Scaling Up our DIY Self-Help Group App with Partners in East Africa

In early 2015, Code Innovation and our partners at One Hen Inc. visited the implementing partners for our Self-Help Group app in Ethiopia and Tanzania. After a successful pilot in 2014, our plan was to scale up the use of the app by 1000% focusing on new users in food insecure areas of both countries.

We met with partners at Tearfund Ethiopia and with Tearfund Tanzania's local NGO implementers, the Christian Council of Tanzania (CCT) to decide on a viable plan for multiplying our impact and rolling out a new-and-improved iteration with content that we estimated would last for about six months worth of weekly Selp-Help Group meetings. According to our previous coordinator, during our 2014 pilot in Ethiopia this was about the time it took for new groups to raise enough capital and develop enough business acumen and group momentum to begin to give their first loans.

This is a write-up of how Phase 2 of the project went, in terms of fidelity to our plan and also around ICT4D best practices and lessons learned. Wherever possible, we'll tie what we're doing and learning into the Digital Principles because we're proud to be one of the endorsing organizations contributing to this emerging field of practice.

An Overview of our DIY Self-Help Group App

(If you're already familiar with our project, feel free to skip this section. You can also read more background here and here.)

For those of you new to the project, in 2013 we began working with the US non-profit One Hen Inc. to digitize and scale Tearfund Ethiopia's successful Self-Help Group model of savings and credit groups, themselves adapted from the model pioneered by Myrada in India. The groups have shown a cost-benefit ratio of approximately 1:100* with long-term and far-reaching social and economic impacts on members and their communities, lifting people out of poverty over time with very little outside support.

Working closely with Tearfund Ethopia, we adapted their Self-Help Group modular curriculum to a mobile interface on a free and open source Android app you can download from the Google Play store here -- although it's very much still in Beta for now. Over a 12-week pilot, we found that the facilitators thought the app was a useful professional tool and facilitation guide and that they'd already begun using it to start new Self-Help Groups not officially involved in our pilot.

Our Plan for Phase 2, a.k.a. How to Scale 1,000% in Six Months

Based on the positive feedback we got from Self-Help Group facilitators, we sought to expand our reach in Ethiopia and begin working in a new country, Tanzania, with the same organizational partners. With DfID funding, we were able to focus on food insecure regions facing hunger because of the failure of the previous year's rains. Due to a poor harvest because of the drought, families and communities in these regions were considered particularly at risk for hazards related to food insecurity. Our partners selected the Humbo and Angacha regions in Ethiopia and Kongwa in the Dodoma region of Tanzania to scale-up our pilot with 25 new Savings and Credit Groups to be created in each country over six months of field implementation.

Because we were working in two districts in Ethiopia, and also to see if closer supervisory support would yield better weekly reporting data, we tried a new approach to coordination, appointing one District Coordinator for each area, supervised by a single Project Coordinator based in Addis Ababa. In addition to regular check-ins by email and phone, the Coordinator visited every other month in person to ensure the District Coordinators were feeling supported with the technology and the savings and credit group formation process.

In Ethiopia, we worked with experienced Self-Help Group facilitators working in new parts of the country starting groups of primarily young people out of school and over age 18. The focus on youth created some challenges because there was an assumption that young people did not have any source of income, although Tearfund's program model specifically addresses this assumption with a reframe of available local resources and close-to-home economic activities. Nonetheless, we did see below average group retention rates in Ethiopia because the SHG system itself was not established in the communities we selected and was, instead, fairly unknown. Accordingly, parents and youth members were quick to get discouraged and to discourage others from attending the groups. In the Nazaret region, where we first piloted, SHGs had been established for over a decade and belonging was considered to be admirable and beneficial, so this was our first time as a partnership facing a situation where people did not show up with motivation because of a favorable context. Also, in some cases youth decided to enroll in school or move to urban areas to look for work during the pilot program period, so SHG membership was more variable than is usual for Tearfund Ethiopia programs.

In Tanzania, our partners at CCT decided to work with entirely new and inexperienced facilitators in regions where Pamoja groups ("Pamoja" means "together" in Kiswahili and is CCT's name for our Savings and Credit Groups) had not yet been established. This created a number of early challenges that were evaluated to be worth the extra effort because of the acute community need for this kind of support system, given the hazards and risks members were facing around food insecurity and with the drought. It meant that our Coordinator spent half of his time directly working with and training facilitators on the mobile technology, app functionality and reporting protocols, but also that the gains that we saw over time there showed that the program can work in a new and extremely challenging use case.

Because we're still early in the app development and digitization process, we continued our system of weekly feedback from facilitators to get specific inputs on areas of the curriculum that worked well and that need expansion. This system continued to give us the real-time, actionable data that we need to make strong iterations between phases, and we anticipate continuing it in the future until we move out of Phase 2 (testing with new countries, partners and in new world regions). Phase 3 will happen when the app can be used by a new, inexperienced facilitator to successfully learn facilitation skills, recruit and start a group, and save and lend while building group ties over time. We have a ways to go, but we'll get there!

