Your Guide to Leading Data Science Teams: Step 1 – Purpose and Vision
TL;DR
I’m writing a set of articles as a “How-To” guide to leading data science teams based on my experiences over the last few years. I’m starting with Defining the Purpose and Vision of a Data Science Team because this is the first step to creating a great team and crucial to get right to make sure your team flourishes in the right areas. You can do this in 5 steps outlined below, which should cover most of your needs, and more reading is available should you want to dive deeper into the subject. I also made a template you can use with your team.
What are all the things I need to do to build a great data science team?
In what order should I approach these tasks?
How will I know that I’m moving in the right direction?
These were the questions I remember asking myself when I first started at Komaza—East Africa’s largest forestry social enterprise—in my new role as Director of Data Science a little over four years ago. I recall scouring the internet for cookbooks, O’Reilly manuals, blogs, or other resources to answer these questions, never quite feeling satisfied with the answers. To be clear, there were many good books, and I’ve relied on many of them—but I wanted something more concise and actionable, like a checklist I could refer to when engaging in this process.
Now, four years and many experiences later, I’m writing a guide for myself about how I approached these questions and the things I’ve learned along the way—if only to grow, improve, and (hopefully) avoid making the same mistakes again. I plan to share this guide piece by piece in a series of articles over the next few weeks, along with worksheets designed to complement the articles.
Data science and team leadership are vast topics, so there are many things I won’t cover. That said, I’d love to hear your feedback—whether this resonates with you, is useful, or if there are gaps or nuances you think I should address.
For my first article, I’d like to tackle what I believe is the foundational step in building a great data science team: Defining the Mission, Vision, and OKRs of a Data Science Team. Let me know your thoughts!
But First, Why…
Is defining a purpose and vision even necessary? Isn’t the goal of any data science team simply to provide great models and insights to the business? What use would a more complicated definition serve?
Well, consider this:
What do “great models and insights” mean for your organization?
What are you going to spend your time on?
How do you know your team is moving in the right direction?
If you don’t know how to answer these questions, you’re likely to spend time on the wrong things or invest in areas or people who won’t meaningfully contribute to the organization. At best, this will feel like spinning your wheels; at worst, it could land you in trouble. (Even teams that go through mission and vision planning can still fall into these traps!)
So, simply building great models isn’t sufficient. Ultimately, it’s better to give an approximate answer to the right question than a perfect answer to the wrong one.
Ok, Great. But How…
do you go about defining these things? The desired outcome of this process is threefold: 1) your team’s north star (mission), 2) your team’s aspirations (vision), and 3) a concrete plan to achieve these (roadmap and OKRs). I approached getting these answers through the following five steps:
1. Understand Organizational Goals
First, you need to understand the broader picture—what’s happening, what’s important, and what’s not. This step may be more nuanced than expected, but it’s the best guide for ensuring your team focuses on work that will be recognized and supported. The aim here is to establish a 10,000-foot view.
Consider:
What is your organization’s north star, and what are its priorities?
Are there company-level OKRs or roadmaps that can guide your focus?
What recent events have shaped your organization’s current state, and what’s keeping leaders awake at night?
Example:
When I joined Komaza, I quickly learned the top priorities were generating revenue and scaling the value chain to secure the next funding round. Data, though seen as valuable, lacked ownership, resulting in inconsistent definitions and misaligned figures across departments. This context helped me establish the top-level priorities that would inform my own planning.
2. Understand How Your Team Will Contribute to These Goals
Once you’ve established the big picture, focus on how your team specifically moves the needle on these issues. Does your team develop a core product component or inform a major strategic direction? Your goal here is to clarify your team’s role in achieving organizational aspirations.
Consider:
What specific needs was your team formed to address?
Are there org-level OKRs where your team plays a critical role? How much control does your team have over these OKRs versus reliance on partner teams?
Is your team’s purpose to develop a product, create efficiencies, or something else?
Example:
To grow our revenue at Komaza, we aimed to sell Carbon Credits, and achieving this required close collaboration between the Corporate Finance and Data Science teams. Additionally, to scale the operations of the Harvest team, new data systems needed to be designed to replace the current manual process. This gave us clear insights on where specifically we’d be investing our efforts.
