12-5:30, Jul 30
Kiva
Kevin O'Brien
kevin@kiva.org
Kiva is a microlending nonprofit where anyone can lend $25 to an entrepreneur of their choosing. Over the past several weeks I've been working on an analysis of Kiva's web event data. My goal was to determine whether there are certain factors of a loan that make it more attractive to new lenders, with the aim of maximizing overall lending on the site.
When I arrived at Kiva, I sat down with my boss, Kiva’s CTO Kevin O’Brien, to discuss the analysis. I asked him a few questions about the structure of the data and about some discrepancies in my analysis and the database, and we resolved those questions.
We then went into a conference room for the weekly “Machine Learning/Data Science” discussion. I heard from some outgoing interns about their project, and from Kiva’s impact evaluation manager on a new experiment Kiva was running. I then presented my findings to the group. Many of my results were in line with expectations, but many people were surprised at the huge impact a certain sort order had on getting people to lend.
In the future, my analysis will hopefully inform some experiments Kiva is looking to put in place to maximize lending activity. Brandon Smith, a product manager, told me that they would like to see the effect of setting the “better” sort order I had discovered to be the default. If the change does indeed drive more lending, even more people around the world will receive capital to help grow their businesses and support their families.
No comments:
Post a Comment