A Day in the Life of a Chief Data Scientist

Author:Murphy | View: 25774 | Time: 2025-03-23 19:53:57

A few weeks ago, we wrote an article about what a typical day in the life of a Data Scientist looks like. What you may not realize if you're new here is that two Data Scientists share this blog, and our work days and roles differ quite a bit. Leah, the author of our previous article is a Senior Data Scientist, while I'm (Ray) a Chief Data Scientist.

The role of a Chief Data Scientist (also known as Data Science Manager at some companies) looks very different from organization to organization. For example, many of my peers in similar roles have a team that is dedicated to a specific business line or product. In contrast, I run an enterprise-level team that works with almost every business line in the entire organization.

Details like these have a significant impact on how my time is spent compared to my peers. My peers may have one client, while I have many. They may get to do more data science, while I get to learn about and interface with a variety of business areas. All in all, I love my work, and I get a good mix of hard and soft skills.

While every day is different, the following tasks represent a typical day for me as a Chief Data Scientist managing an enterprise-level team.

Day at a Glance

4:30–6:30 – Starting My Day
6:30–9:00 – Deep Work
9:00–9:30 – Scrum
10:00–10:30 – One-on-One
10:30–11:00 – Check Email
11:00–12:00 – Office Hours
12:00–1:30 – Workout, Lunch, Nap
1:30–2:30 – Absolute Free for All
2:30–3:00 – Check Email
3:00–3:30 – Product Meetings with Clients
3:30–4:30 – Plan My Next Week (Fridays only)

Before we dive in, I want to emphasize that the flow of my day reflects how I take part in the lifestyle design movement. I realize that not everyone will have the luxury of approaching their work this way, but it is what makes the job a great fit for me. I have a great degree of autonomy with what I choose to do and how I get it all done. This is only afforded because of the quality and quantity of work I produce.

Starting My Day

I generally start my day no later than 6:30 am. I have a sizable family, and if I don't wake up early to work out, my kids will wake me up. Unfortunately for my sleep schedule, I routinely wake up an hour or two earlier than I planned.

When this happens, I go to my local gym, get a swim in, and hit the sauna.

After I get home, I get the kids out the door for school and fire up my computer around 6:30 am. During the pandemic, I became a permanent remote worker, and this has given me a lot of flexibility in the way I structure my days.

Deep Work

Two of my favorite authors, Tim Ferriss and Cal Newport, have had the largest influence on the way I work. Those familiar with these two individuals will see some overlaps in my approach. My day is designed to manage and harness my natural flow of energy throughout the day.

I've found that my brain is firing at its absolute best early in the morning, but by the afternoon, my mental capacity has fallen off. Because of this, I've worked on ways to harness the peak hours and manage the drop-off.

I reserve my morning clarity for the hardest work I have in front of me. While the specific activities vary from day to day, some of the frequent tasks include

Reading academic papers
Solo-storming ideas for problems
Crafting tactful communications for executives or staff
Preparing for meetings
Pair programming
Coding

All of these activities require my full mental capabilities to perform them at a high level. Take, for example, crafting tactful communications. I've been coding for over a decade and consider myself very technically-minded; however, in my current role, soft skills like communication are extremely important. I've found that I need dedicated time to think through delicate and complex situations in managing up, down, and across my peer group.

In my current role, I've learned to slow down and be more mindful of how my communications will be perceived. I've been burned in the past for immediately responding to emails with whatever popped into my head first. Admittedly, these considerations don't come naturally to me, but they are a necessity at this point in my career.

Scrum

At 9:00 am, I join my team's daily scrum. In our previous article, we mentioned that we don't follow a traditional scrum meeting. Rather than saying what we did yesterday, we show it.

This is extremely powerful for me because I don't spend as much time coding or pair programming in my current role. With over a decade of experience, I've learned a lot of things the hard way. While I could share these lessons with team members one-on-one, the rest of the team doesn't reap the benefits of that knowledge. Seeing someone's work allows us all to react to it and provides an opportunity to reinforce best practices, cross-collaborate, and promote a culture of learning.