3. Understand the Organizational Context
Simply knowing your team’s objectives isn’t enough. Organizational dynamics—like stakeholder relationships and perceptions—play a significant role. The goal here is to navigate these dynamics effectively to support your team’s success.
Consider:
What type of company are you joining? Expectations differ based on the company’s growth stage—how you act and collaborate in a startup is going to differ vastly from that of an established company. The book The First 90 Days has some great notes on this.
What’s your team’s history, and are there any expectations or baggage?
Who are your stakeholders? Are they allies, neutral, or adversarial?
Which other teams or projects influence your success?
Example:
I quickly learned that our success as Data Science depended on delivering co-led initiatives with other departments. Consequently, building relationships with these teams was critical. Additionally, the organization I was with was transitioning from the scrappy startup phase to scaling operations, necessitating more structured systems and processes.
4. Assess Your Current Capabilities
Turn the analysis inward to evaluate your team’s current abilities. Delivery of goals—large and small—depends on capability as much as intent. So, it’s crucial to understand what your team can do versus what is expected of it.
Consider:
What are your team’s current capabilities in terms of people, processes, and technology?
What gaps exist between current capabilities and short-term (OKRs) and long-term (vision) needs?
What’s your organization’s data maturity, and where can your team provide the best ROI?
Example:
Through interviews, I assessed team skills, existing platforms, and processes. Without a central data team, we lacked key elements like governance, project intake processes, and clear communication channels. This evaluation highlighted skill abundances (e.g., analysis) and deficits (e.g., data engineering), directly informing hiring needs and giving us clear insights into where we’d need to invest.
5. Define Your Mission, Vision, Roadmap, and OKRs
Finally, for the tricky part—synthesize all the information to define your mission and vision, as well as a roadmap and OKRs that help you move towards said vision. This step uses the data you’ve gathered to align your team’s work with organizational goals and aspirations.
Consider:
Mission: What is your team’s fundamental purpose?
Vision: What does your team aspire to be? Are you trying to be the most bad*ss data science team in forestry/fintech/(your industry here)?
OKRs: How will your team contribute to the organization’s goals? What goals are you delivering that will help meet your organization’s key needs, and how will you grow to meet the need?
Roadmap: What steps, projects, or strategies will move you toward your OKRs and vision?
Example:
After clarifying Komaza’s current state, the team and I were ready to answer key questions about ourselves and establish a framework to prioritize our work. By understanding what Komaza expected and where we could make the most impact, we defined our mission. Knowing what we aspired to as a team, we set our vision. Finally, by aligning with the firm’s short-term priorities and assessing our circumstances, we created OKRs and a roadmap to guide the year ahead.
And What If…
What if my organization doesn’t have clear goals or OKRs?
If your organization lacks clear goals, focus on understanding the broader mission and identifying pain points that your team can address. Talk to key stakeholders to uncover implicit priorities and align your team’s objectives accordingly. This process might take extra effort, but it’s an opportunity to establish your team as a strategic partner.What if my team is too small or new to define a vision?
Even small or nascent teams can benefit from a clearly defined mission and vision. Start simple—define what success looks like in the short term and gradually build your vision as the team grows. A lightweight roadmap can help prioritize immediate needs without overwhelming your resources.How do I balance delivering results with building long-term capabilities?
Balancing short-term and long-term goals is a common challenge. Use your mission and vision to guide decisions. Focus on delivering quick wins that align with your long-term aspirations, and be transparent with stakeholders about the steps needed to build lasting capabilities.What if defining a mission and vision feels like a waste of time?
It’s tempting to dive straight into execution, but a lack of clarity around mission and vision can lead to wasted effort and misaligned priorities. Think of this process as an investment—time spent upfront saves far more time (and headaches) down the road by ensuring your team is focused on the right objectives.What if our goals change frequently?
Frequent shifts in goals are common in fast-paced environments. A well-defined mission and vision provide a stable foundation, even when priorities shift. Treat your OKRs and roadmap as living documents, revisiting them regularly to adjust to changing circumstances while staying aligned with your overarching purpose.
Is That All?
Of course not! Entire books are dedicated to this topic. Consider these sources:
What’s Next?
Please use this template to work through these questions with your team - I’d love to hear your feedback on how useful you found it. And stay tuned for the next series of articles on this topic!
