Impact not method: Unless you’re working in the research arm of tech companies (MSR, Google Brain, FAIR), or in Academia, your boss’s boss likely doesn’t care about what method you used to solve the problem. Just what the impact to the company was. So stop focusing on methodology and start focusing on how you’re helping the business. Remember that you’ve not been hired to write SQL and Python, you’ve been hired to help do one of: (a) make more money for the company, (b) cut costs for the company - which also includes helping others in the company be more effective by building internal tools.
over time, the expectation is for you to become more and more independent. Which would require you to constantly learn new things and adapt to change. Keep in mind that this doesn’t mean you should stop paying attention to using the right methodology for the right problem, but that when you’re telling your story, focus on the impact.
Getting promoted: Unless you’re exceptionally talented or exceptionally lucky, promotions won’t happen without asking for it. Have open and honest conversations with your manager about the expectations for the next level and chart a course to get there. Ask for feedback and continuously improve. As you grow in your career, especially past the Senior level, your ability to influence your peers, and leaders matters a lot more than your technical competency. So identify areas where you can exercise that muscle. It’s very different muscle from being a technically capable data scientist. A mistake I see a lot of young Data Scientists making is relying on the idea that more work will get you promoted. It won’t, at least after a point. What it takes to go from L3 (entry level) DS to L4 (the very next level) is very different from what it takes to get to L5 (Senior) and above.
Find a mentor, then mentor others: If your company has many Data Scientists, find someone who you look up to and ask them to be your mentor. Go prepared to your mentorship sessions, ask pointed questions and ask for their advice. More importantly follow-up with them and show how their advice helped you. Once you’re comfortable enough, pass on the favor to those who come after you. If you’re the only DS in a company - which I’ve seen cases of, find industry mentors. I’ve seen many forums for you to find DS mentors. If nothing works, reach out to people in the industry on LinkedIn asking for mentorship. More than half the time, they will accept.
Strike a balance between technical skills and soft skills: As I mentioned above, the further you get in your career, the less important your technical ability becomes. That doesn’t mean you can afford not to know the basics, it just means you’re valued more for your ability to influence your peers, stakeholders, and leaders than your ability to write beautiful code. So learning after a point has to span both the (almost orthogonal) axes of technical ability and soft skills. You cannot afford to be lax on learning new techniques because the world around us is changing. I started my job knowing only SQL and SAS. Then my toolkit was SQL, SAS and R. I only picked up Python ~2-3 years ago. I’m sure it’s not gonna end here. Learning is a lifelong endeavor in a field like this. At the same time, I was also learning to be a better negotiator, and a better leader.
Focus on quality of life: Remember that I have had over 12 years to do this, and I’m not even half-way into my career. Careers span 30+ years. Learn to take a breather, take vacations, take breaks and disconnect from it all once in a while. An endless rat-race will leave you burned out and unable to focus. It happened to me 5-6 years into my career where I felt overwhelmed everyday. It’s not good for you in the long run. I know this point sounds contradictory to everything I’ve said above about constantly having to learn, but it really isn’t if you put it into a 30+ year time-span. You don’t have to be the best analyst, best machine learning engineer, the best story teller, and the best leader in year 2. You can pick and choose what you want to learn based on your environment and what’s needed immediately.
Note: Yup. Thats me 5-6 years in.
Don’t gatekeep, don’t look down on others, help those who ask: Nothing bothers me more than seeing young Data Scientists/ML Engineers looking down on Data Analysts, calling them SQL monkeys. A lot of them also look down on people who do sales and marketing.