The village Chief's band performs

Durbars are community entry ceremonies that must be done in all of the 11 zones where we are working with Mobile Technology for Community Health (MoTeCH) .  They include bringing offerings to the Chief, telling the community members about MoTeCH, dancing and hopefully getting the community members to formally “accept” MoTeCH as a valuable health service.  Durbars last for several hours, usually take place under a tree and we’re holding them for all 11 zones this week so we can keep on schedule with our launch activities.  Both of yesterday’s durbars were received very differently.  One community was very excited, engaged and asked great questions (several questions from men about whether they could sign up even if they don’t have a wife!).  The community emphatically “accepted” MoTeCH when asked by their chief.  The other community had many questions and concerns about services provided by Ghana Health Service, so wasn’t  able to engage in discussion about MoTeCH.

MoTeCH Project Manager Kirsten dances with the band

These experiences provide an important lesson, that we need to realize that each community will be different and will have a different set of experiences with other NGOs and health care providers that will influence how they feel about MoTeCH in the beginning.

Posted by:  Kirsten Gagnaire, MoTeCH Project Manager

In early 2009, Grameen Foundation went to Uganda with the idea of creating a fluid and effective two way communication channel between rural farmers

CKWs in training

and the world of agricultural experts, development agencies, traders and commercial players. Through this loop, rural small holder farmers would be given livelihood saving agricultural information generated by the experts and the big  players would keep informed on conditions on the farm from adoption of best practices to available produce for sale.

If you know much about agricultural extension services in rural Africa, you could have called such a ambition, well, ambitious. Traditionally, taking agricultural extension services to these farmers who live in far and hard to reach places is slow, expensive and inefficient because it is riddled with bottlenecks such as poor or no road infrastructure and few extension workers spread thin over vast areas.

Grameen Foundation’s technology center had a plan – harness the power of the mobile phone, a  technology that is oblivious to such bottlenecks and combine it with a network of human intermediaries that can be trained to fully leverage its capabilities  for the benefit of the farmer. The Community Knowledge Worker concept was thus born but would it work? That remained to be tested.

In nine months of testing the concept, Grameen Foundation built recruited, retained, and provided ongoing training, support, and monitoring to 38 Community Knowledge Workers operating in two districts. The foundation partnered with 7 organizations to pool expert agricultural information onto a central database and was commissioned by three organizations, including the World Food Program, to collect grassroots data on farm conditions. Together with MTN Uganda, the country’s leading mobile network operator, the foundation prototyped, tested and deployed eight mobile applications to deliver information to and from rural farmers.

The result of all these intricacies? The network of 38 community knowledge workers reached rural farmers in more than 14,000 instances either providing information to them or interviewing them for surveys commissioned by leading players in agriculture. That is; more than 8,000 times, a farmer received livelihood improving information such as  organic agricultural tips and advice developed using local knowledge, agronomic best practices as prescribed by experts, market opportunities and market prices, location and contact numbers for agricultural input dealers. On the other hand, institutional agricultural players touched based with more than 6,000 rural farmers on such crop disease incidence, potential to supply to lucrative markets like WFP and farmer’s knowledge of disease control methods.

Google G1's charging up

From the promising results of the test of concept and guided by the insights got, Grameen Foundation is moving ahead to expand this network of community knowledge worker to reach the rest of the country. Here in the Community Knowledge Worker pilot report, the foundation shares some of the lessons learned on what it takes to sustainably build and expand such a network of information intermediaries.

Posted by:  Lydia Namubiru, CKW Partnership Analyst

A Village Phone operator signs up to be an agent for KerjaLokal

On April 28, AppLab organized a training of 20 VPOs in Tangerang, Western suburb of Jakarta, on becoming Agents for KerjaLokal, a blue collar job search service that can be accessed via the mobile phone.  The 20 new Agents will participate in our initial pilot testing of the KerjaLokal micro-site and the supporting algorithm to match job seekers with jobs they desire.

