Building Powerful Data Teams: On Investing in Junior Talent
Updated: Apr 7, 2021
Coding isn't necessarily a difficult skill to learn. I believe that anyone who truly wants to learn how to code, can. But whether or not people are set up for success in this journey is an entirely different story. Imagine yourself when you were first learning how to code. Whether it was in a college course or on your own through Stack Overflow, you no doubt had some "stupid questions", a lot of confusion surrounding the myriad of acronyms, and many mistakes under your belt (anything from the humble typo error to the "did I close out that AWS instance?").
The difference between the current you and the past you is this: you were able to make mistakes, you learned from them, you moved on to other mistakes. As a Data Director and manager of a 5-person data team, I'm obsessed with how we can more efficiently, and with a lot less shame, go through that cycle of learning from mistakes.
I took over my Data Team in January of 2020 with 3 juniors. It was all of their first data jobs. None of them had significant experience writing any sort of code. Terms like "data warehouse" and "schema" were brand new concepts to them. And facing us was the simple task of building out Sunrise's entire data infrastructure from scratch.
A year later, I would boast that I have one of the greatest data teams on the left. My team is writing sophisticated, clean code that rivals people twice their age. They've created powerful tools, training guidelines, and systems that bring new innovation to the progressive space (that we've often shared with other orgs). And there's no challenge our organizers have thrown their way that they haven't been able to solve.
So how did we do it? Below I want to share a sort of step-by-step recounting of how I built a culture of learning and growth on my data team and spent a year investing in junior talent. My hope is that other managers adopt a similar attitude for their data teams, and that you also can share with me what you have done for your teams!
Note: I use the word "junior" here instead of junior engineer/analyst/dev. I run an interdisciplinary team of generalist, and we're proud of it! Junior here means a junior data person, but you could sub in the flavor of data professional of your choice!
Building a Data Team: Step by Step
Shifting the culture
I made it clear to my team that I wanted them to prioritize professional development during their working hours. I did not want them working a full time job (in the middle of a pandemic) while feeling pressured to juggle professional development outside of work hours. There would be no weekend coding projects on my team. And in order for that to be true, I knew I needed to change the culture of work on our team.
Blocking off time
I instructed each person on my team to block off at least 3 hours during their work week that they could put their heads down, ignore the barrage of Slacks and emails, and do deep work to advance their professional development.
Celebrating taking the time
During weekly management meetings, I would celebrate and uplift anyone who was able to meet their 3 hour threshold for professional development, even if it meant they had to delay other work.
Reflect upon and discuss roadblocks
And if, during management meetings, it came up that my junior was having difficulty taking time away from their projects to focus on themselves, we would open up a space for honest and nonjudgmental reflection on what threw them off in the last week. If it was a question of overloading and project management, we would rework deadlines and expectations. If it was a question of focus and time management, we would go back to the calendar, block off some time, and talk through the anxieties they felt around not attending to their other work.
Being okay with failure
It was through these management meetings and open reflections that it came out that my juniors were afraid of making mistakes and, worse, afraid of disappointing me, their manager. I then took a few days to think through all the ways I contributed to a culture of perfectionism and a fear of failure. For one junior, we introduced "Failure of the week" to our agenda where we each opened up about a failure from the past week. For everyone, I talked openly and often about the mistakes I had made in my journey to being a Data Director, in my complete faith in my team's ability to grow into talented data professionals, and the concept of "failing fast to succeed sooner."—That I needed them to be at a higher skill level than they were currently, that there wasn't a single doubt in my heart that they could do so, and that they needed to be okay with failure early in their career so they could learn from their mistakes and grow.
Putting resources behind the team
After creating a team culture where it wasn't only okay but celebrated to prioritize learning, I needed to move organizational resources to support my team in this work.
