...

From Data Scientist IC to Manager: One Year In


year, I wrote about my first 3-month manager experience. I shared some of the immediate changes I noticed, including more meetings, mentoring and coaching opportunities, a broader scope, and increased visibility into the behind-the-scenes work.

Time flies, and I have now been a manager for a full year. It is a busy year as I adapt to the new challenges — as a result, I’ve gone from writing three articles on TDS each month to just one. Meanwhile, it has been a very eye-opening and rewarding year. My team grew from three to five, and now supports a wide range of functions, from GTM and Operations to Product.

In this article, I will reflect on my first year and share what I believe are the three pillars of being an effective frontline data team manager: prioritization, empowerment, and recognition.


I. Prioritization

People management is about alignment — making sure stakeholders, my team, and I are all on the same page about what matters most and what comes next.

However, data teams these days are often overwhelmed with requests. Taking my team as an example, we have weekly stakeholder meetings to discuss new and ongoing projects; We have a #data-help Slack channel to intake ad-hoc requests; We also get pings like “urgent request, need help now” from time to time. As a result, we always have a longer to-do list than what we can realistically handle without burning out. Therefore, it is critical for me, as the manager, to set the priority appropriately and make sure every party is aligned.

What does this mean for me?

1. Understand business priority

Instead of trying to understand the specifics of every single request, I’ve learned it’s better to start with the big picture. At the end of the day, all departments are evaluated based on their contribution to the business growth, and the data team is no exception. Therefore, project prioritization should be based on the company’s focus and business impacts. I learn this from my manager and senior leadership, and I check in regularly with key stakeholders to understand what’s on top of their minds. All these contexts help me to prioritize work for my team. 

Then I prioritize the tasks based on the business impact and urgency. Generally speaking, the team should prioritize high-impact and high-urgency work, triage or delegate high-urgency but low-impact tasks, schedule and plan for high-impact but low-urgency projects, and delay or decline low-urgency and low-impact asks. Let’s see some examples below:

  1. The Sales team wants a dashboard to automate their manual quota attainment calculation. Does this project have value? Of course. It gives sales rep timely visibility into their performance, and saves someone on Revenue Operations a few hours per week. Is this urgent? Not really, stakeholders can still survive without the dashboard 🙂
  2. The same team also wants to analyze the performance of a new AI-powered automated email channel. Is it impactful? Sure. An automated outreach channel could save sales reps time and potentially lead to more conversions. How urgent is it? Pretty urgent, as this is a new initiative and we need data to understand its efficiency and iterate.

In this case, we will naturally prioritize the second project. 

Another factor to consider is the effort. This helps to understand how many tasks we can realistically take on in each sprint.

2. Delegate and check in:

One great progress I’ve seen in the past year is that once I set up the process and philosophy of prioritization, my team quickly adapted to it and gradually owned this process themselves. This is largely thanks to our clear embedded structure — each member supports a specific business domain and works very closely with the business leads, allowing them to understand each team’s priority well. Therefore, nowadays, my role is mostly to pass along my high-level understanding of the company strategies and help my team connect the dots across domains. I encourage the team to set priorities directly with their stakeholders. I often sit quietly in the cross-functional prioritization meeting, let my team drive the conversation, and step in only when they need it.

3. Be the bad guy

Sometimes this also involves protecting my team’s focus. Since my team works so closely with the business leads and has built strong relationships, they tend to accept more requests than they could realistically handle, which could lead to burnout in the long term. While I always remind them that saying no is an important skill (I learned this the hard way during my IC time), I also step in to be “the bad guy” to negotiate priorities and timelines with stakeholders. Of course, the negotiation again ties back to how each project links to the business impact and the trade-offs we have to make with limited resources.

What could I do better for prioritization? One of my biggest learning is that prioritization doesn’t have to be limited to the existing requests. A great manager doesn’t just triage requests, but also finds scope for the team: identifying high-impact, strategic opportunities and selling those ideas to stakeholders. It’s something I aim to do more of moving forward.