The Village Phne Operators pose for a group photo

We distributed 20 Huawei 6100 QUERTY phones that have a WAP browser.  During the pilot, Agents will sign-up job seekers using the KerjaLokal micro-site.  Accompanying the phones, we distributed a User Manual for the micro-site, that explained how to access the micro-site and perform the basic functions of searching for jobs and registering job seekers to the service.  For many of the new KerjaLokal Agents, they have never experienced accessing the internet on a mobile phone, and for that matter, some of them have never had experience accessing the internet.

For that reason, it was all the more impressive to see these new Agents, many of them housewives and micro entrepreneurs with no more than a middle school education, immediately turn on the phone and open the User Manual and begin exploring how to use the micro-site.  It was great to see that level of interest and initiative to start this service.

We then had two hours of training, that included coaching them on registering themselves as an Agents.  The participants were noticeably excited when they received their SMS confirmation that they are registered to the service.  We then followed that with a small-group role-play exercise where the participants were to practice registering AppLab staff, who were playing the role of a job seeker, via the micro-site.  We could not complete the exercise because the Bakrie Telkom network in that location did not have the capacity for 20 mobile App clients trying to access the network at one time in one location.  We promptly improvised, and used an interactive PowerPoint presentation of the user interface to instruct the KerjaLokal Agents on how to register job seekers.

Program Manager Ross Jaxx leads the training

We distributed marketing posters to the new agents, and they began to discuss immediately where they should locate the poster to gain the most interest of potential job seekers.  It seems to be working.  Within the first 5 days, 13 of the 20 new agents had already registered a total of 44 job seekers on the KerjaLokal system.

Posted by:  Ross Jaax, AppLab Indonesia Program Manager

Joyce Ndago and Maria Nuela, two of our MoTeCH field staff

In December we had our first workshop to introduce and test our mobile phone technology for MoTeCH to community health workers (CHWs) in the Upper East Region of Ghana. Prior to this workshop, much of our field research and testing has focused on building content for our “Pregnant Parents” application, but today we were focused on how MoTeCH can help practitioners deliver high quality antenatal and neonatal health care.

Our goals for the workshop were to learn as much about the challenges of entering patient information on simple mobile phones using structured SMS messages and also learn more about the types of features (notifications/reminders, queries, and reports) that would be helpful for CHWs to serve their patients better. We have had discussions with some CHWs in the past about MoTeCH functionality, but we had a feeling that until they saw the technology in action, they wouldn’t be able to completely grasp what we were building and how it could help them with their daily work. Hands on demonstrations would be essential for getting their creative juices flowing to contribute to our design process.

We started the day by having about 15 participants register their mobile phone numbers with us, load their phones with airtime, and receive ‘templates’ of the SMS message formats directly on their phones. These messages were saved to each person’s drafts folder so they could easily retrieve them later and enter patient encounters in the required format without needing to enter the full message each time. Messages looked similar to the following template:

Type=ANC
FacilityID=
RegNo=
Date=
VisitNo=
TTDose=
IPTDose=
HIVResult=

With Joyce Ndago and Maria Nuela, two of our MoTeCH field staff, facilitating the session, they showed everyone how to use the information provided on sample paper registers and enter this on the mobile phone. During the registration process, the MoTeCH team gave participants training on how write text messages on their phones, for those who were less familiar with SMS. We continued this training by working side-by-side with the CHWs as they entered information from the registers onto the phone. It was amazing how quickly even those who had not used text messaging in the past managed to begin entering data in the required formats.

To simulate the interaction CHWs will have with the system with data entry triggering notifications and reminder messages, we had a modem capture incoming SMS messages. It was amazing to see the surprise and delight of the CHWs as they received a reminder that “Amina Philip (c/o Azia’s house, SAH 64, Gaani) is overdue for her fourth TT immunization”. Suddenly, MoTeCH made sense to them! As soon as the first person received her reminder from MoTeCH, she showed it around and soon everyone was waiting for their messages to come from MoTeCH also, and quickly entering information to see the feedback as soon as possible on their own phones.

One of the best moments of the workshop came when Cecelia Addah, the very first community health worker for the CHPS system in Ghana, proclaimed how useful the system would be for catching and proactively reaching out to defaulters of their 3rd IPT dose – where compliance drops off dramatically from the first and second dose and women often deliver before receiving the recommended third dose. It was amazing to see that she could see the potential of the system to help with a wide variety of use-cases relevant to her work beyond just what was demonstrated.

