An Overview of Our Self-Help Group App Pilot
This is the second article in our series about the Self-Help Group app project. In the first article, we explain a bit about the Self-Help Group as a model and explore "Why the Self-Help Group Program Model Is Ready for Mobile." If you'd like to receive our updates by email, subscribe here.
Appifying the Self-Help Group Model
In the summer of 2013, Code Innovation began speaking with development economist Courtenay Cabot-Venton about the exponential impact she had assessed in a Self-Help Group (SHG) project funded and implemented by Tearfund Ethiopia. As we talked about the potential of using mobile to adapt the approach and help take it to scale, Courtenay became Director of International Programs for the U.S.-based the financial literacy and microfinance educators at One Hen, Inc. We worked together with Tearfund Ethiopia to create a project plan for an early iteration of the app in Phase 1. The goal was to develop, test and iterate a Self-Help Group app that trains facilitators how to start and lead a strong and self-reliant SHG, because we see mobile as being the best way to take this innovative and impactful approach to scale.
Once we had secured funding and agreed on an implementation plan, the next steps were to begin the process of user-driven design with Tearfund's expert facilitators and program team. Through extensive Skype and email conversations, we began to understand the existing ecosystem that Tearfund, their partners and the SHG model work within. We also began gathering all of Tearfund Ethiopia's SHG content: everything from facilitator training materials and trust-building games to the sample passbooks each SHG members carries to keep track of her savings and loans. Tearfund already had a project model with a cost-benefit ratio of 1:100 that grew steadily at 20-30% per year -- it was up to us to build on what was already working and to adapt their approach for m-learning. Over months of collaboration, we developed a mobile curriculum that mirrored the basics of the SHG process with modular content for self-guided group development and a basic structure for each meeting.
We worked with a group of application developers and technology educators in an organization called eMobilis based in Nairobi, Kenya. Because the app would be tested in East Africa, we chose capable developers from the region who were able to input around user interface and user experience (UI/UX) with confidence because they had a clear understanding of our target market and user base.
We believe that the open source standard is a game-changer for international development and humanitarian aid, and should be mandatory in all innovations and ICT4D projects, so the Self-Help Group app is built on an open source platform and released under a GNU General Public License. We are publishing our open source code on GitHub in a couple of months. We also made our content material free and open under the Creative Commons. Some of our partners were not familiar with the open standards of ICT4D, and we took the opportunity to share about the importance of the aid and development community building around the open source ethic. We found that communicating the value of open source helped to align our partners around a similar vision for the end-product of our collaboration: a free open source Self-Help Group app that trains facilitators to seed and lead the SHG process.
Phase 1 of the Self-Help Group App
You can find the Self-Help Group App free on the Google Play store here -- but please note that although this is the latest version, it is not intended for use by untrained or unsupported facilitators yet. If you are interested in using it and testing it out, please contact us at info@codeinnovation.com so we can advise you when we'll have a version that's ready for that use case.
Here are the core areas we created for Phase 1 of the SHG app:
Modules, Lessons and Steps: The basic content for the SHG process, from a group's first meeting to when they are ready to give their first loan, happens here. We divided our content into three Units that mirror basic group formation, from setting out structures for participatory decision-making to deciding on bylaws and what to do if a member taking out a loan is delinquent in paying it back.
Each SHG meeting has a series of steps. To make the basic structure of meetings easy to follow, we created a Meeting Checklist screen that helps facilitators to ensure that the group has covered everything for that week before the meeting adjourns. Although our facilitators were extremely experienced, they found the Meeting Checklist helpful to guide the SHG meetings, using it as a memory aide when they needed it.
By the way, we're changing the language of the content sections to Units, Modules and Steps in Phase 2 in order to move away from overly school-based curricular language. We wanted to ensure that the facilitators don't feel the need to lead or "teach" the content, but rather to participate collaboratively without hierarchy in the process; this was just one of many ways we sought to do that.
Facilitator Preparation: Becoming a skilled facilitator is not something you can learn from reading material in an app; it takes attention, focus and lots of practice. All the same, we seek to create a strong training for facilitators who want to create and catalyze their own SHGs. In Phase 1, this material was very basic and in some places, non-existent, as we were focused on getting the basic content of SHGs covered and not yet on facilitator support.
Supplementary Materials: We wanted to be sure that the SHG app didn't replace or substitute for any of the paper-based bookkeeping and accounting practices of the SHGs. To ensure that they were maintaining a paper-based analog system for their savings and loans, we uploaded versions of forms and tables that they might find helpful in their bookkeeping process. There can be a temptation to use technology for everything that it is capable of; but if we had encouraged groups to share their sensitive financial data with the app, we would be decreasing the transparency of their current process, decreasing the number of people in the group who gain financial literacy through practice and exposing it to risk (via lost or malfunctioning hardware) and potentially exposing their data to theft as well.
