Catalyst is a fully-remote organization with members living all over North America. Our floating cyborg heads are besties, but we’re privy to the power of an analog hang. As such, we host an annual member retreat. This year we chose to meet in Mexico City because it seemed like fun and because it overlapped well with this year’s CSV conf (see Zane’s talk here!)
We booked an Airbnb with a few buffer days so folks could explore the city before getting down to business hashing out Catalyst’s future. We perspired, ate popsicles, and pinched ourselves a few times to make sure we were actually there. Even though we had a lot of work planned for the next few days, we agreed that the retreat would be a success even if all we did was get together.
Work Day One
The goal for the first day was to uncover the aspects of our work that harbored uncertainty or contention. To guide the conversation, we adapted and simplified a role playing game called Microscope (reach out if you want to learn more about our version!).
The results from each round of the game became the fodder for a larger discussion about our focus for the next few days. Emulating the facilitation tactics used by our friends at RMI, we ended the game by asking everyone to share any lingering fears with the group.
Some pertained to Catalyst:
- What if we miss opportunities due to a lack of organizational direction?
- What if we focus too much on technical issues and forgo our needs as employees and humans?
Others pertained to the retreat:
- What if our ideas are too abstract to be actionable?
- How will we follow up on our discussions here?
Then, we pivoted to our hopes.
For Catalyst:
- Maybe we can inspire others to establish autonomous, democratic organizations.
- Maybe we can have a tangible impact on the energy transition.
And the retreat:
- Hopefully we’ll have a rubric for future decision making.
- Hopefully we’ll have a clear path forward for a few key sticking points.
We arranged our hopes, fears, and core themes from the Microscope game along our Airbnb’s windows and grouped them into categories. By working backwards from the most granular concerns we derived a core set of conversation topics to guide the rest of the retreat. We knew we couldn’t tackle everything in the remaining two days, so we voted on our top two.
- How will we sustain PUDL financially?
- How can we learn more about our users?
Gleanings from Work Day One: Games are a great way to model important conversations, but rules take a long time to get the hang of. In-person meetings are much more likely to run overtime than virtual ones–if you plan a whole day in 15 minute increments….well…just don’t. Airbnbs don’t usually have rules about using their windows as dry erase boards.
Work Day Two
The next morning we gave ourselves some time to explore the city, opting to work indoors during the afternoon heat.
After lunch, we dug into topic number one: sustaining PUDL financially.
In the past we’ve leaned on grants and the support of generous clients to keep PUDL up and running. As PUDL matures and more people come to rely on it, we want to ensure that it stays reliable, open, and up to date. To do this we need a more consistent source of funding and a closer relationship with users.
We broke into small groups to discuss how we might crowdsource the funds to sustain PUDL, but we found we had more questions than answers. What might a supporter get in return for their contribution? Would we consider restricting access to PUDL in any way? How would we balance the desires of Catalyst, clients, users, and financial supporters? How would we formulate and maintain an ever-changing PUDL budget.
As a team, we came to a few important conclusions:
- We don’t want to restrict access to PUDL data or give anyone preferential access to data.
- We want to give financial contributors the ability to weigh-in on important issues without relinquishing Catalyst’s autonomy over the project.
- We want our budget to reflect a minimum viable version of PUDL without any frills.
- We want to retain the possibility of raising excess funds to work on non-essential projects.
- We want to be transparent about the cost of maintaining PUDL and build trust with users.
- We want to provide more structure and expectations surrounding PUDL releases and updates.
Armed with these insights, we created a subcommittee to continue the conversation and further develop the funding model after the retreat. (We’re excited to share our progress with you! If you’re interested in learning more or contributing please reach out).
To celebrate a long day of discussion, we ordered an ungodly amount of tacos and went to a Lucha Libre fight where we heckled, screamed, and rattled our feet on the steel floor like a bunch of machista tap dancers. At least one of us stayed up until 2am in a kayfabe-induced Wikipedia cenote.
Gleanings from Work Day Two: Plan to adapt your agenda on the fly. (Hint: it’s a lot of work!) Breaks are essential (lots of little ones). Facilitation fatigue is real. Live note-taking is a life-saver (you might black-out everything you did that day). There is a limit to how many tacos you can eat in one go.
Work Day Three
We dedicated the better part of day three to discussing topic number two: gathering user information.
We love open data because it’s easy to build on top of, you don’t need to aggregate it to share it, the methodology is transparent and under constant inspection by users and maintainers, and it makes research more equitable by not charging access fees. There are a few downsides, however, including the fact that we collect almost no information about our users. This is challenging because we’re left guessing which data sources are most important, what type of organizations (with what level of technical expertise) are using our data, or what data gaps are most important for us to fill next. We thus began a conversation about how we could better understand our user base.
We emphatically rejected the idea of a paywall, but we wondered whether asking users to register with their email to gain access to the data was an acceptable middle ground.
From this arose a compelling back and forth that touched on the technical limitations of accessing data behind a registration wall and the ethos of “open data”. In the end, we found an unexpected alternative. We decided to create a user dashboard with Apache Superset (an improvement on our current Datasette instance) that requires registration to use. This will enable us to collect user stats without hindering programmatic access to the data stored in our public S3 buckets. In addition to the datasets you know and love, Superset will allow PUDL users to query our larger datasets like EPA CEMS and quickly create custom charts and dashboards. We’ll be sharing more information about our Superset deployment in the next couple of months, stay tuned!
Gleanings from Work Day Three: Move around your workspace; don’t stay at one table or in one room. Give yourselves more time to discuss fewer things. Teams that cook together, stay together–but restaurants with long lines are also probably worth it. Don’t forget to stock up on local snacks before you leave!
Take-Aways
We didn’t get to discuss everything we wanted, but we did accomplish our retreat goals and avoid our worst fears. Perhaps most importantly, we deepened our relationships as coworkers and friends. We have a lot of niche interests in common, and it was fun to nerd out and explore together; many of us decided to stay in Mexico after the retreat and continue hanging out!
Aesthetically, this retreat was the opposite of the last year’s gathering in the mountains of Colorado. Who knows, maybe next year we’ll pendulum back into isolation. I’ve heard rumors that Alaska is on people’s minds… perhaps a camping trip is in order? Get ready, Ben (our resident Alaska co-oper), we’re coming for ya!
Until next time–
Love,
Your favorite deranged, dedicated data detectives: Austen, Ben, Christina, Dazhong, Ella, Katie, Trenton, Zach, & Zane
