Building a 30-60-90 Day Plan for a New Data Job

Building a 30-60-90 day plan for a new data job is worth the hour it takes. Research on onboarding suggests that new employees get about 90 days to prove themselves, and that the faster they feel prepared, the faster they contribute (Bauer, SHRM Foundation). Data roles have a particular challenge in that window: before you can produce anything trustworthy, you have to learn the company's data, which is usually messier and less documented than anyone admits.

What research says matters in the first months

Talya Bauer's guide for the SHRM Foundation describes four building blocks of onboarding, the "Four C's":

• Compliance: the basic rules, policies and access.
• Clarification: understanding your job and what is expected of you.
• Culture: the formal and informal norms of how things are done.
• Connection: the relationships and information networks you need.

It also identifies self-efficacy, confidence that you can do the job well, and role clarity as key levers for new employees' success. Your plan should deliberately build all of these, not just technical knowledge.

Good employers structure some of this for you. GitLab, for example, runs onboarding through a checklist whose main tasks are due within 30 days, with role-specific tasks and an onboarding buddy to ask for help (GitLab: onboarding). If your employer has nothing like that, your own plan fills the gap.

Days 1 to 30: learn

Goal: understand the business, the data and what success looks like for you.

• Clarify expectations in week one. Ask your manager: what does a great first 90 days look like? Which two or three problems matter most? How will my work be judged? Write the answers down and confirm them.
• Learn the business from the people who use data. Meet stakeholders and ask what decisions they make, which reports they rely on and what frustrates them.
• Map the data. Find the main sources, the warehouse or lake, the key tables and who owns them. Start your own notes on definitions; for example, what exactly "active customer" means here.
• Get your environment working: access, tools, repositories and how work is deployed or published.
• Deliver one small thing. Fix a broken report, answer a quick question, document a table. An early small win builds confidence on both sides.

Days 31 to 60: contribute

Goal: own a meaningful piece of work end to end.

• Take on a project agreed with your manager, ideally one of the problems they named in week one.
• Check numbers against trusted sources before you share them. Reconciling your figures with an existing finance or product report is the fastest way to find data traps and to build credibility.
• Share work early and in draft. Show a first version to a stakeholder before polishing it.
• Learn the team's standards: code review, naming, testing and documentation practices.
• Keep building connections across teams you will depend on, such as data engineering, product and finance.

Days 61 to 90: improve and own

Goal: show judgement, not just output.

• Deliver your main project and explain its result and limits clearly.
• Propose one improvement you have noticed: a recurring manual task to automate, a confusing metric to define, a missing data quality check.
• Write something lasting: documentation of a data source, a guide to a dashboard or a short note on a metric definition.
• Review progress with your manager: what went well, what to change, and goals for the next quarter.

An example plan for a data analyst

For a data engineer, swap in pipelines and on-call processes; for a data scientist, an experiment or model with a clear baseline.

Common mistakes in the first 90 days

• Redesigning everything in week two. Learn why things are the way they are before proposing changes.
• Working silently for weeks and presenting a finished product that misses what was needed.
• Trusting a table without checking its definition and freshness.
• Measuring success by hours instead of by the outcomes your manager named.

Share your plan with your manager in the first week; it shows initiative and gives you both a basis for the 90-day review. When you reach that review, how to get promoted in a data team explains how to build towards the next level, and planning a 5-year career in data covers the longer view.

Habits for checking and documenting data, essential in any new role, are part of the Data Analytics course.

Sources

• Onboarding New Employees: Maximizing Success (Talya N. Bauer, SHRM Foundation Effective Practice Guidelines, 2010) — SHRM Foundation (copy hosted by the University of Georgia)
• GitLab Onboarding (The GitLab Handbook) — GitLab