Worldreader, Accra

Going to work in Accra is unlike going to work in other cities. First of all, you have to know really well where you are going, as there is no convenient way to orient yourself. Most of the streets do have street names, but those are rarely displayed, so it does not really help. What you need to know are landmarks. I knew that the Worldreader office would be near 'Beijing Clinic' in the quarter of Osu, which most taxi drivers were meant to know.

Fortunately for me, that morning I was on the phone with Worldreader's director of technology, Alexander Sulzberger, who offered to pick me up. He easily found my hotel as it was 'near the US embassy', a rather ugly concrete bunker of monstrous extent. We arrived at Worldreader's office, and settled down to strategize our plan for the day. The office was a humble slice of office space, shared with the internet service provider Ecoband. Several sets of Kindles were being charged here, waiting for being brought to the field, so the new students who had entered the e-reader classes would be a able to discover the joy of reading. Shortly later Joseph and Sofia arrived, and began laying out their plans for the week. Sofia was meant to go back to 'the field', the pilot villages', while Joseph would coordinate matters from Accra.

Alexander and I, however where to settle connectivity matters with Vodafone. We were off to the Vodafone office near Accra Kotoka International Airport. We met with two of our contacts, Alex from the engineering team and Daniel from the data department. The two had been working with the device and SIM-IDs that we had provided over the week from the field. Together with them we worked through the scenarios of error that I had drawn up. The challenge was to find out why some of the e-readers would never connect to mobile networks, others connected for some weeks and then failed for some weeks and yet others (the majority) never had problems.

(Watch out, we're getting technical now) My scenarios basically explored possible reasons of why the devices were behaving differently from one another. We could quickly exclude the 1st scenario, a hardware problem, as that did not explain why some devices were connected on and off. Likewise the second scenario could be disproven, it also was not an issue of network coverage. I had observed in the field that the devices that had problems, had them everywhere, regardless of their location. Vodafone had run network analytics and showed us that the coverage in the villages was just fine, despite the many hundreds of Kindles accessing the network from there. One of the pilot villages had even recently been upgraded to 3G. This had been a great success for Worldreader, because not only could children now download their favorite books faster, but also would the whole village profit from the improved mobile infrastructure. Scenario 2 was not an explanation, either. So far for the obvious ones. My 3rd hypothesis was that the devices differed in having different types of mobile SIM-cards, and indeed, Vodafone's tests could confirm that for some of the devices. They had found out that the devices that never connected were running on a set of SIM cards which were simply not provisioned for roaming. The network in Ghana would basically notice them, but then not know what to do with them. We learned that these SIM-cards could most likely be remote activated for roaming, which would fix the problem. What remained were the devices which sometimes connected for a couple of weeks, and then did not for a while. I had a 4th scenario up my sleeve, which was based on the theory that a not a technical issue, but a business process was to blame. Alex and Daniel from Vodafone could confirm that all of those SIM-cards were enabled for roaming and so they supported that theory. We went through a couple of possibilities, narrowing down the problem further. We decided on next steps, and the additional contacts to involve, as a number of departments and parties would have to work together on restoring connectivity for this group of devices.

Back at the office I recapped: We were in no bad position - from a somewhat intangible connectivity problem we had narrowed it down to two very specific issues, with clear explanations and next steps. I was confident that those would soon be resolved. My learning process had also been one for Worldreader. For those reasons Worldreader runs such extensive pilot projects; so they can learn all there is to learn about usage behavior and infrastructure, before rolling out on a large scale.

I would spend my afternoon meeting with Joseph and Sofia, sharing my experience and takeaways about how kids use their e-readers at home, and to structure our case study on the topic. Likewise we would discuss my impressions of the 'classroom observation'.

Two plans had been made, and I was looking forward to them: Worldreader would roll out its initiative, and by 2015 enable 1 Million children in developing countries to read and learn with e-books. The other plan was to go to Duncan's in Accra that evening with Joseph and Sofia, and enjoy the worlds most delicious grilled Tilapia.