Some of the other ideas for how our system could assist with the day-to-day work of CHWs in the Upper East Region of Ghana were as follows:

• Notifications of all types of defaulters: Overdue childhood immunizations, overdue TT immunizations, overdue IPT and IPTi treatments, missed deworming, etc.
• Reminders to follow up with patients that had been referred from the facility to another clinic: CHWs want to follow up with the patients they have referred both to ensure the patient has actually followed instructions to receive care at another facility as well as to learn from the outcome of the referral (how the patient was treated, whether the treatment was successful, etc) to educate themselves around effective referrals for the future.
• On-demand queries to help fill in their monthly reports

Data entry was not without its challenges though. With so many different phone models, it was impossible to anticipate and provide instructions on how to do things on each individual phone. We quickly found that the Nokia 1100, one of the oldest and most simple Nokia models owned by 3 participants, did not like the “=” signs in the message templates and prevented users from scrolling past the symbol to enter the value of the field. Other phones added in extra line breaks or strange characters when the messages were sent and received by our modem. But, seeing these realities first-hand by having users use their own phones (as we will do in our MoTeCH pilot), was invaluable to ensure we come up with a good design that works for all of our CHWs.

Based on the workshop, there are several things we will need to look into operationally. Some of the questions we will need to design around are:

• How will CHWs be able to confirm that all of the messages they have submitted for the day were successfully received by the system, so that their reports are accurate. Could we potentially provide a summary SMS message each day where they can easily see how many of each type of message was submitted and cross check the totals are complete?
• What will we do about the fact that many CHPS facilities do not have electricity and CHWs cannot charge their phones regularly? Many were concerned that their batteries will die after entering just a few messages per day.
• CHPS compounds are often ‘communication hubs’ with CHWs using their own phones to receive and make many calls to coordinate their work activities. There is often only one phone per CHPS compound. How will the CHWs manage to send and receive SMS messages to MoTeCH while constantly being interrupted by calls?

The workshop ended with a fun-filled contest where participants competed to enter messages as fast as possible. The first person to submit a message without any errors won a radio. Even after the winner was declared, everyone continued fervently racing to be the next one to get finished successfully. Our winners managed to finish in just 2-3 minutes per message – not bad for their first time using structured SMS messages of this sort! Our next steps will be to analyse the data that was entered to learn more about speed and quality of the data and use this as feedback to iterate on the design of the messages, as well as develop more detailed use-cases for reminders and queries that will help CHWs with their work. All of this, and more, will be re-tested in a mini-pilot with 4 nurses in Bongo district at the beginning of January where nurses will enter real patient encounter information on their phones as part of their regular routines so we can see how the system will operate in a real-life setting. All in all, our first round of design and testing with CHWs in the Upper East was a resounding success.

Posted by:  Aliya Walji, MoTeCH Ghana Technical Program Manager

HBS Team in Jakarta

We are being assisted this week by a crew of Harvard Business School (HBS) graduate students who have volunteered to assist Grameen Foundation and its sister social enterprise in Indonesia, PT Ruma, on developing business plans for some new pro-poor products. The HBS students will spend 2 weeks in Indonesia. During that time, they will review and further develop the business model for the AppLab-sponsored jobs market application, Kerjalokal.com.  Their recommendations will fit into Grameen Foundation’s ongoing work to incubate and launch sustainable social enterprises that make use of ICT innovations to help the poor improve their lives and livelihoods.

Posted by:  Ross Jaax, AppLab Indonesia Program Manager

Village Phone Operators

As we get underway with our application development in Indonesia, we are looking at those Applications that will assist people in the informal job sector, the source of livelihoods for most poor Indonesians.  The informal sector encompasses the lower end of the labor market, for those those working as maids, gardeners, drivers, and other day laborers. It also includes millions of small and medium entrepreneurs (SMEs) who engage in such businesses as selling snack food and sundries from roadside kiosks, cooked food from push carts, and pre-paid airtime for wireless communications.

