The first ninety days of a new job decide more than most people realise. Not because anyone is formally grading you — though probation reviews are real — but because this is the window in which your colleagues build the mental model of who you are, and that model is stubborn. The reputation you accidentally earn in the first three months can take two years to revise. Which means the early weeks reward deliberate strategy far more than raw effort, and most people get the balance exactly wrong: they try to prove themselves by producing, when they should be trying to understand.
I have watched talented people stall in new roles for a reason that has nothing to do with talent. They arrived determined to make an immediate impact, started changing things in week two, and trampled relationships and context they did not yet have. The opposite mistake — disappearing into quiet observation for three months — is rarer but just as damaging. The art is sequencing: learn first, contribute visibly but modestly, then change things once you have the standing to.
Weeks one to two: become a student on purpose
Your only real job in the first fortnight is to understand how the place actually works, which is never how the org chart says it does. Who holds informal power. Whose opinion the boss quietly checks before deciding. Where the last person in your role came unstuck. What the team is secretly proud of and what they are defensive about. You learn none of this from documents and all of it from people.
So spend the first two weeks in conversations rather than deliverables. Ask your manager directly, in the first week, what a great first ninety days would look like to them — and then write down the answer, because their definition of success is the one that counts, and it is frequently different from what you assumed when you took the job. Ask every person you will work with what they wish the person in your seat understood. People will tell you remarkable things if you ask early, while you still have the licence of being new.
Questions worth asking in your first week
- To your manager: "What does success in this role look like at ninety days, and at a year?"
- To your manager: "What is the one thing you would most like me to get off your plate?"
- To peers: "How do things really get decided around here?"
- To peers: "What should I be careful not to break in my eagerness to help?"
- To anyone senior who will spare you twenty minutes: "What do you wish you'd known when you joined?"
That last question is disarming and almost always generous in its answer.
Weeks three to six: find the early, visible win
Once you understand the terrain, you need a win — but a specific kind. Not a sweeping reform. A small, concrete, visible contribution that solves a real irritation for the team and signals competence without threatening anyone. The report nobody had time to fix. The process step everyone complains about. The introduction you can make because of where you used to work.
The point of the early win is partly the result and largely the signal: it tells the team that hiring you was a good decision, which buys you trust, which is the currency you will spend later on the bigger changes. Pick something you can actually finish inside this window. A half-done grand project at day sixty reads as a warning sign; a small thing fully delivered at day thirty reads as reliability.
The relationship that matters most, and the ones people neglect
Your relationship with your manager is the obvious priority, and the single most useful thing you can do is establish, early, how they like to communicate and be kept informed. Some want a weekly written summary; some want a two-line message when something is on fire and silence otherwise. Guessing wrong here creates friction that has nothing to do with your work. Ask them outright how they prefer to stay in the loop.
But the relationships people neglect are the lateral and downward ones — peers and the administrative and support staff who actually make an organisation function. The colleague who shows up only for the boss and treats everyone else as scenery gets quietly marked, and that mark is hard to erase. The people who control the calendars, the systems, the budgets and the institutional memory will determine whether your job is smooth or a constant low-grade fight. Treat them as the important people they are, from day one.
Weeks seven to twelve: start to shape, carefully
Now, and only now, you have enough context and enough credibility to suggest real changes. Frame them as questions and proposals rather than verdicts — "I've noticed X, here's a thought, what am I missing?" — because the fastest way to lose the goodwill you have built is to announce that everything was broken before you arrived. It usually was not, and the people who built it are still in the room.
This is also the moment for an honest checkpoint with your manager: how am I doing against what we agreed in week one? Asking for that feedback before your formal review does two things. It surfaces any quiet concerns while they are still small and fixable, and it signals that you take the role seriously. Most people avoid this conversation out of fear of what they will hear. The ones who seek it out are the ones who course-correct before a problem hardens into a reputation.
The mindset under all of it
Here is the thing that took me too long to learn: the goal of the first ninety days is not to be impressive. It is to be trusted. Impressive fades; trusted compounds. The new hire who listens more than they talk, delivers one real thing, treats everyone with respect and asks for feedback before they are forced to looks, on day one, less dazzling than the one who storms in with opinions. By day ninety, it is not close.
You will not get all of this right, and you do not need to. You need to get the sequence right — understand, then contribute, then change — and to remember that the people around you are forming a story about you whether you intend it or not. Spend the first three months making sure it is a story you would be glad to live with for the next three years.