What you'll learn
Quick Answer
Most Indian IT service companies run an annual cycle: you write a self-appraisal, your manager rates you, then a calibration meeting adjusts ratings across the whole team against a forced distribution (the bell curve) before hikes are decided. Your manager's opinion of your work and the rating you finally get are not always the same thing — calibration can move it either way.
The annual cycle, step by step
Most Indian IT service companies run the appraisal cycle on their financial year, which for a large share of them is April to March. The cycle formally starts with goal-setting — you and your manager agree on KRAs or KPIs for the year, sometimes tied to a project, sometimes to broader competencies like technical skill, delivery quality, and collaboration. A mid-year check-in is supposed to follow, where progress against those goals gets revisited; in practice, this step gets rushed or skipped at a lot of companies, which is exactly why waiting for it is a bad plan.
The cycle proper starts at year-end: you write a self-appraisal — your account of what you did against your goals — and your manager writes a rating and narrative based on your work and their own observations. That rating then goes into a calibration meeting, where ratings across a team, sometimes across an entire vertical or account, get compared and adjusted against each other before anything is finalized.
Only after calibration closes does a rating become final, and only after that does a hike or promotion decision get attached to it. Each of these stages can be separated by weeks, which is why the appraisal conversation you have with your manager and the number that eventually shows up in your salary can feel disconnected from each other — they usually are, procedurally, even when nothing has gone wrong.
The bell curve, and why it isn't really about you
Forced distribution — commonly called the bell curve — means your manager isn't free to rate everyone on the team “exceeds expectations” even if, in their honest opinion, everyone earned it. Most systems specify roughly what proportion of a team can land in each rating band, so a fixed share has to land in the middle or lower bands regardless of absolute performance. That's a structural fact about how the system is designed, not a comment on any individual manager's judgment.
The practical effect is that your rating is partly relative to your specific peer group in that cycle, not purely a measure of your own output. Being on a team stacked with strong performers can pull your relative rating down even if your work didn't change; being on a comparatively weaker or smaller team can do the opposite. This is one of the least intuitive parts of the system for people early in their career, because it means two people who did objectively similar work in two different teams can walk away with different ratings.
Calibration is where this actually happens — a cross-team meeting where ratings get adjusted to fit the required distribution before anything is finalized. Your manager advocates for your rating in that room; you're not present, and you generally don't get to see how the conversation went, only the outcome.
Why the rating and the hike letter don't always arrive together
Here's what surprises almost everyone the first time through a cycle: getting your rating communicated is not the same event as your hike showing up in your salary, and the gap between the two can be substantial. Ratings are typically finalized and communicated within a few weeks to a couple of months after the cycle closes. The actual salary revision often trails further behind — some companies stagger implementation by band, level, or even quarter, and during periods of margin pressure, the industry has a track record of delaying or partially rolling out increments across entire levels, sometimes by many months.
So if your manager tells you in April that you got a strong rating, and your July payslip doesn't yet reflect a raise, that's not automatically a sign something went wrong — it's frequently just the normal rollout timeline for that company's process. Ask your manager or HR for the expected implementation month up front, rather than assuming a rating problem when it's really a timing one.
Also worth separating clearly: a good rating and a band or designation promotion are not the same decision. A strong annual rating can come with a healthy hike and zero change in your title or band — promotions usually require a separate nomination and panel review, and don't happen automatically just because your rating was good that year.
What actually moves the needle in your self-appraisal
The self-appraisal is not a formality your manager will override anyway — in a calibration meeting, it's often the only written record your manager has to argue your case from, especially in bigger teams where a manager oversees people across multiple projects and genuinely won't remember every detail from ten months ago. A weak self-appraisal gives your manager less to work with, no matter how good your actual year was.
The difference is specificity. “Supported the team during the migration project” tells a calibration panel nothing. “Owned the data migration for one module, reducing manual reconciliation effort from three days to four hours per release, across six releases” tells them exactly what happened and why it mattered. Track what you actually did through the year — a running note, updated monthly or after every notable piece of work — rather than trying to reconstruct twelve months of contributions from memory during appraisal week, when everyone is trying to do the same thing at once and remembering less accurately than they think.
Two things carry more informal weight than most official rubrics admit: onsite opportunities and visible cross-team work, even when they aren't explicitly listed as KRA criteria. Certifications sometimes count directly, sometimes just signal initiative informally. Rather than guessing which of these matter at your company this cycle, ask your manager directly what's actually being weighed — the answer varies enough between companies, and even between managers at the same company, that guessing wastes effort.
How to prepare, practically
Ask for a mid-cycle check-in explicitly rather than waiting to see if one happens on its own. A short, scheduled conversation three or four months into the cycle — how you're tracking against your goals, and what would move the needle in the second half — gives you time to actually act on the feedback, instead of hearing it for the first time at year-end when it's too late to change anything.
Ask a more specific question than “how am I doing”: ask what would move you from your current rating to the next one up, concretely. Vague encouragement doesn't tell you anything actionable; a manager who says “take ownership of one more module independently” or “get visible in the architecture discussions, not just execution” gives you something to actually do differently.
If you disagree with a final rating, use whatever documented feedback or appeal window your company provides, and use it promptly and in writing — most HR processes have one, even if it's rarely advertised. A calm, specific written note referencing concrete accomplishments tends to get taken more seriously than an informal complaint made weeks after the fact, once the calibration record has already closed.