Community: This is the social section of the app, the place where groups can add some text about who they are and upload a photo to share with other SHGs. This section also included a section for the SHGs to record audio and video stories.
Resources: This includes training material, a list of good games to use during SHG meetings, a loan calculator and a list of the financial forms and templates references within the app content as "Supplementary Material."
We designed the app with a simple UI/UX because our primary users are not smart phone literate and we knew that the majority of them would be using smart phones for the first time. Tearfund Ethiopia provided the smart phones, Vodafone Smart II’s that had been donated by Vodafone. We built the content so that no data connection was required, and because the app wasn't on the Google Play store yet, facilitators had to learn how to use a free APK Installer to manage their versions. Luckily, because our facilitator cohort was so small, this did not prove the technical hurdle that it could have been.
Field testing of the app's content within SHG groups lasted 12 weeks, or three months, with three experienced facilitators (two women and one man) each leading two groups of 15-20 young people, some in-school and some out-of-school. A project coordinator, hired and supervised by Tearfund, trained and supported the facilitators and served as an intermediary between Code Innovation's team and the implementers on the ground. The coordinator sent weekly written feedback reports to the group that the Code team would follow-up on via email. Weekly, each facilitator filled out an Amharic feedback forms for each group she facilitated. These were then translated by an external translator with no connection to the project, to minimize the potential of editorializing our primary source data. At the end of the pilot, we collected information from the youth in the SHG groups personally, finding out what they enjoyed and what they didn't about using a smart phone in the SHG process.
The feedback schedule was heavy during Phase 1 because we did not have an ICT4D specialist on the ground to input on how things were working and where we were failing. We needed regular feedback to be as detailed and real-time as possible, so that we could make decisions on project implementation to fix any issues in-process and before they became distracting to the process.
After the pilot, we debriefed the facilitators screen by screen, ensuring that our Amharic translations were up to scratch and also that the flow and content of the UI/UX mirrored the expertise of how facilitators preferred to lead the formation and growth of an SHG. We were most surprised by the overwhelmingly positive feedback from our facilitators. Aside from some content adjustments, Amharic translation issues and technical issues here and there, they were happy with what we had created and considered it a useful, valuable tool in their work.
What We Learned from our Self-Help Group Pilot in Phase 1
Overall, the Phase 1 of testing the SHG app was a success, in that the app worked reasonably well at doing what we wanted it to do, namely to help facilitators move through the process of forming a new SHG. However, we also had numerous failures that we've learned from that are strengthening the next iteration in Phase 2.
Here is a snapshot of the biggest lessons learned. We're happy to share in more detail about any of them -- just email elie@codeinnovation.com.
ICT4D Lessons Learned
These are the lessons learned that pertain directly to ICT4D's Principles of Digital Development and that we think might be most relevant to the ICT4D community.
Group learning with a one-to-many approach: At Code, we like to up-end the industrial assumption that technology should be one-to-one, especially in low-resource environments. If even one person in a community has a smart phone, it can be used as a powerful tool for change. The SHG app leveraged this same model, relying on a facilitator with access to a smart phone or tablet, as a way to impact the entire group or groups with information and guidance on how to establish and strengthen an SHG.
Facilitators determine when to use the app in SHG meetings. Our Project Coordinator was right to encourage the facilitators to use the app as they saw fit and to decide for themselves whether to use it actively in a meeting. The experienced facilitators working with us on the pilot rarely needed the detailed content prompts, but did use them as a tool to prepare for meetings and to check in and make sure they'd covered everything. This was a good move, as it avoided the technology taking center stage and distracting the group from the content and purpose, namely the creation of strong SHGs. As we start using the app with less experienced facilitators we’ll keep a close eye on how it alters the normal dynamic of meetings.
Translations done as locally as possible: It was challenging to work in Amharic for Phase 1 and our developers in Kenya hired Amharic translators whose language didn't meet the criteria of our partners at Tearfund Ethiopia. Because of this, we spent a lot of time and energy tracking and fixing text line by line that would've been better focused on quality activities like improving UI/UX and content. Next time around, we will source our translations with the input of our implementing partners.
Amharic script on Android phones: In our Community sections, where we expected the groups to create a brief introductory profile, we hadn't thought to enable the smart phone keyboards for Amharic typing. This was an oversight, but because we didn't discover the issue until the debrief session after the pilot, we weren't able to fix it during field testing. Where different scripts and alphabets are involved, we cannot assume that an appropriate keyboard will be integrated automatically.
Poor speaker quality for audio materials: Several times within the content of the app, we relied on audio case studies that the groups would listen to. However, speaker quality on the donated Android phones wasn't good enough for the groups to understand what was being said. In Phase 2, we are supplementing audio case studies with written transcripts and our partners are eager to improve the quality of the locally-purchased hardware to include better speaker quality.
