What should you actually do in the first 90 days of a new job?
Learn before you change anything — because the credibility you spend early cannot be earned back, and the most common failure of a new starter is arriving with answers before understanding the questions.
The first two weeks: understand the system.
Meet people individually, widely, including outside your immediate team. Ask what they do, what gets in the way, and what they wish someone would fix. This single set of conversations produces more useful information than any handover document.
Learn how decisions actually get made, which is rarely the org chart. Find out who is consulted, who has veto, and where things get stuck.
Find out what the previous person did well, which is politically important and frequently skipped — arriving as though nothing worked before is the fastest way to alienate a team.
Understand what your manager is measured on, since your job is partly to make that go well.
Weeks three to six: establish competence.
Deliver something small and visible. An early, unglamorous, completed thing buys permission for larger ones. Early wins are about credibility rather than impact.
Ask the naive questions while you still can. There is a window in which "why do we do it this way?" is curiosity, and after it the same question sounds like criticism. Write down everything that seems odd in the first month, because you will stop noticing.
Weeks six to twelve: form a view.
Agree explicit expectations with your manager — what success looks like at six months, in concrete terms.
Now propose changes, with evidence gathered rather than assumed, and having understood why the current arrangement exists. Things that look stupid usually have a history, and knowing it is the difference between a proposal that lands and one that does not.
What the evidence on new starters suggests: the strongest predictors of a successful transition are relationship-building breadth and clarity of expectations, rather than technical performance — which is not what most new starters prioritise.
On probation: understand the criteria and ask for interim feedback rather than waiting for the review.