Things I Learned Doing Mobile Work in Africa

From DevSummit
Jump to: navigation, search

Facilitated by Camille Ramani


Camille recently returned from a tour working with Refugees United in Nairobi, Kenya. In this session, she'll relate what she learned and what surprised her during her time across the Atlantic.


Camille spent 6 months in Nairobi building mobile tools for Refugees United


  • Camille
  • Tomas
  • Dirk
  • Seth
  • Dan
  • Dwalu
  • Hillary
  • Dorrit


Camille has a background in non-profit programme, and is not a technologist Accidental techie, took a job with Refugees United (Copenhagen-based non-profit)

On the ground in Nairobi to implement an existing tool with 2 local partners

  • Kenya Red-Cross Society
  • Refugee Consortium of Kenya

Working in 2 large refugee camps in Kenya (Dadaab, Kakuma)

The Tool itself

A web-based tool and a mobile tool. Sometimes explained as a "facebook for refugees" Free online tool, for refugees to create profiles (that could be partially anonymous), i.e. people can put in information about themselves, but this information could be nicknames, private details (scars etc.). The premise is that people can then find individuals based on these profiles.

Working with 3 populations:

  • Urban Refugees - Often illegal, living in major urban centers
  • Internally Displaced People - Internal refugees having moved from one part of the country to another
  • Refugees in Camps - People who are not allowed to work, just living of aid


  • Each group thinks very differently about this (Congolese, Somalians, Sudanese etc.)
  • A lot of cultural differences weren't built into the tool
    • Names of fields in the tool were sometimes strange
    • Example: What is your primary e-mail where the word primary relates to primary school
    • i.e. I didn't have an e-mail when i was in primary school
  • The tool had already been built before this information was gathered on the ground
  • Mobile tools are extremely useful, and broadly used (example m:pesa) but, that doesn't necessarily translate to non-profit sectors.
  • Some people don't necessarily want to find their family,
  • Articulating the value proposition to refugees can be difficult
  • RU was working with community monitors who were in the field working with refugees, but even these sometimes ran into tribal issues, since people didn't necessarily trust people from their own countries if they were from different tribes
  • Using a technological tool in a relationship based society can be difficult, and explaining the value proposition in terms that make sense culturally is extremely difficult.
  • Stigmas relating to the atrocities that some refugees went through mean that they didn't necessarily want their families to know what they'd been through.
  • There are so many subcultures layered upon each other that it can be impossible to map out and react to.

Understanding the individuals needs and contexts

Many Amazing anecdotes were told.....

In education we are good at thinking about psychology, sociology and cultural contexts, but when delivering tools to adults we forget these same things

The Operational side

  • The main tool was to built for a phone with a browser, and this didn't work for most of the refugees. Connectivity was a problem. Screen sizes were minute, and the tool wasn't built for this. Logistically this was a huge problem.
  • The tool worked in urban areas, and when it worked in camp, it was through using the youth as proxies to explain the issues. Going to colleges and getting young volunteers to explain and leverage this issues.
  • Working with women: Large communities of conservative muslims in the Somalæi communities. It was very difficult to recruit female monitors, and the male monitors could not talk to women.
  • In traditional societies you must not underestimate the role of the village chief/imam/pastor/elder etc. They often need to give all permissions.
  • Ramadan


Never come to a project with the preconceived notion that a mobile app is the solution. A different approach vector is needed.