Pay decisions run in a sequence: a pool is set, split across the business, allocated to individuals, then approved and communicated. Once the split has happened, moving your number means taking it from somebody else, which is why late asks fail. The request that works lands before allocation, and the conversation that makes it work happens a cycle earlier still.
The decision is made before the meeting
You walk into the review conversation with a prepared case and leave with three percent. It feels like the case was weighed and found ordinary. Usually it was never weighed at all, because the number already existed before you sat down. Understanding that sequence is worth more than any amount of preparation aimed at the wrong moment.
Pay decisions run through four distinct stages every year. A pool is set as a share of next year’s payroll, decided before anybody is named. It gets split across divisions and teams, still with nothing about you in it. Managers then allocate their share to individuals, which is the last point where your number can move without taking money from a colleague. Finally the totals are checked and the letters go out.
By the time the review conversation happens, you are usually somewhere in the fourth stage. The answer of next cycle is not a brush-off at that point; it is generally an honest description of where the money is. Which means the conversation that decides your increase happens weeks or months earlier. The useful question is not when to ask but when to give your manager what they need.
When to actually raise it
Six to eight weeks before your organization’s review cycle is the window that works. That lands while budgets are being planned rather than after they have been fixed, and it gives your manager time to build an argument. Any later and you are asking somebody to reopen a settled allocation. Any earlier and it fades before the decisions get made.
Ask your manager directly when the cycle runs and when the decisions are actually made. Most will tell you without hesitation, because it is an administrative fact rather than a negotiating position. Knowing that date is worth more than any phrasing you could choose for the request itself. It is also the single most commonly skipped step.
What to give them
Give them a short written record of what you delivered, with numbers wherever numbers exist. That document is not really for you and it is not a performance review. It is what your manager carries into a room you are not in, competing against several colleagues for the same pot. A manager arguing from specifics does measurably better than one arguing from impressions.
Keep it to a single page and write it so it can be repeated aloud. Anything longer will not be read by the people who actually decide, and your manager will end up summarizing it anyway. If they have to summarize, they will do it from memory and lose the numbers. Do the summarizing yourself and the figures survive.
The moments that genuinely help
Just after finishing something visible and successful is the strongest single moment available. So is any point where your scope has genuinely grown, whether that is new responsibilities, a larger team, or an area you now own outright. A documented market gap works at any time of year, because it is a fact about pricing rather than about the calendar.
Each of those gives the decision a reason attached to a specific moment. That is much easier for somebody to act on than a general request arriving at an arbitrary point in the year. Reasons travel through a process; requests do not. Attach your own request to something that actually happened.
The moments that do not
Immediately after the budget has been set is the worst timing, because there is genuinely nothing left to allocate. During a hiring freeze or a round of layoffs the answer is structurally no regardless of the case. Neither of those is about you and neither improves with better argument.
Two others are worth avoiding for different reasons. Immediately after a visible failure the conversation will be about the failure. And raising pay in the same conversation as a complaint about workload converts a compensation discussion into a grievance discussion. Those are separate meetings with separate outcomes. Keep those two conversations apart from each other deliberately.
Off-cycle requests, and when they work
Off-cycle adjustments exist as a category at most organizations and they need a reason the annual cycle cannot wait for. A genuine and documented change in scope qualifies easily. So does a documented market gap where your pay has fallen behind what the employer now advertises. A competing offer qualifies, with all the complications that carries.
Without one of those three, an off-cycle request is usually deferred to the cycle anyway. That spends your ask without gaining you any timing advantage, and it means the cycle conversation starts with a question already answered. Save the request until one of those three exists. The mechanism is real but it needs a trigger.
The follow-up that does the work
If the answer is no or not yet, ask what would change it and by when. A defined review with written criteria converts an indefinite wait into a date and a standard you can work against. Those two things are what make a deferral worth accepting rather than merely absorbing. Ask for both the date and the criteria together.
Then put the date in your own calendar with the criteria attached to it. The most common failure after a deferred raise is that nobody ever returns to the subject, and the person with the most reason to remember is you. Raise it a few weeks before the date rather than on it. That way it arrives while the next budget is being planned.
Find out when the cycle actually runs
Organizations differ enormously here and the difference decides your whole approach. Some run on the calendar year, some on a fiscal year that ends in the middle of it, and some give anniversary-based increases tied to your own start date. The last arrangement means your timing is personal rather than organizational.
Ask HR or your manager directly rather than inferring it from when letters arrived last year. It is an administrative question with a factual answer and nobody will think twice about being asked. Getting it wrong means arriving reliably after every decision has been made. That single fact is the difference between preparation and theater.
The longer game
Ask early in the year what would justify a strong increase at the next cycle, then spend the year doing exactly that and keeping a record as you go. That converts the eventual review from a negotiation into a confirmation of something already agreed. It also removes the awkwardness, because you are reporting against criteria rather than making a case from nothing.
It makes your manager’s job straightforward, which is what actually produces movement in a process where somebody else does the arguing. The question also tells you something immediately. If the honest answer is that the band has no room left, then no amount of timing helps and the real conversation is about level rather than percentage.
Common questions
When should I ask?
Before allocation, which is usually a quarter or more ahead of when letters go out. Ask when the decisions are made and work back from that.
Why is the answer always 'not this cycle'?
Because after the split, raising your number means lowering someone else's. It is often a genuine constraint rather than a way of saying no.
Does the review meeting matter at all?
For the rating, yes, and the rating drives the allocation. For the number itself it is usually too late.
Can I get a raise outside the annual cycle?
Yes, through retention adjustments, market corrections and reclassifications. They need a trigger and an extra approval, so they are slower and rarer.
How do I find out the cycle?
Ask when increases take effect and when decisions are made. Both are ordinary factual questions and asking them commits you to nothing.
When should I ask for a raise?
Six to eight weeks before your organization's review cycle, so it lands while budgets are being planned rather than after they are fixed.
What should I give my manager?
A one-page written record of what you delivered, with numbers. It is what they carry into a room you are not in.
What if the answer is not yet?
Ask what would change it and by when, get written criteria, and diarize it. The usual failure is that nobody returns to a deferred raise.