Your Guide to Leading Data Science Teams: Step 3 – Processes
TL;DR
Great processes are a common factor in any high-performing team. Without structure, attention gets diverted to low-impact work, valuable learnings are lost, and growth stalls. This article explores prioritization, learning, knowledge management, data governance, and meetings—key processes that drive efficiency, improve decision-making, and enable long-term success. The right frameworks help teams scale, minimize friction, and stay focused on high-impact work. Use this worksheet to assess your team’s operational gaps. Next up: leveraging technology to streamline and enhance these processes.
Setting the Stage
What’s the common thread between top-performing creatives, dominant sports teams, and flying airplanes? If you guessed processes—bravo! Even the best in their fields thrive with (or perhaps because of) great processes. Routine allows artists to focus on creativity, marginal gains and compounding improvements drive championship teams, and checklists ensure pilots follow the right steps every time. Likewise, for your team to succeed, you need to provide structures that enable continuous growth and improvement. This is where creating great processes comes in.
But First, Why…
Are well-defined processes even important? If a team has survived so far by being nimble and agile—or by hiring smart people who just figure things out—why invest time in defining processes?
Two main reasons: scale and efficiency. At a small scale, peer-to-peer communication and individual knowledge holders might work. But as teams grow and requirements become more complex, relying solely on talented individuals will break down.
Then there’s efficiency. If your team has to rethink everything—how to prioritize, who to engage, what resources to use—every time a new task arises, friction builds. This might not be obvious at a small scale, but as needs, stakeholders, and complexity increase, it becomes a major drag. Ultimately, the right processes set your team up for success.
Ok, Great—But How?
How do you build effective processes, and where do you even start? I won’t attempt to cover all aspects of organizational processes, but here are a few key ones your team must master:
1. Workflows & Prioritization
Filter and prioritize requests effectively. Having a backlog or board is elementary; as a manager, your role includes shielding your team from unnecessary requests and demanding stakeholders. Consider:
How are you ensuring everyone is on the same page? Use measurable systems like agile sprints or another structured approach to methodically choose, discard, and prioritize work.
How comfortable are you saying no? Do you have a clear framework to decide which requests to accept or refuse? The biggest productivity hack is knowing when to say no.
Are you distinguishing between important stakeholders and merely loud ones? OKRs, KPIs, and strategy help keep priorities in check.
Example: One of my biggest mistakes in a past role was giving too much attention to the loudest person in the room, which meant that more significant work got delayed. A clearer prioritization framework would have helped me say no more often and allowed us to focus.
2. Learning & Growth
Even the best teams in the world have room to grow. In fact, the best teams might be those that learn and grow the most. Learning, however, must be actively practiced, not just aspirational. Consider:
Is learning baked in or bolted on? The easiest habits are the ones you don’t have to think about—like brushing your teeth. Build learning directly into team workflows (e.g., retrospectives, lunch-and-learns).
How do you ensure your team stays up to date? If the nature of requests or work changes, how do you make sure the team can adapt? As we’ve witnessed, the emergence of LLMs is completely upending workflows.
Who do you turn to for guidance? Having mentors, industry peers, or forums prevents stagnation. It’s incredibly powerful to have an outside voice.
Example: We implemented periodic retrospectives (adapted from agile) to ensure every review became an opportunity for growth. This meant every monthly planning, quarterly review, or other key touchpoints were also opportunities for improvement.
3. Knowledge Management
Great Knowledge Management (KM) often goes unnoticed when done well but is glaringly obvious when poorly managed. From siloed information to lost time spent re-explaining or rediscovering knowledge, poor KM can be a major drag on productivity. Your goal is to move from tacit and implicit knowledge to explicit knowledge. Consider:
How are you making this knowledge easily and obviously accessible? If it’s not used, then it doesn’t create value.
Who owns the repository? A key failure of many repositories is that no one maintains them.
Can it serve multiple stakeholders? While primarily useful for your team, it could also benefit others.
How can you standardize onboarding, workflows, and key requests? A well-maintained knowledge repository reduces repeated questions and serves as an outward-facing resource.
Example: By creating a well-maintained knowledge system, our team was able to reduce onboarding time, create standardized request templates, eliminate silos, and minimize knowledge loss.
4. Data Governance
Knowing who owns data, who is responsible for it, and what standards apply is crucial in a data-driven organization. Without clear governance, expect conflicting "single sources of truth," unrestricted access to sensitive data, and loss of control over data strategy. Consider:
Who are your data owners? This is the first step in establishing a real single source of truth.
Is your data securely managed? Compliance with GDPR, PHI regulations, etc., should be proactive, not reactive.
Can governance improve performance? Done well, DG isn’t just an administrative burden—it creates consistency and efficiency.
Example: Due to unclear data ownership and poorly documented data changes, our team spent seven months reconciling competing "single sources of truth" just to calculate one key business metric—customer count. A little upfront investment in governance could have prevented this wasted effort.
5. Mastering Updates & Meetings
Remote work is a game-changer, but it comes with a risk: death by meetings. Working effectively across distances requires skill—especially in ensuring time isn’t wasted on endless updates. Consider:
Teach your team how to run and participate in effective meetings. Pre-meeting expectations, structured agendas, and clear takeaways help tremendously.
Master async tools. Reducing unnecessary real-time discussions frees up valuable time.
Leverage AI tools for automation. Tools that assist with note-taking, documentation, and task tracking can reclaim hours.
Example: By establishing a clear set of meeting principles (pre-sharing documents, focusing on questions rather than updates, and more), we transformed long, unfocused meetings into decision-driven sessions. Meetings became what they were meant to be—a forum for decision-making.
And What If…
There’s already a workflow system in place? Great! The key question is whether it meets your team’s and organization’s needs.
People feel like there’s no time for growth? Even small changes—like occasional retrospectives—can make a difference.
Knowledge management feels like unnecessary overhead? Make it incremental. Start with just a few principles: where to store information and how to make it accessible.
No one wants to own processes? (Left blank intentionally—it’s a common reality!)
What’s Next…
Use this worksheet with your team to start defining the core elements of your most critical processes. Over time, this may help identify other areas where a lack of structured operations is slowing your team down.
Finally, look out for the next article, which will discuss the final part of this team-building trio—technology.
