 To collect data you need a lot of people, a lot of people to administrate the questionnaire in combat with the precious data you need. The second phase and the hardest when we produce a rigorous impact evaluation is the data collection phase. So once you know what methods you want to use, the research design that you're going to apply, you go out and collect the data going door-to-door. This is usually done either through survey firms or recruiting directly enumerators. We train them on how to interview people and how to run surveys but also how do you get to village A where do enumerators go once they get to the village. Those sort of very practical questions all have to be solved at this phase. Before going to a given village to conduct face-to-face interview, we first explain to the local authorities the goal of the survey in order to get the approval and support. One of the main challenge in collecting data is to make local authorities and population to understand that the recommendation that will result from this story may eventually improve public policy. Once we get the approval, the feed enumerators go to the households and conduct the interview by following a well-defined sampling process. At the end of each working day, we do the debriefing where each enumerator has the opportunity to talk about his challenge and the solution we found. This is also a great opportunity for each of us to share our life experiences and this was very great. We are very proud and honored to have worked with this team because we could benefit from their experience and the knowledge they have and we don't. Data is the cornerstone of evidence-based policy-making so we have to have reliable data and that is really depending on the human quality of those people who are collecting data for you. The third stage is the data analysis and the data cleaning. Which is maybe not as fun as a data collection that is equally as important.