As DonorsChoose.org grows so does the complexity of questions that we ask of our data. The volume of questions has also risen sharply. From the beginning, we were obsessive about democratizing access to our data through Looker, running on-boarding training, building staff support community on #Slack and involving project managers in investing their teams’ time in being self sufficient with data needs. Our mission was simple: make our two data scientists obsolete. Despite best efforts, we were still too involved in supporting daily business needs that we strongly believed others could do better than us!
So we opened a data analytics course to the entire staff. We called it Data Masters. Keep reading this post, or skip to the course deck.
We knew we had to teach 3 things for the org to be self sufficient with data:
- thinking like a data analyst
- coding and in-depth understanding of how Looker works
- planting a data master on each team that would help answer their colleagues’ burning data questions
WHAT?
It’s a 3-month long, 6-part data bootcamp to teach you coding and thinking like a data analyst, followed by you applying your skills to your daily work.
WHY?
For colleagues: this is a chance to take your interest in data to a master level, with focus on transferable, hard skills, including coding.
For the org: with your newfound skills, you are committing to being the front-line responder to data questions from your team members, helping them to achieve their goals faster.
For the data team: shift focus from internal business intelligence to publishing and research.
HOW?
Curriculum will largely be driven by participants. All skills will be learned on our data in the context of your team’s questions and will be immediately applicable.
CURRICULUM
- Month 1 (Asking questions): Databases; SQL Runner; Coding in SQL; Git; Thinking like a data pro: state hypotheses, split complex questions into steps, challenge assumptions
- Month 2 (Answering questions): Looker frontend; Looker backend: LookML; Dashboards; Docs; Data visualization; Calculations
- Month 3: Analytical process; Designing experiments and running statistics
We put together a quick application, asking folks for what they were most excited to learn and what their current pain points were. That helped define the course. The material was taught through examples where each person would work independently for 5-10 mins, then someone would volunteer to show to the group how they solved it. They then were involved in a process of peer teaching. It was glorious. We learned that people did not like working in groups. Ok. Ramble ramble… instead, check out this deck with practice questions, application challenges, analytical process, course philosophy, looker and sql learning resources and much much more. It’s all in here.