This latter product, airtime, sold by electronic vouchers, is very popular among new micro-entrepreneurs because it can be sold in small denominations, is purchased frequently offering regular cash flow, requires holding no physical inventory, and continues to grow along with overall demand for wireless communication services in Indonesia. We have learned this in part due to our ongoing Village Phone program in Indonesia, which has assisted hundreds of poor people to launch micro-enterprises that resell airtime in their communities. By studying these entrepreneur’s needs, and the markets they operate in, we have been able to refine our thinking around what sorts of applications would be most useful and relevant for their day-to-day sales activities.

With these needs in mind, we are currently designing two applications that will help automate and simplify the airtime sales process, and provide additional basic small business tools for Village Phone operators, as well as other micro-entrepreneurs who are selling physical products.

Posted by:  Ross Jaax, AppLab Indonesia Program Manager

CKW's in training

Right now we’re in a planning phase—which ultimately means we’re wrestling with the “big” challenges that become even more significant at scale.  We’re building partnerships to begin recruiting Community Knowledge Workers (CKWs) in early 2010 and that has me thinking about one of those challenges: how do we ensure that female farmers have an equal opportunity to participate as CKWs and that they have the same access to services offered through the CKW channel?

In the pilot we learned that women do most of the manual labor in farming in Uganda and will often be the decision makers when it comes to adopting new agronomic techniques.  Our observations also show that they’re less likely to own their own phones and to approach a male CKW.  During the pilot, 33% of CKWs were women, but if we really want to reach female farmers we will need to recruit equal women and men to participate in the program and ensure that both enjoy the same benefits.  So how do we do it?

It will take time to answer that question but we can start by documenting and understanding the barriers that prevent women from becoming CKWs—recruitment criteria such as literacy, fluency in English, and even the timing and location of recruitment sessions make it more difficult for women to participate.  Once they make the cut, they face additional challenges when compared with their male counterparts.  For example, Agnes, a CKW in Mbale, wakes up at 5am to go dig in her garden, where she takes care of everything from planting to weeding to pest and disease management. She returns home around 9am and cuts grass for the cow and then fetches water for the cow and the house, cleans the compound, and goes and gathers food for the day.  Around noon she prepares lunch for her family. 

CKW Sarah talking to farmers in rural Bushenyi

While we haven’t yet solved the dilemma, by talking with Agnes and other female CKWs, we were able to understand the barriers to women’s participation and the additional challenges they face as CKWs.  By starting with input of female CKWs themselves, we’ve gained insight on how we can support women, what motivates them, and ideas for how we can address gender bias over time.

Posted by:  Whitney Gantt, ICT Innovation Technical Program Officer

Health Clinic Registration Desk

How do AppLab programs get started?  How do you really understand the best way to address the problems that people in poor rural communities face?  The approach we have consistently taken for AppLab projects is to conduct a broad “needs assessment” survey at the very outset of the project.  We work with experts in ethnographic research who spend hours and hours interviewing people in the field.  The end result is qualitative data which helps to guide and inform our project work.

Sometimes, the findings validate our initial assumptions – other times they can be quite surprising.  For example, when we went to ask people in Uganda what their current sources are for information, we learned that there was no word in the local languages that corresponded to “information” – there was a notion of “knowledge” or “advice”, but not “data”.  Our MoTeCH project in Ghana originally called for simply sending pregnant women messages via SMS.  Our needs assessment research identified the importance of using voice to deliver messages as well as the role that men play in access to mobile phones and health services.  As a result, we now consider “pregnant parents” as our target audience and are specifically tailoring some of our messages to address men.  While we will be providing message content in both SMS and pre-recorded voice, our expectations are that voice will be the most popular delivery mechanism.

To read the full mHealth Ethnography Report, click here and download the PDF. This report was prepared for The Grameen Foundation by Patricia N. Mechal and The Dodowa Health Research Center in August 2009.

