Onboarding to a data science team

A checklist for your first days on the job

By Lisa Cohen and Matt Storey

You’ve landed your data science dream job — congratulations! Now what? Don’t just assume that your first days or weeks will be perfectly organized for you. There are a number of actions you can take to set yourself up for success and make a good first impression. In this post we’ll share some suggestions for how to get off on the right foot.

Photo by Raphael Renter on Unsplash

Get an onboarding buddy

An onboarding buddy is typically a peer mentor who your manager helps identify. Meet with this person one or two times per week for the first few months, and then decrease the cadence as you ramp up. Beyond a meeting recurrence (where you can batch up your questions), you can also ping this person in chat to help unblock on quick questions at other points during the week. This onboarding buddy can help with newbie questions, acronyms, and getting acclimated, so that you can make more efficient use of the time with your manager for other deeper topics. This will also help you avoid wasting time figuring out where to find things or how to go about simple tasks.

Once you’re ramped up and feel more established, replace your onboarding buddy with career mentors. Learn more about how to do this in our previous article on mentoring.

Be a sponge

Identify priorities, gain context, learn vocabulary, and understand how the organization operates. Your first 30 days are a great time to ask lots of questions. Here are some ways to maximize this window of opportunity and absorb as much as possible:

1. Join new employee orientation or any onboarding training that your company offers. This may include mandatory trainings as well, covering topics such as privacy, security, ethics, the employee handbook, and so on. Ask for pointers to any onboarding guides, checklists, or internal wikis that might be available. Update these documents as you go along, based on your onboarding experience, to help keep them current.
2. Sync with your manager on a shared vision of success. Get the team’s mission and vision statements so that you can align with them. Share your vision for what a successful day 90 would look like, and get feedback regarding priorities as well as your approach. Document this in your company’s performance tool of choice. Ask what challenges they expect you will face so that you can start building a plan for how to overcome them. Find out the performance criteria you will be evaluated on as well as principles or expectations on how to go about your work. Ask about criteria for a successful project and some specific examples. Ask what the planning and/or sprint cycles are, how projects get prioritized, and how impact is measured.
3. Ask your manager what his or her working style is as well as background (which often influences their approach). Ask how best to communicate updates, the team’s processes for reviewing research, and so on. Find out their likes, dislikes, goals, pet peeves — you get the idea. This is also relevant if you find yourself in the position of helping a new manager of your own onboard to the group.
4. Ask about team norms that you should know about. What are the planning rhythms and “intake” processes for new projects? What are the team’s engineering expectations (such as source control, code review, live site processes, and others)? What are the organization’s meeting expectations (such as arrival time and recording policy)? What are the communication norms (i.e., channels and times of day)? Find out how the leadership team prefers to review your work (such as via slides, documents, Jupyter notebook), when to share pre-reads before a review meeting, and how to best provide updates along the way. There may be templates, team or organizational branding, or a data visualization library for you to use. Learn more generally about the organization and team culture.
5. Read about prior work on the company’s external blog and YouTube channel (if you haven’t already done so during the interview process). This will help you learn about the organization’s domain and prior approaches. Now that you’ve joined, also research internal repositories where the group publishes analyses, model documentation, and experiment results. If such repositories don’t exist, this is a great opportunity for you collect examples and establish one, which will be highly valuable to the team and future hires. Similarly, check out recent “all hands” meeting recordings to learn about leadership priorities.
6. Join internal email subscriptions and other chat channels to stay up to date with current topics. This includes employee resource groups that foster a diverse and inclusive workplace. These communities can help you meet people, offer support, and provide context about other parts of the company. (See professional networking for people who hate networking.)
7. Schedule 1:1 and team meetings with your manager and your team (if onboarding as a manager). Get invites for upcoming and recurring series that you should join.
8. Meet key contacts. Ask your manager and teammates for a list of key contacts, based on who you will engage with in your role and who the decision-makers are. (You may also be able to browse the organizational structure using internal tools.) Meet with team members, stakeholders, and leaders to understand their lens on the business or product. There’s a balance here on timing — try to do some homework first, so that you know what to ask about. At the same time, don’t wait too long and end up wasting cycles researching the wrong things. Don’t be afraid to ask questions about why things work as they do — often a fresh perspective yields new insights and can be helpful for the team as well. If meeting remotely, keep your video on to help establish a connection. Beyond 1:1 meetings, join relevant group forums (and find people who can get you included) so that you can learn from and establish yourself among colleagues. Leaning in to help others can be a way to make connections and gain experience. Finally, introduce yourself to your leadership team. Join a leadership team onboarding lunch (if available) or set up an initial “skip-level” 1:1 meeting. Use this meeting to share more about your experience and skills, primary projects, observations of the team, career goals, ideas for the future, and more.
9. Brace yourself. Starting a new role can be overwhelming, and you’re not alone if you feel that way. There’s a lot to learn, including technology stacks, products, businesses, terminology, culture, people, and organizations. (That’s why we call it “drinking from the firehose.”) Be assured, knowing that it will get easier over time.

