On Monday, we hosted The Bay Area Women in Machine Learning & Data Science Meetup Group.
We were thrilled to welcome almost 100 people to our San Francisco office to talk about how we use algorithms and machine learning to build social movements and connect people to causes and organizations they are passionate about.
If you were able to join us, we hope you had a great time! If you weren’t able to make it, here are 8 tidbits that were shared on Monday night:
Change.org has a LOT of data on social activism: we’ve been around for over 8 years, we have 130 million users spanning 196 countries. We’ve had petitions on all 7 continents (including Antarctica) and we’re localized in 16 different languages.
The median number of signatures it takes to reach victory on Change.org is a relatively modest 388.
If you try hard enough, a multi-arm (Bayesian) Bandit algorithm implementation can be optimized (It’s not easy, but it can!)
Change Politics, our newest product, brings a thoroughly exciting, data-driven approach to the US elections this year
Algorithms help us personalize a user’s experience and connect them with the activities, causes, people, and organizations they care about.
Starting a petition is a deeply meaningful and personal experience: A large number of our starters are first time starters -- they aren’t experienced advocates -- and there is a deep sense of satisfaction around the victory experience for all users, especially petition starters
President Obama is the Decision Maker on more petitions than any other DM our site.
Our President and COO, Jen Dulski, was instrumental in conducting what might possibly have been the first A/B test on the internet -- back in 1998.