Posted by:  Tim Wood, Director of Mobile Health Innovation

The "MoTeCH Song" being performed

As a part of our MoTeCH initiative, we are holding content workshops to learn more about what types of information local women want and need in relation to their health. In Accra, Ghana, Eve’s Pregnancy School has offered lessons about pregnancy and motherhood to women  for over 10 years, and has seen over 2,000 mothers through safe pregnancy and delivery. The founder, Florence, gives bi-monthly classes to pregnant women and mothers; she attended our first content workshop recently and had lots of fantastic input for our program. One of the key takeaways from the workshop was that singing and music is an important part of Ghanaian culture and this class opened with the women singing a local song about womanhood, empowerment, and feminine beauty. It definitely got people excited! Part of the class was then given over to the group chanting “positive utterances” such as “my body is strong”, “i will have a safe delivery”, “my family loves me”, and “my husband loves me”. Teaching was interspersed between these more fun and participatory sessions.

These workshops, along with previous trials of interactive voice response systems in Uganda, have shown that people want the information services that we create to be fun as well as informative – not simply recorded information. As a result we are including some local songs in the audio recordings as a way to engage users. The video below shows a singing group in Vunania performing songs they prepared for the MoTeCH project about ANC, exclusive breastfeeding and importance of male involvement, in both Kassim and Nankam (main local languages). They put these together in just a day and it sounded great!

Live performance of the MoTeCH song:

For your interest, a translation of the songs is below:

NANKAM

First Song: Antenatal Care

If a woman is pregnant and agrees to go for antenatal clinics and takes all the necessary injections, on the day of her delivery she will deliver safely and a healthy baby as well. This will make her husband happy. This is an advice to all pregnant women and potential mothers.

Second Song:  Exclusive Breastfeeding

When you give birth don’t give water to the baby, the doctor said it. When you give birth don’t give water to the baby, the nurse said it. One breast is food for the baby. One breast is water for the baby. Giving the baby good health until six months.

KASIM

First Song: Antenatal Care

If a woman is pregnant, she should attend ante natal clinics for it has great benefits. If you do this on your day of delivery you will give birth to a healthy child and safely.

Second Song: Care Seeking Behavior

The health workers are here for us (the CHOs, community health volunteers and all health workers) to help us and our new born babies. All men should wake up and help us help us achieve this so we can get more help. Men should help their wives go for antenatal care and take their children for child welfare clinics.

Posted by:  Jessica Osborne, Program Officer

Calling into Question Box

During the test of concept phase of our Community Knowledge Worker initiative, AppLab Question Box (AQB) was one of the services that the CKWs provided to rural communities. Grameen Foundation worked with Appfrica Labs, and US based NGO, Open Mind, to pilot this service.  AQB is a live, local language hotline service that brings the Internet and expert advice to the homes and market stalls of individuals who may never see a computer, visit an agricultural specialist, or read in English.  Between April and September, villagers in Uganda’s Mbale and Bushyeni districts had access to the service to ask agricultural, education, recent events and other questions.

Villagers access AQB via CKWs who connect to local-language telephone operators.  Once a question is asked, an operator searches for the answer in English and calls back with the answer, translated back into the local language. The operators can answer many questions by searching pre-approved websites and doing and local database searches, but, when necessary, can also escalate questions to identified experts.  AQB currently works with Uganda’s National Agricultural Research Organization (NARO) to bring accurate and relevant agricultural information to smallholder farmers.  NARO also provided the list of approved agricultural websites and a wealth of locally relevant resources from a range of agricultural organizations.

Between April and September, the hotline received almost 3000 calls from CKWs and our analysis of the questions provides great insight into the questions that are of greatest importance to rural communities.  Roughly 2/3 of all questions asked focused on agriculture, with health, education, and news being the next most popular subject areas.  Diving deeper into the data we are able to learn more about what, specifically, farmers are asking about agriculture.  The following table provides a breakdown of the agriculture-specific questions:

This information is extremely helpful for us. For example, the fact that such a high proportion of questions focus on pest and disease control may mean that we should focus our efforts on new applications that help farmers address the most frequently encountered problems.

More Question Box users

As we plan for the next phase of the project, we are taking special care to ensure that we will be able to collect data, like the above, on a regular and reliable basis.  It is only by listening to farmers that we will ensure that CKWs are providing services that meet the real needs of the communities that they serve.

Posted by:  Adam Taylor, Project Officer