I was inspired to implement this version of scrum on my team after reading the book Creativity, Inc. by Ed Catmull and Amy Wallace. They discuss the concept of "dailies" which is a daily meeting where directors give feedback that the entire staff of artists can learn from. This allows directors to ensure consistent implementation of their style across all team members. One of my favorite parts of these meetings is I often learn something new.

One-on-One

I run a fairly large data team of Architects, Data Engineers, Software Developers, Data Analysts, and Data Scientists. The book Team Topologies by Matthew Skelton and Manuel Pais describes four types of teams:

Stream-aligned – A team "aligned to a flow of work from (usually) a segment of the business domain."
Complicated subsystem – A team "where significant mathematics/calculation/technical expertise is needed."
Enabler – A team that "helps a Stream-aligned team to overcome obstacles. Also detects missing capabilities."
Platform – "A grouping of other team types that provide a compelling internal product to accelerate delivery by Stream-aligned teams

My team is a combination of these four teams. Since we own our solutions end to end, we are designed to collaborate to deliver various data products for every business line in our organization. This means we tend to have more specialization on our team, but there are generalists as well. The stream-aligned team is frequently comprised of generalists, while the other types of teams range from generalist to pure specialization.

Keeping your thoughts organized can be challenging with a large team. For any Cal Newport fans out there, I like to keep a Kanban board with a column for each of my one-on-ones.

An example Kanban board. Image by author.

This allows me to track what we have previously discussed and generate an agenda for our next meeting. In addition, this helps me protect my team members' flow states (for more on that concept, check out Flow by Mihaly Csikszentmihalyi). Instead of sending a distracting message when I have a question for them, I simply add it to their column on my Kanban board.

In a perfect world, I meet with each team member once every two weeks. However, some team members need more attention and development, so I meet with them more frequently. Because I'm a permanent remote worker, I often go for a walk during these meetings. I have bone conduction headphones that allow me to hear my surroundings while I'm enjoying a chat with the team. I find the fresh air energizes me when I return to my desk.

Throughout my one-on-ones, we discuss a variety of topics including what's on the horizon for the team, how they fit into the bigger picture, what tasks they'll work on, and what skills they want to develop.

Unfortunately, not all of my one-on-ones are pleasant. Throughout the pandemic, there were deep discussions about managing mental health, with multiple of my team members navigating the loss of an immediate family member.

I also had to manage a very tough labor market. I lost multiple team members to salary hikes in the market that my organization was not willing to match. Through this, I learned valuable lessons on being more prepared for similar situations. The past few years have been some of the most challenging and fulfilling of my life. Having a positive impact on others means the world to me, and I hope I'm serving my team well.

Check Emails

Let me start by saying I'm not a fan of email. I've found that many people expect email to be used for rapid, back-and-forth communication like instant messaging, but my goal is to generally attend to each email within 24 hours of it hitting my inbox. Rarely, something is truly that urgent, but if it is, I can be reached via instant message or my phone.

One of my favorite hacks is to disable the sounds and visual notifications that play/show when an email arrives in my inbox. This helps me prioritize my time and get into a state of flow. I believe you should be in control of your time rather than constantly reacting to an incoming stream of messages.

Now, let me unveil my inner nerd. After working at my company for a few months, I scraped my inbox, analyzed patterns, and configured various rules to handle most emails that come through. As a result of this analysis, I've found that I can effectively address almost all emails by strategically reviewing them only twice a day. To avoid getting caught up in the time-consuming vortex of emails, I avoid checking my inbox for the rest of the day.

Before setting up these concentrated email sessions, I found that my email habits followed Parkinson's Law which states that "work expands so as to fill the time available for its completion." Switching to concentrated email sessions has allowed me to batch together very quick responses to simple emails, similar to stacking a lot of small chores together. If I encounter an email that requires my full attention, I will save it for my deep work time.

If you're interested in learning more about these concepts, I highly recommend Cal Newport's book, A World Without Email.

Office Hours

As one of the most experienced Data Scientists in my organization, I feel I have a duty to help others outside my team. Joining a team of Data Scientists at my first job was crucial for my learning and development, but I've also experienced the pain of being the only Data Scientist in a department or even an entire organization. Setting up individual meetings with Data Scientists outside of my team wouldn't scale well, so instead, I hold office hours.