Social: Our social sections were embedded within the content of the app in Phase 1, but didn't take into adequate account the privacy and security issues of our users. This is an area we'll build out and test more in Phase 2. One Hen in particular seeks future features where youth SHGs share their stories, challenges and best practices with each other, so this is something we're building slowly with each phase. We noticed, also, that it’s difficult to anticipate when a social (outward facing) activity will seem appropriate within the normally internal and private process of a group. Our efforts to guess at the right moments were not successful; so we are moving the social content out of the core curriculum and into its own area. In this way, groups that identify an eagerness to try out a social function can navigate to the appropriate section and find structure for their experience when they are ready.
Internet connectivity: We meant to build the app so that it did not require any data or connectivity to operate, but users who navigate to certain social sections found their phones trying to connect—or worse, using data. Version two will include tight controls around the phone’s appetite for data so that no user costs are incurred in its routine use. Because facilitators will have periodic access to free wifi to send their data and download new versions as we scale in Phase 2, the latest version of the SHG App will be available on the Google Play store for our facilitators to automatically download, so they don't have to learn how to manually install APKs.
Self-Help Group Content Lessons Learned
This section explores the content of the app as it follows the standard SHG process. It includes how well we were able to adapt the analog process into a digital experience, and places that we're already building out for Phase 2.
Time before first loan: After our document discovery process, we were able to draft basic materials that help a facilitator establish an SHG. Over the 12 weeks of our pilot project, we were able to test the entirety of this content, but in reality a group takes much longer to establish. Group members need time to develop and explore business ideas and to save enough capital to be ready to give their first loans. In reality, the app needs to stretch to about a year of content, although there will be a range in how fast or slow some facilitators might like to explore the content modules. This means that our field testing is going to run longer in Phase 2, for a minimum of six months, and we'll be getting weekly feedback on the filler content we develop and how it works in the field.
Cash box vs. banking vs. mobile money: Most groups in rural areas use a cash box to hold their savings and develop a consensus-based protocol to establish trust and transparency around who holds the money and how they secure it. In urban and semi-urban areas where banks are available, groups are encouraged to start their own bank account. In the future, we see SHGs electing to use mobile money as a way to hold and manage their group savings. This was content detail we didn't focus on in Phase 1, but that will be increasingly relevant as we scale and move to two-country implementation in Phase 2.
How to show impact for in-school SHGs? We realized in our pilot debrief that parents of in-school SHGs are very eager that the groups not encourage small business loans to individual members because they want their children to have extra incentive to stay in school. Because the standard metric of success for SHGs, on a group and a community level, are the small businesses that are created by its members, we are focusing on finding new ways to measure impact and success for these young people, including group income-generating activities (IGAs) that could raise the capital holdings of the group without creating distractions from school and also ways of encouraging in-school youth to support their own further education and training.
Implementation Lessons Learned
These are things we learned about how to be more effective and efficient in our working process. It's great when ICT4D projects can test and iterate quickly, and we saw areas to improve around these specific points.
Coordinator buy in to ICT4D: We realized midway through the process that our Project Coordinator was not supportive of the ICT4D process and did not think that technology should be used in resource-poor contexts because it created needless attention towards disparities. This meant that we were often having philosophical discussions about the value of our pilot and the SHG app with the person we were looking to coordinate the implementation of testing the app in the field. A project coordinator position is regularly difficult to hire for as it requires cross-cutting ICT4D and content-specific skills, and to make things a layer more political, it is usually done by the partner organization so we do not have direct supervisory control over the person who plays the largest role in determining the quality of the process. Perhaps we will write more about this in a separate article in the future, as our lessons learned about how to hire for ICDT4 projects with partners may be useful to the wider community.
Facilitator feedback: We asked for detailed feedback from facilitators after each of their two weekly SHG meetings. Despite training them on the need for and importance of high-quality detailed feedback, it was difficult to get answers from them about specific areas where they relied on their own expertise to fill in the gaps that the app didn't cover. It was also difficult to avoid duplication of their reports across both meetings; in most cases, the content was almost entirely identical. Facilitators only spoke Amharic so conversations around quality feedback had to be mediated by our Project Coordinator, who was not entirely bought in to the process. All the same, we received enough quality feedback to make important changes for our Phase 2 iteration.
We hope that this in-depth look at Phase 1 of our SHG app project is useful to ICT4D practitioners and to those working on similar m-learning projects. We are currently in process with Phase 2, scaling 1,000% in Ethiopia and Tanzania over 2015 . We'll be posting updates here and continuing to document our work for the community.
*
Thanks for reading! For more information about our work with mobile education, ICT4D and the Self-Help Group app, email info@codeinnovation.com. You can subscribe to future updates from Code Innovation here.