The first three months in a new role decide how the next two years feel. Here is a plan that works whether you are a junior on your first job or a senior who just changed domains.
Week 1 — get the machine running and meet people
Your only real goals this week:
- Get a working local environment. Build the project, run the tests, run the app. Write down every step you had to figure out — this is your first useful contribution (fix the onboarding docs).
- Ship something tiny. A typo fix, a log message, a test. Getting one change through the whole pipeline — branch, PR, review, merge, deploy — teaches you the process and builds confidence.
- Map the people. Who owns what? Who reviews your code? Who do you ask when you are stuck? Have a short 1:1 with each teammate.
- Do not try to understand the whole codebase. You cannot, and trying will just make you anxious.
Weeks 2–4 — build a mental model
- Follow a request end to end. Pick one real user action and trace it through the UI, the API, the services, the database and back. Draw it. This single exercise replaces a month of confused reading.
- Take the boring tickets. Bug fixes and small features across different parts of the system are the fastest way to learn its shape.
- Ask "why" questions in review. "Why is this done this way?" is not a dumb question in month one — it is expected, and the answers are gold.
- Keep a decisions log. Every time you learn how something works or why, write one line. In two months this is your cheat sheet.
Weeks 5–8 — start owning a slice
- Pick (or get assigned) a small area you can become the go-to person for — one service, one feature, one part of the pipeline.
- Deliver a medium feature end to end, including tests and docs. This is the point where you stop feeling like a visitor.
- Pair deliberately. Ask to pair with the person who knows the tricky part you keep bouncing off.
- If you are drowning in a specific technology gap — the team moved to Kubernetes and you have never touched it, say — this is the moment job support pays for itself. A few focused sessions with someone who runs it in production is faster than weeks of solo struggle.
Weeks 9–12 — deliver and get feedback
- Ask for explicit feedback. "What would you like to see more of? Where am I slower than you would expect?" Better to hear it now than at a review.
- Reduce your questions. Not to zero — to the ones that genuinely need another person. You should be unblocking yourself more often than not.
- Write something for the next person. A runbook, a design doc, an updated README. It cements what you learned and it is visible.
Things that help the whole time
- Timebox being stuck. 30–45 minutes trying alone, then ask. Longer than that is rarely productive and often just stress.
- Ask in public channels, not DMs. The answer helps the next person and shows you are engaging.
- Protect your evenings. Cramming after hours for weeks leads to burnout, not competence. Steady daytime effort wins.
- Expect to feel behind. Almost everyone does for the first two months. It passes.
When to get outside help
Get job support or a mentor if: you have a specific technology the job needs and you cannot close the gap fast enough alone; you are consistently missing estimates on the same kind of task; or the anxiety is affecting your work and you need someone in your corner who has done it before. It is not a sign of failure — changing stacks under deadline is genuinely hard.
Frequently asked questions
How long until I feel competent in a new role?
Most engineers feel functional at 4–6 weeks and genuinely comfortable at 3–6 months, depending on how different the stack is from what they knew.
Should I tell my team I am struggling with a particular technology?
Yes, framed as a plan: "I am ramping up on Kafka this week, I have started with X, I will ask if I am stuck." That reads as ownership, not weakness.
Is it normal to consider quitting in the first month?
Very common, and usually the feeling fades as the codebase stops being a fog. Give it to the 8–10 week mark before making a real decision.
What is the single highest-leverage thing in week one?
Getting one tiny change all the way through to production. It de-risks the process and everything after it feels easier.