Similar to how professors and teaching assistants hold office hours at universities, I block off two hours each week where anyone can join a meeting and ask me questions. For the most part, any topic is on the table. Whether it's a problem someone is facing, a request for advice on the next steps a department should take on its analytics journey, or a truly random topic, we cover it all.

I start with the first person to join the meeting and work my way down the list from there. The topics are entirely driven by the participants, and it's been fascinating to see how many people join just to learn from other's questions. There have even been plenty of times when I don't know the answer, but someone else on the call does. I can honestly say I've learned a lot from these meetings.

Managing My Energy

From noon to about 1:30 pm, I'm in energy management mode. After I eat, I get extremely sluggish in the afternoon, so I will often do some kind of light workout to get the blood moving again. Sometimes I even give myself the grace to take a nap if I'm really feeling out of it.

I don't spend an hour and a half every day doing these three activities, but I usually spend 30 to 90 minutes doing some combination of them each day.

This results in improved energy levels as I knock out my second large working session of the day. Although I don't usually have the same mental energy that I do in the morning, my dedication to energy management makes my afternoons as productive as humanly possible.

Absolute Free for All

The reason I call this time of day an "Absolute Free for All" is because I don't protect this time as much as the time in the morning. I often try to use this time as another deep working session, but I have limits on the amount of deep mental work I can complete in a day. In fact, most individuals are unable to maintain more than four hours of intense focus per day.

While a good workout or nap can often give me the energy for a second round of deep work, this time is normally taken by people who need to have a quick meeting with me to sort through something. I don't mind the interruption if I've had a productive morning. Many times these distractions bring stimulating analysis or problem-solving opportunities, which I truly enjoy.

Product Meetings with Clients

I meet with each of my clients on a similar cadence to my one-on-ones with my team. In these meetings, we typically discuss the status of the product and the next set of features, enhancements, or technical debt that we need to address in our roadmap. I'll also have my team members provide a demonstration of a new feature we have been working on.

I've found that these meetings are also an excellent opportunity to educate my client on the data profession. Since most of my clients are specialists in another area, they don't know how Machine Learning works. Education is a key part of my job.

In addition to these client meetings, I will often hold hour-long meetings once a month with the leadership team over a particular business area at my company. These meetings allow us to cover major accomplishments and discuss larger opportunities to partner together as departments.

A recent example of this is a meeting I held with the head of User Experience (UX) at my organization. We had a great discussion about sharing our roadmaps and packaging our services together. We also discussed how we can use data science to help build personas for clients and analyze unstructured text from surveys. Lastly, we brainstormed how our teams could share knowledge of our respective fields through lunch and learn sessions.

Planning the Week Ahead

One of my longstanding rituals at the end of every week is to plan my upcoming week. I like to look ahead to all of my scheduled meetings for the next week, resolve any overlapping meetings, and reject any that I know I shouldn't or can't attend. With the available time left in my schedule, I use block scheduling to set aside time for accomplishing the highest priority items in my Kanban board.

This aligns with how my mother taught me to plan for school the night before. When I start work on Monday, I feel like I have a head start on the week because I already know what I need to do. I don't waste time figuring out what to do next. I'm prepared and ready to hit the ground running!

Conclusion

Although "Data Scientist" is in my job title, the majority of my day isn't spent doing data science. After about six months into my role as Chief Data Scientist, my boss jokingly said, "I bet you didn't realize how much selling you would have to do in this job!" Companies new to machine learning are often skeptical of its value, and great sponsors can help reinforce your story.

Overall, the job is extremely rewarding because I get to have a large impact on both my organization's development as well as many individuals' career development. The cherry on top is staying up to date on the advancements in data science and discussing my vision for the future of my organization with leadership. Between the enjoyment of my work, the people I'm working with, and the flexibility of how I get my work done, I'm extremely privileged.

If you enjoyed this overview of my typical work day and are on the journey from Data Science Student to Data Science Professional, check out my workshop where I teach you the skills you don't learn in school.