The First 90 Days in a New Job: The Quiet Moves That Set Up Your First Raise

Everyone tells you the first 90 days are about “making a great impression.” Sure. But impressions fade by February. What doesn’t fade is documentation — and the strange, useful truth about your first 90 days is that this is the only period of your entire tenure when you can capture a clean “before” picture of the job. Every raise case you’ll ever make at this company is a before-and-after story. Most people only ever collect the “after.”

I learned this by accident. In my second big marketing job, out of pure new-girl nervousness, I wrote a memo in week two describing the state of everything I’d inherited: email open rate 14%, no attribution on the events program, a six-week campaign turnaround. Eleven months later that nervous little memo turned a vague review-season claim (“things are going well!”) into a case — open rate 14→27%, turnaround six weeks→nine days — and the biggest single raise of my employed life. Nobody disputed the “before” numbers. I was the only person who’d bothered to write them down.

So here’s the 90-day plan I now run deliberately: three phases, each with a public layer (do the job well) and a quiet layer (set up the raise).

Days 1–30: baseline and recon

Write the baseline memo. One page, plain descriptions, zero blame: what you inherited, with numbers wherever numbers exist. Metrics, processes, response times, backlog counts, error rates. If a metric isn’t tracked, that fact is itself baseline (“no attribution existed for the events program”). Date it, email it to yourself, and — this is the finesse — consider sharing a diplomatic version with your manager as “here’s my understanding of the current state; am I reading it right?” You get alignment, they get a thoughtful new hire, and your before-picture becomes co-signed.

Learn your manager’s scoreboard. Ask directly: “What are you on the hook for this year? What would make this team’s year a success to your boss?” Every task you’re handed for the next year will secretly be graded against that scoreboard, and most new hires never ask to see it. Bonus: this question, in week two, marks you as senior-minded faster than any deliverable can.

Days 31–60: the visible win and the map

Ship one visible, finishable thing. Not the biggest thing — the finishable thing. A cleaned-up report everyone dreads, a process doc that kills a recurring meeting, a small automation. The first concrete win does two jobs: it converts your hiring from decision into vindication, and it becomes receipt #1 in the file.

Start the receipts file now, not at review time. One line per win: date, what, metric, link to evidence, who saw it. Fifteen minutes on Friday afternoons. This file is the raw material for everything downstream — the Promotion Doc is basically a curated export of it, and it’s the data layer for the annual review I run on my own career. Memory is a terrible archivist; February-you will remember maybe a third of what September-you did.

Map the influence graph. Notice whose opinions move your manager, who gets consulted before decisions, who has quiet veto power. These are your future sponsors — the people whose two-sentence endorsement in a calibration room outweighs a page of self-assessment. You don’t “network” them; you find one real way to be useful to each, then let time pass. Sponsorship is compound interest on small deposits.

Days 61–90: scope, in writing

Get your actual scope documented. By day 60 your real job has diverged from the job description — it always does. Write the delta down: “In addition to the posted scope, I’ve picked up X and Y.” If the divergence is upward (it usually is — new people are handed the orphaned work), this becomes the seed of a future case that you’re operating above level. Scope drift that’s documented is leverage; scope drift that’s invisible is just unpaid work.

Have the expectations conversation. Before day 90, ask your manager the question that pre-writes your first review: “Looking ahead to the review cycle — what would ‘exceeds expectations’ concretely look like for this role by then?” Write down the answer and email a friendly summary back: “Great conversation — capturing what we discussed so I can aim at it.” You have now converted a subjective future judgment into a semi-written contract, months before anyone else on your cohort thinks about reviews. When the cycle arrives, your case opens with their own definition, met.

The script for the two conversations that matter

Because the whole plan hinges on two questions asked out loud, here’s exactly how I phrase them.

The comp-cycle question, dropped casually at the end of a week-three 1:1: “Totally future-focused question, no agenda for now — how does comp planning actually work here? When do decisions get made, and are first-year hires in the regular cycle or prorated?” The disclaimer up front matters; it tells your manager you’re orienting, not agitating. Almost every manager answers this straightforwardly, and many will volunteer the internal politics of it too, which is worth more than the calendar.

The expectations question, in month three: “I want to aim at the right target. If we sat here at review time and you said this hire was a clear win — what would have happened between now and then, specifically?” Then be quiet and take notes. The word “specifically” is the working part; without it you’ll get “keep doing what you’re doing,” which is kind and useless. With it, you’ll usually get two or three concrete outcomes — and those become your operating plan, straight from the person who grades you.

The honest failure modes

This playbook has edges, and I’ve hit most of them:

• Baselines can read as blame. If your predecessor still works there — or worse, got promoted — a memo cataloging inherited problems can make enemies by week three. Keep the language clinical (“current state,” never “mess”), attribute nothing, and share selectively. The version for your own file can be complete; the shared version should be kind.
• The finishable win can typecast you. Fix the reporting mess brilliantly and you may be crowned Reporting Person forever. Pick a first win adjacent to work you want more of, not just whatever’s most broken.
• Some companies genuinely won’t move in cycle one. Prorated eligibility, comp freezes, rigid bands — sometimes the machine says “not yet” regardless of your case. The 90-day groundwork still pays: the same file that couldn’t move a frozen budget becomes, word for word, the evidence pack for a negotiation somewhere it can.
• Don’t optimize so hard you forget the job. The quiet layer is maybe four hours a month. If the receipts file is thriving while the actual work slips, you’ve built a beautiful case for a raise you won’t get.

None of these moves is dramatic, and that’s the point. Ninety days of unglamorous documentation, one visible win, and two well-timed questions — that’s the whole machine. Run it, and your first review stops being an evaluation and starts being a reveal.