability.” Read the research.

More than 75% of 38 transparency and accountability initiatives interviewed in Kenya and South Africa (some of which were MAVC-funded) weren’t happy with the technology tools they chose for their projects. Read the research.

Of 24 organisations using technology in transparency and accountability projects in Kenya (18 of which were funded by MAVC), 65% expected easier buy in from project’s stakeholders, gatekeepers and users. Read the research.

Evidence suggests that projects using technology to improve transparency and accountability are more likely to succeed when:

Governments already want to get (and act on) feedback from citizens. Read the research

Organisations already know the people or government actors they were working with - and have good connections with them. Read the research | Read the research

Projects’ advocacy strategies act at several different levels: offline as well as online (Read the research | Read the research): and targeting people and institutions from different sectors at the same time. Read the research

We asked people at MAVC’s event ‘Appropriating Technology for Accountability’ how we can we set more realistic expectations in future.

Researchers found that organisations they interviewed in Kenya “lacked the technical capacity to fully understand or anticipate issues associated with a given technology” - in 59% of the projects assessed, participants said that a lack of technical knowledge was a barrier. Read the research.

In a review of MAVC-funded projects in Kenya, several “admitted that they may have been over-reliant on the advice of the tech experts they had engaged” and held them responsible for designing tech solutions that did not fit their context. Read the research

.

Of the seven organisations working on mHealth in Indonesia, Nigeria, South Africa and Tanzania that spoke to researchers, three felt ill-equipped to coordinate with tech providers and telecommunications companies. Read the research

Only 3 of the 38 Kenyan and South African projects interviewed in 2015 did research in all three areas relevant to tech project design: the accountability problem, the technology options and the people who would use it. “Organisations were not able to identify or fill in their knowledge gaps, instead focusing on areas where they already had strong knowledge.” Read the research

Evidence from MAVC’s research projects suggests that it’s hard for organisations to find technical support from providers that understand their needs. However, when they do, it can make a big difference.

For example, the Ghana Center for Democratic Development and the technical provider VOTO Mobile took “far longer than anticipated” to design an interactive voice response (IVR) survey, going through eight iterations, but found that more people - particularly women - completed it as a result. Read the research.

Another group in South Africa with a community organising background, “struggled to find appropriate technologies for a project monitoring public service delivery. They had much more success when they found a technology partner that they felt shared their values... to guide their understanding.” Read the research.

This suggests a need for organisations to understand and connect better with technical providers. For example, South African organisation the Foundation for Professional Development found that “there’s a need for a translator” between programme and technical staff. Read the research.

MAVC’s experiences suggest that active communities of practice are one way of helping organisations build this understanding. Read the research

We asked people at MAVC’s event ‘Appropriating Technology for Accountability’ how we can learn better about accessing the right skills to use tech.

Only 15 of the 38 transparency and accountability initiatives in Kenya and South Africa that researchers interviewed in 2015 did any user research before choosing a tech tool. Read the research.

Separately, a review of MAVC grantees in Kenya found that “many grantees conceived tech solutions that did not take the constraints of the targeted environment into account” citing problems such as “a service problem notification platform [that] only worked on laptops, which very few people had.” Read the research

.

Researchers looking at 24 organisations in Kenya in 2016 found that “most projects were poorly grounded in their contexts.” Read the research

When organisations do user research, they usually find out essential information.

Looking back, 79% of the Kenyan organisations interviewed in 2016 said that “adapting to context mattered a lot for the project.” Read the research

For example, the South African organisation Yowzit found that its users would only find the citizen participation platform convenient if they spoke English, were digitally literate and already used similar ratings platforms. This allowed them to retarget their work more effectively. Read the research.

Meanwhile, those of the 38 Kenyan and South African organisations that did user research and trialled tech tools with their intended users were much more likely to be happy with their choice. Read the research.

In these cases, the user research typically took longer, but could actually save money and time overall. In South Africa, the Foundation for Professional Development (FPD)’s user research showed them that they needed to focus on an app focused on clients’ needs, rather than the more ambitious programme they had planned. Read the research.

We asked people at MAVC’s event ‘Appropriating Technology for Accountability’ why user research doesn’t happen more often, why it’s difficult, and what can be done about it.

In Kenya, researchers found that many projects “reported that key stakeholders and important issues were not discovered until late in the implementation of the project, often in its last phases.” Read the research.

Researchers found that the nine tech hubs they investigated in sub-Saharan Africa “often...avoid policy engagement altogether, or are constrained to doing so in ad hoc, superficial or premature ways.” They cited a “lack of mutual trust, understanding of incentives and shared buy-in” as reasons behind this. Read the research.

Projects built on existing relationships may be more likely to succeed.

The MAVC-funded mHealth projects that had already earned the trust and cooperation of government actors felt that they were more successful in reaching their accountability goals. Read the research. Research on the Kenyan community mapping initiative Map Kibera found that the trust networks it built over time “created the context that enabled specific outcomes” such as policy changes around education. Read the research.

This means thinking longer-term, where organisations build connections over time: to succeed, research suggested Yowzit needed to understand the “organisational infrastructure, culture and inner workings of the specific public entities you hope to affect” in South Africa. Read the research.

We asked people at MAVC’s event ‘Appropriating Technology for Accountability’ how they thought civil society organisations could connect better with different kinds of groups.

Of 24 projects researched in Kenya in 2016, “the majority of the adaptations [within their projects] had a reactive character, happened late, and were not properly communicated or coordinated well among the partners... Frequently, they were “not fully aware of the high degree of adaptiveness” needed to run a successful project. Read the research.

Researchers interviewing organisations in Kenya and South Africa in 2015 found those that built tools themselves were more likely to experience technical problems and delays, and were in many cases not prepared to handle unexpected changes. Read the research.

Organisations that can adapt during a project tend to see themselves as more successful, but most cannot do so - because of funder constraints.

The Kenyan and South African organisations that tested and trialled technology tools with their users were more likely to be happy with their selection. Read the research.

But researchers interviewing Kenyan organisations in 2016 noted that they had the most scope to change course in their early stages of projects. They found it much harder when they inevitably needed to make changes later on. In these cases, donors generally didn’t support projects to adapt, by failing to design monitoring processes that included feedback or disrupting project plans themselves. Read the research.

For projects to be genuinely adaptive, they need to be supported with flexible funding and transparent tools for communication with funders. Projects like Yowzit and FPD are examples of organisations that adapted as they went through their projects (and were supported to do so by MAVC). Read the research.

We asked people at MAVC’s event ‘Appropriating Technology for Accountability’ how technology could contribute to transparency and accountability work, if it’s used effectively.