Leveling Up Your Data Science Career - Part Two
A few weeks ago, I had a great conversation with Mark, a fellow data scientist, about how to build a meaningful and sustainable career in data science.
What started as a casual chat about degrees and project work turned into a deeper discussion about what really matters in this field. I posted Part One of my takeaways from our conversation, below is Part Two.
đ¤ Where Data Science Projects Come From
Every team operates differently, but data science projects can come from multiple sources. In my experience, itâs usually a mix of direct requests from your business partners (product, marketing, sales, etc), top-down projects from your boss/leaders, and problems that you identify as worthy of solving.
Your business partners donât always know whatâs possible with data, so all data teams need to take the time to learn their area of business so they can identify the opportunities for projects that solve real problems - this can lead to your most impactful work. I find doing a regular sync with the teams you support is helpful - ask what they have going on, whatâs coming up, what questions they are pondering, etc.
Itâs also good to show off the work you do, even if itâs for another team - often that can spark a new idea. Some teams will do a regular demo session to show off their projects, or you can make sure you give your work visibility in other ways, such as sharing it in internal newsletters or during town halls, or co-present with the team you supported.
âł Balancing Big Projects with Quick Tasks
Most analytics teams balance big long-term projects (some spanning months or over a year), and smaller requests that can be turned around in a day or two. How much time you spend on the big projects versus smaller tasks will evolve - expect to do more small tasks early in your career, and take on more big projects as you get more senior.
Also, donât always expect a blueprint or template for your work - sometimes you might be given a task or project that no one on the team has ever done - and itâs up to you to figure out how to do it! This is why data analytics isnât always viewed as an entry-level field. You sometimes need the perspective of being able to scope out solutions and build the process from nothing.
đ§ The Projects That Really Matter
One common question from folks trying to establish themselves in this field: What type of projects stand out?
The vague but true answer: The ones that solved a real problem or had a business impact.
My guiding principle: Make your stakeholder the hero. What are they grappling with? What does success look like to them? What are they trying to improve? Use that as your guide.
When you can connect your work to business success - saving time, driving a key result, helping tell a story of why something succeeded (or didnât) - you become indispensable. And learning how to make that connection starts with learning about their part of the business - another plug to have regular conversations or syncs with the teams you support.
đ Measuring Impact Beyond Accuracy
One mistake I see a lot of folks make is delivering their work and moving on to the next thing - and they forget to check and see if their insights mattered.
If youâre lucky, you can do an A/B test to measure if a change had an impact, but in practice, itâs not always possible. Iâm going to sound like a broken record, but talk to your business partners - close that feedback loop - ask if they used your insights, if they implemented your recommendations, and what the outcome was.
Not only does this help you learn what matters, but it can give you something to talk about when it comes to your actual impact, and it continues to build your relationship with the business and show you are just as invested in their success as they are.
đ If I Could Start My Career OverâŚ
If I could go back, I wouldnât necessarily change anything - I used to worry that I âwastedâ my years in marketing before switching to analytics. Now I see it as an advantage - having been on the business side gave me empathy for the teams I support as well as a unique insight into how to connect insights to business value, and my experience in public relations and marketing helped my communication and storytelling skills.
However, I would have tried to branch out sooner, specifically around growing my network to learn about other career paths and to find potential mentors.
The longer Iâm in this field, the more I believe the ârightâ path isnât linear. I think doing a career pivot makes you a better candidate for the right role for you - but it can be daunting to navigate a pivot, so try to find support.
Stay tuned!
Iâve always enjoyed networking chats with other data pros, and now that Iâm recording them, Iâm loving listening to the convos and sharing them with all of you. Iâve got a few great ones that I need to edit and post - one with the author of a data science book, who has a more indirect career path than me, another with an aspiring data scientist who asked some great questions about building your career, another with a fellow social media data career content creator, and an upcoming scheduled chat with the Head of Analytics at a tech company.
Who else should I record a chat with?? Leave a comment or send me an email with your suggestions!
Really smart breakdown on what actually drives career growth in data science. The point about making stakeholders the hero is underrated, most early-career folks focus on model accuracy when the real skill is translating technical work into business impact. I'd add that proactively identifying projects based on businesscontext (not just responding to requests) is what seperates senior ICs from mid-levels. The feedback loop piece about measuring if insights were actually used is crucial but almost no one does it systematically.
Great article! We tend to think of our non-traditional background as a weakness, but now I see it as a strength, where we learn so much about the business and daily problems that people using our data faces.
And just like you, if I could go back in time, the only thing I would have done differently would be that I would have switched careers earlier :)