My first salary negotiation lasted 47 seconds. I walked into my manager’s office, said “I think I deserve more money,” and was told “we’ll revisit this at your annual review.” That was in March. The review was in January.
That’s not a negotiation. That’s a deferral. And it happens constantly, because nobody teaches you the mechanics of how to ask for a raise in a way that actually moves the conversation forward.
Why timing matters more than most people think
The mistake most people make is treating a raise request as a standalone event. It isn’t. It’s a budget conversation, and budgets have cycles. If your company runs annual planning in Q4, asking in November means the money might already be allocated. Asking in September gives your manager something to argue for.
The three moments that work best, in my experience: right after you’ve shipped something big, during the review cycle where you have a chance to shape the narrative before it gets written, and at the 12-to-18-month mark in a role when you can point to a clear gap between what you were hired to do and what you’re actually doing.
Avoid asking right after a company-wide cost-cutting announcement. That one seems obvious, but a surprising number of people do it anyway, usually because they’re frustrated and the frustration is legitimate. Legitimate frustration and good timing are different things.
Building your case with real numbers
The frame that works: this is a market correction, not a reward request. You’re not asking for a raise because you’ve been there long enough or because you work hard. You’re asking because the gap between your current comp and market rate has widened, and you have data to show it.
For data sources, Levels.fyi is the most accurate for software roles because the submissions are tied to offer letters rather than self-reported estimates. For non-engineering roles, Glassdoor and LinkedIn Salary both work reasonably well, though I’d filter to your metro area and company size rather than using national medians. The Bureau of Labor Statistics Occupational Outlook Handbook is useful if you want to cite something your manager can’t easily dismiss.
Beyond market data, you want three to five specific accomplishments from the past year. Not “led the migration project” but “led the migration project that cut p99 latency from 2,400ms to 340ms, which unblocked the enterprise contract renewal with Acme Corp.” The more specific, the harder it is for anyone to argue against.
Also document scope creep. If you were hired as a senior engineer and you’ve been running two people informally for eight months, that’s a real argument for a title change and accompanying comp adjustment.
What to actually say in the room
Short version: “I’d like to talk about my compensation. Based on what I’ve been doing over the past year and where the market is for this role, I think there’s a gap. Can we talk about closing it?”
Then stop talking. Don’t fill the silence with qualifiers. Don’t say “I mean, I’m happy here, but…” That undercuts the whole thing before your manager has said a word.
The number question is tricky. Some career coaches will tell you to never name a number first. I disagree with that for salary discussions (though it’s solid advice for job offers). If you’ve done the market research, name the range. “Based on comp data for this role in this market, I’m looking at something in the range of X to Y.” It shows you’ve done the work and it anchors the conversation.
The responses you’ll get and what they mean
“Let me look into this” means your manager needs to check with their manager or confirm what budget they have. This is actually a good response. It means they didn’t say no.
“You’re already at the top of your band” is more complicated. It might be true. It might mean the band itself needs to be revisited. Ask directly: “Is this a band issue, and if so what would it take to move to the next level?” That reframes it as a career conversation rather than a dead end.
“Not right now” should always be followed with a specific ask: “When would be a better time, and what would I need to show between now and then?” If they can’t answer that, the problem isn’t timing.
“No” with no explanation is a signal. Push for specifics. If the answer is budget, ask whether a mid-year review is possible when budget opens up. If the answer is performance, ask what exactly needs to change. If the answer is vague, that’s information too.
When the answer stays no
Some companies genuinely don’t raise salaries outside of annual cycles regardless of what you do. I’ve seen this at late-stage startups and large enterprises both. If you’ve made a strong case and hit a structural wall, the honest question is whether the role is worth the gap between your pay and what you’d make somewhere else.
Not everything needs to be a base salary fight either. Title changes (which help future offers), equity refreshes, a professional development budget, or an extra week of PTO have real value and are often easier to approve than base comp changes. They’re not a substitute for being paid fairly, but they’re worth asking about if the base conversation stalls.
The thing nobody tells you: your second raise conversation is always easier than your first, because you’ve already established that you have one.