Set up your data science environment

1. Set up your computer, log-in details, and remote access. Acquire necessary hardware, software downloads, and get tech support, if needed. Set up phone access, if needed. Join communication channels such as Teams or Slack, depending on what your organization offers.
2. Request data access on Day 1. Given that this may take some time to process (and block you in the meantime), it’s best to kick off as soon as possible. Be proactive about asking your manager and others working in the domain about the datasets you need to ensure you get access to them. Also join any necessary security groups, which may enable access to certain data.
3. Download software and tools considering the environments you’re most familiar with and what your team frequently uses (i.e., R Studio, SQL Management Studio, Git, Power BI, Tableau, Azure DevOps, Jira, Trello, and so on). Check your company’s policies to ensure that the tools you download are compliant.
4. Get a list of key websites and portals (internal or external) that you’ll need for your work and other company functions.
5. Learn the domain. Any data science analysis, machine learning model, or experiment will require understanding the domain. This helps ensure that you’re interpreting the data correctly, giving relevant recommendations, and have the context to come up with new ideas. Gain context through product training materials and trying things out yourself. (See our prior data scientist toolbelt article for more tips.) Familiarize yourself with key concepts in the dataset such as the customer definition, the broader data model, and other commonly used constructs. Ask up front about current data challenges so that you’re aware before you run into them.
6. Explore the dataset. Exploratory data analysis is a key foundation for any data science project. Get to know the ranges, distributions, and cohorts. Review existing code and queries that the team has developed in this space. Explore documentation and/or walk through it with a technical subject matter expert to ensure you understand what the fields represent. Related historical analyses are also a great resource for gaining context and inspiring future research.
7. Identify internal trainings or brownbag recordings to gain instruction on internal approaches and datasets. Practice sample queries and follow internal documentation to learn about the datasets and ensure your environment is working.

Establish yourself

A key measure of success in your onboarding is whether you become known for the areas for which you’re responsible, and establish a solid reputation for your approach and deliverables.

1. Share your picture and bio, including your past experience, expertise, and a few personal anecdotes. Ask your manager to send to the team so that others are aware that you’ve joined and have some topics of conversation with you. Share what you’ll be working on (at least whatever you know at this point) so that others know how to engage with you.
2. Be mindful of how you show up and make a good first impression. Prepare for meetings and take some time to reflect on the topic so that you have ideas to contribute. Practice self-care (sleep, exercise, diet, water, meditation, and others), so that you have healthy energy levels to engage. Do not get distracted by trying to multitask in meetings. Be sure you understand the topics under discussion and ask questions (whether publicly or with your manager or onboarding buddy separately) if something isn’t clear.
3. Speak up and share your thoughts. While you may be new to the company, you were hired because of what you have to contribute. Draw from your experience at school or in the industry to offer ideas and approaches. The things that come easily to you based on your background or approach will be highly valued in your new environment. This can include technical skills, organizational processes, and more. There is a short window before you acclimate to the new environment, so leverage that time to raise topics that stand out with your fresh perspective.
4. Deliver quick wins. This is important to demonstrate value early on and get early feedback that can help you hone your approach in a spirit of continuous improvement. Scope work into milestones so that you can share and iterate along the way as you make progress. Before long you will have a portfolio of deliverables that reflect your contributions.
5. Ensure that others know about your work so that it can have greater impact by others leveraging it. Publishing your work, putting links in relevant newsletters, presenting to the team, and hosting a brownbag about the approach are all great ways to go about this. You can also modify the content (by removing confidential information and respecting all customer privacy scenarios) and post it externally to gain broader reach. (Make sure to check any company policies on acceptable channels and processes ahead of time.)
6. Become known as a subject matter expert for the area you are responsible for. Speaking up in meetings about your area and delivering results (both covered above) are a couple of ways to go about this.
7. Ask for “360-degree” feedback from your manager, peers, business partners, mentors, team members, and others. Identify people you can trust to provide honest comments. Leverage the knowledge you gain to become the best you can be.

Congratulations again on your new role! Data science is an exciting field, with great opportunity for impact, and many career paths available. We hope this list can help ease your transition and provide reminders on key things to do in your early days.

Photo by Jeremy Vejgman on Unsplash

And if you’re the manager…

While this post is geared toward individual contributor new hires onboarding to a data science team, the same topics are relevant to managers who are onboarding data scientists.

Onboarding is a critical period for your relationship with your new team member. In a number of ways, a new hire’s first day is — so far — probably the crowning achievement of their career. This is your chance to celebrate that with them. A new hire is at their most vulnerable during those first few days and weeks. Now is your best chance to demonstrate empathy for their situation and to foster a relationship of trust. It’s your one chance to make a good first impression. You’ve invested weeks and months in recruiting the best possible talent. Don’t blow it all because you decided to prioritize a meeting that you won’t even remember by the end of the week.

Here is a checklist for managers to help ensure new team members are set up for success:

• The week before the new employee starts: Prepare the employee’s machine, monitors, accessories, desk, login password, and welcome card and/or gift.
• Day 1: Meet your new employee for lunch. Send a welcome mail to the team and stakeholders, with a picture, bio, and areas of ownership for the new employee. Make sure you or someone else is available to unblock tech questions during set up on the first day. Provide an onboarding “checklist” for the days and months ahead (you can use this article as a start!). This should include a list of required software, datasets, and access instructions. Ensure your employee is registered for any orientation and onboarding training sessions available. Introduce the employee to their onboarding buddy. Forward relevant meetings. Add the new hire to key distribution groups, team chats, and security groups. Approve data access requests. Set up a 1:1 meeting series with your new employee.
• Week 1: Communicate expectations, including a vision for success for the first month. Provide a contact list so that the employee can set up 1:1 sessions with key people. Share relevant resources that can help provide context (on the business, datasets, company, product, strategy, goals, culture, norms, engineering standards, project workflow, and performance expectations).
• Month 1: Hold a regular onboarding lunch with the leadership team for new onboarding cohorts.
• Month 2: Now that the employee has some context, have an initial session to document priorities and plans for the next few months in the company’s performance management tool.
• Month 3: As onboarding ramps down, connect the employee with some career mentors.

Onboarding is a critical period to ensure the employee’s success in the first one to two years at the company and a worthwhile investment. It also ensures you are maximizing productivity from your team. It’s a memorable period and your employee will appreciate the investment and resources you can provide during this time!