Hiring a tutor
I got extremely lucky. I knew I wanted to hire a contractor who could meet 1:1 with each of my 3 juniors and tutor them in SQL, Python, and all things data. What I did not expect was to meet the absolute best person for the job within minutes of making a Slack post in a progressive data community. Sunrise is incredibly blessed to have Jason Wiener on board with us as a contract mentor/tutor for my juniors. Every week, he sits down 1:1 with all 3 of my juniors individually to tutor them in a self-directed curriculum (with heavy input from me as their manager). Jason is not only a skilled data professional himself, but has both spent time in the progressive data world and experience teaching at a college. My team has made it clear that their time with Jason has been one of the best opportunities they've ever had to grow professionally.
Professional Development and Technology Funds
I am also lucky that I work at an organization that also believes in investing in people. Everyone at Sunrise has access to a pool of money for professional development and a separate pool of money for technology expenses. First, I advocated for my team to have an increased technology fund given the natural of their role (we can't run scripts on a Chromebook!). Second, I took the time in my management meetings to ensure everyone on my team knew about the two funds and how to access them. I then sat with each of my team members and made a plan on how they would spend their professional development money—some wanted to attend online data bootcamps, others bought statistics textbooks, and still some other used their resources to purchase books to learn more about the intersection of data and racism.
Being okay with this taking time
I started at Sunrise in January of 2020. The first few months were spent solely on transitioning our CRM, and so it was only in March of 2020 that my team officially kicked off as the Data Team at Sunrise. I made it clear to my team then that I did not expect them to know how to do all parts of their jobs then, but that they were going to learn. I understood that this process was going to take time, but I had to do more work to undo their internalized workplace shame and to set up structural guardrails to support them in their journey.
Shifting my role
I hold the strong belief that it is on me as a manager to provide my team with the support, resources, and guidance to be good at their jobs. In late 2020, I realized that there was one last piece to the puzzle that I had been neglecting. For most of my time at Sunrise, I had spent the bulk of my time directly contributing to projects and holding most of the high-level technical work at the organization. I realized that if my team were going to continue to grow, I needed to give away some of the work I was holding and transition into more of a behind-the-scenes, full-time management role.
Became a full time manager
Supporting a team takes time. It takes significant time. And I ultimately realized it took more time that I had to give. Many months into my time with Sunrise I realized I needed to shift my role away from direct contribution work towards more of a full-time management role. This realizations kicked off a series of sliding-puzzle decisions to juggle the mountain of work on our teams plate at the time. It started with a promotion of one of my juniors to Deputy Data Director, and then a hire underneath her to backfill the work. The transition worked, and with a Deputy leading more the day-to-day coding work of the organization, I was able to devote 100% of my attention to supporting my team.
Code reviews, pair programming, and more
In my new management role, I set new expectations with my team. We would all (finally) learn git and GitHub and all major code would go through a Pull Request/code review system. I would also make myself available for up to 2 hours every week per person for pair programming. I also implemented a brand new project management system featuring Asana that would allow me to have more insight into the work of my team and in turn provide them with more feedback and support.
Could I have just hired juniors that knew how to code? Maybe. I can't say for certain if that would have been a viable option because it never crossed my mind. I had the opportunity of a lifetime to invest in 3 junior data people, build them up into something great, and set them up for success later on in their career. I also work at an organization that shares my belief in investing in young talent.
But there's another reason I took this work so seriously and spent so much time, money, and resources to build up my team: I am only where I am professionally because of the investment other people have made in me. I started off in this field as a fundraiser with a distant dream of working in data. After discovering progressive analytics, I cold LinkedIn messaged one of the cofounders of Data for Progress, Colin McAuliffe, with a couple of humble questions about entering the field. For some reason, the Colin saw something in me and invited me to the Data for Progress community Slack. From there, I spoke openly about my interest in a data career and asked a lot of really silly questions in the #coding-questions channel, in which extremely patient, more experienced data professionals took the time to answer. Eventually, someone noticed me and offered me a grant to conduct research on what I called "the intersection of data and climate." This was my first real data project in the field, and from there I was given more opportunities to prove myself and received more investment (and a lot of links to Stack Overflow) from the community.
It was only through this series of random and more established data professionals seeing something in me—believing in me—that I wound up in the incredible position I am today as the Data Director of Sunrise. And to turn around and my team that same opportunity, that same faith and compassion...that is the ultimate realization of my role as a Data Director.