II. Empowerment

A mentor once told me that a key skill in management is to “give advice confidently on things that you don’t fully understand”. It might sound risky and counterintuitive at first. But to be clear, this does not mean one should pretend that they know everything. Instead, it means being comfortable making decisions and giving guidance based on incomplete information. It’s about two key manager skills to empower the team — get the context quickly and unblock the team.

1. Get the context quickly

Let’s be honest, a manager won’t be the subject matter expert on everything their team works on. But we still need to know enough to reason through trade-offs, risks, and priorities. In that sense, a good data team manager should be a generalist who knows a bit of everything. For example, one person on my team supports the Marketing team, but I haven’t worked directly with the Marketing team as an IC in the past. As a result, I had to pick up essential marketing data knowledge quickly through reading key metrics dashboards and attending marketing business review meetings. Though this does not mean I know all the details of our Multi-touch Attribution model, it helps me to understand the landscape well enough to ask good questions and offer support.

2. Unblock the team

When someone on the team is blocked, as a manager, my first step is to understand what the blocker is. If the ask is not clear enough, I can clarify it with stakeholders; If it is a technical challenge, I will brainstorm with the team, or even do research myself to find the best approach; If it is due to a dependency on another team, I can escalate the ask to get it resolved faster, etc.

Empowerment also means equipping the team with the skills they need to succeed. This, of course, covers both technical skills and soft skills.

  1. Technical skills: When I first became a manager, our employee engagement survey highlighted a gap in learning and development (L&D). Since then, I have introduced a monthly poll to determine which technical topic the team is most interested in, and then I host an L&D session to dive deeper into the topic. So far, we have covered topics including experimentation, causal inference, time series analysis, AI use cases in DS, etc.
  2. Soft skills: One way to grow the team on the non-technical front is to give them autonomy and trust. As I mentioned above, I encourage the team to lead cross-functional meetings to enhance business communication. I also give them opportunities to present their work during team meetings so they can practice in a safe and supportive environment.

Is there anything I could do better for empowerment? One area I am still learning is how to balance between autonomy and support. Sometimes I may be too hands-off — I don’t check in very often to avoid micromanaging. However, in some cases, my team might instead appreciate early guidance or feedback.


III. Recognition

Earlier this year, I went through my first annual review cycle as a manager, and I was able to promote one direct report on my team. Promotion is one of the most powerful forms of recognition. However, it is not always feasible given company budgets, team size, tenure, etc. There are a couple of additional ways I think a manager can utilize:

1. Shout-outs and kudos

I try my best to highlight the impact of the team, regardless of size. It could be an insightful analysis, a great presentation, a detailed documentation, or even a creative idea. Celebrating these wins publicly in team meetings, Slack channels, or via emails is always a great way to show appreciation and keep the team morale. I also encourage my team to give kudos to each other to foster a collaborative environment.

2. Give credit upward

It is also important to share the team’s achievements with leadership, attributing project success clearly. This increases the team’s visibility and paves the way to future promotion.

3. Support career goals

Career growth is another form of recognition. I think managers should fully understand the career goals of each direct report and help them to address the gap. Therefore, I have monthly career growth check-ins with everyone to discuss this topic. For example, one of my reports is interested in transitioning into a Data Engineer role. Since we sit in the same org as the DE team, there are lots of opportunities to collaborate. Therefore, I encourage her to work closely with DE and take on small DE tasks within her domain step by step, and keep a running list of all the DE-related work she has done to build a case for the transition.

Anything I want to improve for recognition? From the IC perspective, performance review could be a myth. Now that I have experienced a performance review cycle as a manager and got some insights, I want to help my team better understand how it works and how to better position themselves for success.


With only one year into management, I still have plenty to learn. But I’m grateful for the lessons so far, the team I get to work with, and the opportunity to keep growing — both as a data professional and as a manager.

Do you have any tips for new managers or have lessons learned from your own experience? I would love to hear your thoughts!

Source link

#Data #Scientist #Manager #YearIn