I graduated from college with a degree in Operations Research and Management Science. While it sounds really impressive to take more than two seconds to say your major, it left a lot of people confused, myself included. The Data Science major had just started being offered my senior year, but through my major, I ended up taking all the classes necessary for me to start a career in data.

I started off with a part-time job as a research data analyst at the Carnegie Mellon CyLab. It was my first time experiencing data outside of the sterile confines of academia. While at school, professors tried their best to give us exposure to raw data, but you always knew that there was an answer, a path forward. It wasn’t realistic. While performing research, the skills I learned became the foundation on which I built my career. It wasn’t enough to be able to extract and clean data, like I had in school. I needed the data to tell a story. The findings I presented every week had consequences. It was my first taste in helping shape a narrative. It helped me put data into practice, learning what was noise and what was a story.

After a year, I got my first full-time job at CircleCI. From the beginning, even being new, I was asked a lot of hard questions. And, “I don’t have an answer for that” was not a valid answer. Let me describe the reality of data at CircleCI. Every company has something called technical debt, but I’d like to present it to you as technical history. When I first joined CircleCI, we were sunsetting 1.0 and moving to 2.0. As with any major update, the things of the old don’t translate directly to the new. To serve the needs of the company, we had to carry along some of that history in terms of metrics. And it was up to me, the keeper of data, to become a historian and create bridges to help people along the journey.

So what was that like? Well, early on, one of the first questions the VP of Product asked me was, “What’s your backlog?” and I almost got offended. To me, still in the academic state of mind, it was a travesty to not complete my assignments on time. I thought he was questioning my performance. But to him, this was a normal and expected reality. In a startup that’s growing fast, there are more questions than there are people to answer them. Coming onto the data team effectively grew the team by 25% (from 3 to 4 people). There definitely was a shortage of people who could answer the questions. It’s the perennial problem of scarcity.

Having a degree in Operations Research and Management Science, I saw it for what it was, an optimization problem. I had to learn to wade through the backlog with a few tools. These tools helped me not just navigate prioritization, but also helped me to navigate the sea of data I was trying to tame.

Here are four lessons to help data analysts stay afloat:

  1. Build trust
    Be transparent about your processes and decision-making. As the data analyst, I regard myself as one of the messengers of the data. Our fates are tied. When something breaks, I become the harbinger of bad news regardless of whether or not I have any control over the matter. Since data touches every part of the platform, understanding impact is a cross-collaboration effort. Resiliency is hard when you’re one person, but if you have a band of people you trust, and who trust you, you can be successful.

  2. Data doesn’t lie
    Speak up. When I needed to speak up, I did, and I found that my voice mattered. It was because I wasn’t speaking about opinion. I was speaking the language of data and I was confident about what the data told me. That was what I was here to do. Through data, I could focus the voices, sort out the opinions, and direct us toward what was valuable and actionable. With data, it changed the mood of the meetings from what people wanted to do, to what the data told us that we should do. It becomes about skill and storytelling.

  3. Be willing to learn
    The world of data is new and exciting. Being in a startup is new and exciting. Working with a product that gets continuously better is new and exciting. When everything around you is changing, you have to be proactive about learning all the new things, and it helps to commit to this everyday. Being knowledgeable of your company and its product makes the work of interpreting data less daunting.

  4. Take a break
    Whew! I may love data but there are times when I just need to take a step away from the work and breathe fresh air. For that, I’ve been using my PTO to pursue my personal passion for mentoring college students and traveling. I’ve volunteered at UC Berkeley (my alma mater) and at University of Washington as well as traveled to Korea and Japan while working at CircleCI. And whenever I come back, my personal motivation levels are recharged so I am ready to tackle the next data problem.

When it comes down to it, data is about people - how you interpret actions and events, how you quantify behavior, and how you share stories and trends using numbers. Remember that as an analyst you play the role of historian and storyteller, and there’s a level of subjectivity in that. With these four lessons in mind, it becomes easier to prioritize data in discussions, learn as you go, work through large data sets without getting burned out, and provide value to your team and your company.

Learn about what processes can be used to make data science, machine learning, and AI code more reliable with CircleCI.