What Went Well in our Rapid Scale-Up

We are happy to report that a number of key areas went extremely well. We're going to summarize them here, but do get in touch (info@codeinnovation.com) if you'd like to hear more details as we're keen to share what we know with our ICT4D community.

App Functionality and Usability: The app did not require repeated training for new facilitators to use, especially around the key curricular areas of meeting content.

Expanded Content around Case Studies, Games and Stories: We hoped to include content in the app that would take group members well into six months of weekly meetings, and we succeeded in doing that with our expanded curriculum around social business skills development, conflict resolution and disaster risk management/disaster risk reduction. Facilitators and group members enjoyed the illustrative content in particular, and over the course of Phase 2 we've collected a wealth of additional content to help us build out the curriculum further.

Facilitator Training: Our new module created a step-by-step training guide for new facilitators to learn basic skills, recruit group members and develop self-organized learning for their own professional development. We heard from facilitators throughout field implementation that it was an appreciated part of the content.

Facilitator Preparation: Before each module throughout the content, we expanded the information needed to prepare facilitators for their weekly meeting. We heard that this was an extensively used part of the app this time around and were requested to continue building it out as a resource for planning meetings.

Hardware: We selected locally-purchased Tecno tablets available for about $200 in Ethiopia per device and about $100 in Tanzania per device. The higher cost in Ethiopia is due to national taxes on ICT, as the tablets themselves were almost identical. Every device continues to function without damage at the time of writing, a testament to the care with which our facilitators treated them and also to the durability and appropriateness of the tablets themselves in rural East Africa.

Reporting and Supportive Supervision: Weekly reporting kept facilitators, coordinators and us in close contact to problem-solve proactively and ensure that our content and UI/UX was meeting their needs in running groups and also in their own professional support and development. In Tanzania, reports were sent via facilitators' Gmail accounts and our users created a What's App group on their own initiative to share success stories, keep in touch and help each other resolve group, tablet or meeting challenges.

Secondary Benefits of Accessible Mobile Technology: In most cases, facilitators used their tablets for professional and personal development, including engagement with LinkedIn, online news and Facebook social networking. In many cases, facilitators began to pass around the tablet during meetings so that members took turns facilitating the key discussion points during group meetings. In a few cases, facilitators made their tablets available to community and group members to access the internet, creating strong secondary benefits in areas that did not previously have easy access to mobile technology.

What We Learned for Future Partnerships

There were some key areas for lessons learned as well, detailed in brief here. Again, please do get in touch (info@codeinnovation.com) if you're keen to hear more about these, as we'd love it if no one in ICT4D ever made these same mistakes again!

Solar Chargers: In all cases where hardware is provided, we will be advising partners to purchase locally sourced solar chargers to enable the tablets to be charged directly by the facilitators whenever needed. Relying on local charging stations is both time-consuming and expensive, and could in the future be a source of low motivation to use the app.
App Updates: Because access to mobile data is so slow and wifi is often completely unavailable, we needed a new system to update new app versions so that facilitators would be sure to be using the latest app version. We are using our Coordinator's laptop and installing APKs directly onto tablets during field visits in the future. But this is a function of our beneficiaries being in unusually remote areas underserved by electric infrastructure. If we were targeting robust growth in an urban area, this recommendation would likely not apply.
New Group Formation: We had anticipated that 25 groups would be fairly easy to form over six months in each country, but in fact we will only reach our target in late 2015/early 2016. In Ethiopia, working in a region where SHGs were not known by the community made their establishment slower than anticipated. In Tanzania, new facilitators were only ready to create new groups after their existing ones had been established for about three months, so relying on facilitators to create multiple groups should anticipate some phasing delays. At the moment in Tanzania, new group creation is on hold because members would not have the income needed to contribute to savings, since it is the very end of the dry season and family resources are very scarce. A few weeks after the short rains begin (in December or January, we hope), members will once again have the financial resources and be able to begin group savings.
UI/UX and Usability Testing: Secondary app functionalities were not as easy for new users to navigate as the curricular modules, namely our Community (or social media) section and our group login system. Based on usability testing directly with facilitators in Tanzania, we have a great list of priority fixes in this area.
Multimedia When Possible: Adding photos and illustrations, especially for case studies, will help to make the content more personal and come alive for members and facilitators. It was a repeated ask from our group interviews and something we're looking into while keeping in mind that we don't want the app (already around 10MB) to become too heavy to download in low-bandwidth areas.

What our Self-Help Group App Group Members Had to Say about the Project and our Process

During a recent field visit to CCT's Pamoja Groups in Kongwa, Dodoma region in Tanzania, we were able to interview nine groups in four village areas.

