Tuesday
August 18
2009

Tayo Akinyemi

Bridging the Communication Gap in Uganda: Appfrica Labs

When the subject of successful, scalable BoP models comes up, two will inevitably be cited: mobile telecommunications and microfinance (ala Grameen). While the reach of cell phones at the BoP is indeed widespread and impressive, one still has to consider what the barriers to access are for this technology and its ancillary, the internet. Is it cost, cultural context, infrastructure, or language? My guess is that it’s all of the above. That’s why the work of ICT pioneers like Appfrica Labs is so critical. Having collaborated with Grameen, Google, and MTN to roll out a suite of SMS applications including QuestionBox, the company is poised to address these challenges and many more. Needless to say, I was pleased to have the opportunity to chat with Jon Gosier, founder of Appfrica Labs and 2009 TEDGlobal Fellow. Let’s learn more about Appfrica Labs…

Tayo Akinyemi, NextBillion.net: How did the idea for Appfrica Labs emerge?

Jon Gosier, Appfrica Labs: There’s a lot of talent here in Uganda but an extreme lack of opportunity. At Appfrica Labs, we incubate and mentor software entrepreneurs while offering them a productive environment in which to work and learn from their peers.

Tayo Akinyemi, NextBillion.net: How does Appfrica Labs work?

Jon Gosier, Appfrica Labs: We have a two part model-we’re a software company that does contract work and we’re also an incubator that provides seed capital. We don’t have external funding so we cover our own costs doing software development for short term cash, and the returns on investment provide long term cash flow.

Tayo Akinyemi, NextBillion.net: We’ve also heard about your partnership with Grameen and Google. How did that evolve?

Jon Gosier, Appfrica Labs: The opportunity came from Grameen as it was developing SMS applications in partnership with Google. They were looking for something different, another way to reach people. It was by chance that QuestionBox entered Uganda around the same time, and they decided to roll out our service as part of their pilot. It added a human element to all of the high tech stuff they were offering to people who were used to low-tech solutions.

Tayo Akinyemi, NextBillion.net: So what’s the nature of your relationship with Question Box?

Jon Gosier, Appfrica Labs: If we were to display our relationship on Facebook it would read “It’s complicated.” First and foremost, I’m Question Box’s Chief Technology Officer. I started Appfrica Labs, which incubates several projects and QuestionBox is a project we’re incubating. Eventually, it’ll leave the incubator and be its own thing, but for now it’s all under the same umbrella.

Tayo Akinyemi, NextBillion.net: Remind us what Question Box does? What problem does it solve?

Jon Gosier, Appfrica Labs: Question Box aims to democratize the acquisition of information, especially here where literacy rates are very low. We want to provide disenfranchised populations with the information they want to know. However, there are two challenges. First, many of the people that we reach can’t read, or are otherwise uncomfortable with English. QuestionBox allows them to talk to operators who speak and understand their language. Second, some of the people that we reach are too poor to own their own mobile phones, much less send SMS’s. For people who can afford mobiles, they can just call into the service. For the extremely poor, there are physical boxes placed in their villages (Question Boxes) that they can use to make calls for free.

My personal passion is using technology in ways that allows the billions of people living at the base of the pyramid to share their perspectives. Things like the real-time web (Twitter) and mobile applications (apps for the iphone) are literally changing the way the world works for a very small segment of the human population (the Global North), but there are more ways that these technologies can be applied than are currently being explored.

Tayo Akinyemi, NextBillion.net: What types of questions are people asking? What are you learning?

Jon Gosier, Appfrica Labs: The information we were collecting was fascinating and I thought it would be interesting to share it with the world. I built the site World Wants to Know to make that information public. Anyone can go there to see a real-time feed of the questions as they come in to our call center. You can also subscribe to the questions via RSS.

Some callers are kids using the service to ask about homework. Women ask health-related questions because there is very little local medical facilitation. Men overwhelmingly call to ask about agriculture and sports. They ask these questions because they’re patriarchs in the farming villages. The questions about Manchester United and Arsenal indicate that people everywhere, despite their status in life, enjoy entertainment. Every question teaches us something new about the areas we’re in and the people who live there.

Tayo Akinyemi, NextBillion.net: What drives this question-answering power?

Jon Gosier, Appfrica Labs: Appfrica Labs built a localized database that stores information relevant to a particular subject or location. It’s essentially a search engine that works offline on servers that can run even when the power is cut (using generators). The Question Box operators take calls in the languages that the callers speak. They search the database in real-time, often finding answers while the person on the other end is still talking. If the answer can’t be found in the local database, the software searches the web. Over time the search engine gets ’smarter’ because it’s indexing the answers and the questions. Thus, we get better at answering questions the more often people call us.

The information we’re searching comes from very specialized sources. For instance, Grameen gave us massive amounts of data related to Ugandan agriculture, market pricing, the local government, health and more. This is information that doesn’t exist online. It’s information that traditionally only researchers and development organizations have had access to because they were collecting it for their own purposes. In some cases the database has more information than even the local government has!

A few weeks ago there was a discussion on the website Slashdot about QuestionBox. People criticized the service for being the same as searching Google. Actually, searching the web is a last resort because there are no algorithms for making the search more accurate over time. We also collect a large amount of demographic information about the callers, the type that other organizations have expended significant resources trying to collect.

Tayo Akinyemi, NextBillion.net: The QB operator job sounds challenging. How do you train people to do it?

Jon Gosier, Appfrica Labs: The operators undergo a lengthy training process so that they don’t answer questions with opinions or provide information that can’t be verified. They’re timed, so if a call takes longer than three minutes, they ask the person to call back. Additionally, we needed the operators to speak multiple languages, so each of the women we’ve hired is conversant in up to five of the regional dialects. We’re also working on a rather sophisticated voicemail system to allow people to call back to retrieve answers.

Tayo Akinyemi, NextBillion.net: It seems like your work has implications for telecom providers at the BoP. How well do you think they’re doing at reaching customers and meeting their needs?

Jon Gosier, Appfrica Labs: Everyone is looking for new customers. Many people can’t afford to be customers nor can they read. QuestionBox bridges that gap. During a recent focus group, the overwhelming message from callers was that they enjoy the service because of the human element. All the tech heavy stuff goes on behind the scenes. Good technology is like that, it’s like your car engine. Most people don’t know how an engine works, but they find a lot of utility in what a car does. People find a great deal of utility in this service as well.

To answer your question, SMS is the way to reach the urban poor, but I’m not so convinced it’s the way to reach the rural poor. I don’t think that telecom providers are doing a good job of that. There are between 7 and 8 million mobile subscribers in Uganda out of 30 million people. Most mobile owners are the wealthy, the middle class and the urban poor. However, the rest of the country, the ’market’ that the telecoms want to reach, suffers from poverty, poor education and a lack of technical literacy. I think the ’human-ness’ of services like QuestionBox is a big draw, and it also gets back to the basics.

Rose Shuman (the founder of QuestionBox) always says, “You can get people to take one step, maybe two, but after that you lose them.” (You may recall that fellow NextBillion blogger Francisco Noguera profiled Rose and QuestionBox in an earlier post.)

To learn more about Appfrica Labs and its work with Question Box, read PCWorld’s feature, visit the company’s site check out the QuestionBox blog, or view Jon’s TED talk.

Categories
Uncategorized