In our Self-Help Group model, each group member contributes weekly through buying two types of shares, social shares and savings shares. Each week, each member contributes one share to the social fund, for use by group members in emergencies. In addition, they can buy savings shares at a minimum and maximum set by the group.

In all Pamoja groups, the social fund is repaid without interest and had minimum 1 share @ 500 Tsh contribution per week, slightly less than $0.25 at the time of our visit.

In Mautya Village in Kongwa District, Dodoma Region, participants reported:

"We are facing hunger in our families and communities because we are primarily agricultural and because of the lack of rains last year and the failure of our crops."
"We are using the social fund to buy food."
"Group social ties give us strength to face the challenges of the drought together. We do not feel alone."

In Nguji Village in Kongwa District, Dodoma Region, participants reported:

“We are facing hunger in our families and communities because we are agricultural and because of the lack of rains and the failure of crops.
"We are using the social fund to buy food and pay school fees."
"Because of the group, we are not facing too much hunger at the end of the dry season and we feel supported by each other."
"Belonging to the group helped to improve my existing business and my profits have increased."

One-third of the participants in Nguji owned their own mobile phones and 80% had their own businesses.

In Machenje Village in Kongwa District, Dodoma Region, participants reported:

"There is no rain, and everyone here are farmers. Bad harvest means hunger. Lack of rain increases the price of food."
"If I don’t have money, I can take a loan to invest in a business and use the profit to pay back the loan and buy food for my family."
"Our economy is so much affected by the drought because we depend on agriculture and there is no rain or harvest. It is difficult."
"Because I now have a small business, I can buy food and eat with my family."
"If someone is sick, a loan [from the social fund] can take them to hospital and pay for their immediate needs."
"If a group member has any emergency, anything in life, we can support them."
"These groups are good. We encourage anyone to join. However, know that if you take a loan, it can be challenging to pay it back so that another person is able to take a new loan."
"The community originally thought that these groups were a trick, but now that they've seen our success and the capital we've raised, they themselves want to join."
"I had a business before, but I was inexperienced. Belonging to the group helped to improve my business skills and share with others. Now, I am helping my family to have a good life."
"Belonging to the group has really helped my family. With a loan, I have been able to expand my tomato selling business."
"This is a bad year because of the lack of rain, so buying shares every week is difficult, especially now that it is dry season. My savings come from collecting firewood in the bush and selling it in the village."

30% of the group members had businesses before joining and 46% do now. 58% own their own mobile phones.

In Laikala Village in Kongwa District, Dodoma Region, participants reported:

"Life is difficult. I joined this group to get out of poverty."
"In most cases, we struggle to pay for school fees and because of the group, we make sure that we pay for all the school expenses."
"Using the tablet has introduced me to new things and ideas, and it is good for me."
"The social fund is for problems or unexpected disasters."
"Our group made an IGA whose profit goes back into the social fund, because we depend on it so heavily now. The IGA involves buying sugar and rice wholesale and each members sells some and returns with the profit."
"People should join groups because they are sustainable. Members are there for each other and will continue to be there to help each other."
"This community depends on agriculture. Without rain there is no food. We have hunger and no money to buy commodities. We are all affected."
"Without a good harvest, there is no money and without money, you cannot buy anything. There is no water for gardening."
"Without food at home, after a poor harvest, loans help our families to eat."
"With the problem of the lack of rain, most people are bankrupt so others can’t help, but the group can help, especially with a small business."
"I didn't have a business, but then I took a loan and now I have a profitable small restaurant."

13% of the group members had businesses before joining and 52% do now. 55% own their own mobile phones.

Next Steps for our DIY Self-Help Group App

We are in discussion with CCT, Tearfund Tanzania and Tearfund Ethiopia to continue to scale up with their new and existing Pamoja and SHG facilitators in the coming months and into 2016. In addition, we have a new partnership with World Vision Tanzania working with their Volunteer Savings and Loan Associations with groups in the Babati regional cluster (of Tanzania). Stay tuned for more developments early in the new year, when we'll be releasing a new-and-improved iteration based on what we learned during this Phase 2.

Our own goal at Code is to scale the Self-Help Group App impact to 1 million direct beneficiaries within three years. Because of the economic and social need, we hope to concentrate mostly in sub-Saharan Africa but it will depend where we find implementing partners and funding. Of course, as we iterate closer to a stand-alone app with full functionalities, , our own inputs for consecutive iterations will become less necessary.

We hope to find partners in all parts of the world, but in Africa in particular, who are interested in using our Self-Help Group app to train and support facilitators starting their own groups in their own communities, helping to empower people to create social and economic support systems that reduce their vulnerability to stresses, shocks and poverty.

Want to partner with us on this or other projects? Get in touch (info@codeinnovation.com)!

* Cabot Venton, C et al (2013). “Partnerships for Change: a cost benefit analysis of Self Help Groups in Ethiopia.” Tearfund, Teddington, UK.