The published band and the closed offer are two different numbers

The published band is assembled by people who aren't, at that moment, trying to close anyone. It smooths thousands of data points into a range that's true on average and precise about no one. The closed offer is the opposite: it's the single number that moved one specific candidate, with one specific alternative in hand, on one specific Tuesday. The two answer different questions, and only one of them signs.

This matters most in the senior AI, data and IT roles where the interesting hires live above the median by definition. When a committee anchors its budget to a published midpoint, it's budgeting for a candidate who, almost by construction, isn't the one they want. The person worth chasing is already being quoted numbers the survey hasn't seen.

The lag is built into how a salary survey is made. A comp house collects submissions from participating employers on a fixed annual cycle, cleans and matches them to job codes, ages the data to a common reference date, aggregates it into percentiles, and only then publishes. Each of those steps is deliberate, and each one adds distance between the number on the page and the number being signed. Aon's 2025–26 India survey draws on more than 1,400 organisations across 45 industries; Mercer's Total Remuneration Survey covers over 8,000 roles at more than 1,500 companies. That scale is exactly what makes the output authoritative, and exactly what makes it slow. You can't aggregate 1,500 companies quickly.

There's a second effect underneath the cycle: self-selection. Surveys are populated by employers stable enough to participate, submitting roles that map cleanly to an established job code. The mandate that's genuinely re-pricing a role, a new title, a hybrid skill set, a competitive counter-offer, is the one least likely to be sitting inside the sample. So the survey reports the market late and, worse, systematically under-represents the very transactions setting the new price.

By the time a range is compiled, checked and published, the mandates that pushed it have already closed at numbers the compilation never captured. In a steady segment that lag is harmless. In a segment that's re-pricing, which is most of the ones a GCC actually competes in, the lag runs in exactly the direction that hurts you. Roughly a quarter or two of drift, all of it on the roles you can least afford to under-price.

Drift concentrates in the roles that are re-pricing fastest, and those roles cluster in one place. Demand for AI talent in India is projected to exceed 1.25 million by 2027, against a base of roughly 600,000–650,000 in 2022, according to the Deloitte–Nasscom report on bridging the AI talent gap, which also found only about 16% of IT professionals were AI-skilled. When demand roughly doubles and the qualified pool stays thin, the clearing price moves faster than any annual survey can track.

You can see the movement in the postings before you see it in the bands. Naukri's JobSpeak index recorded AI/ML hiring up around 45% over the year, with the senior 20-plus-LPA band up about 16%, a job-posting signal that leads the compiled survey by definition, because a posting is real-time and a survey is not. The aggregate national picture reinforces the pressure: India's Net Employment Outlook was +68% for Q2 2026, the strongest since 2008, yet 82% of employers reported difficulty finding the skills they need, per the ManpowerGroup Employment Outlook Survey. Strong intent meeting thin supply is precisely the condition under which a published midpoint goes stale fastest.

This is why anchoring a senior AI or data budget to last year's published midpoint is a slow miss. The band is describing a market that, in this segment, no longer exists at that number.

The fix keeps the benchmarks in place. Benchmarks are the most defensible starting point you have, and the convergence between the major houses gives you a solid floor: Aon projects India salary increases of 9.1% for 2026 and Mercer projects around 9%, two independent gold-standard surveys landing in the same place. Treat that consensus as the base of the range. It tells you where the broad market sits; it doesn't tell you what the person in front of you will sign.

Correcting the band is a discipline. Start from the published midpoint, then age it forward by the drift you can actually observe: the offers your own recent mandates in that exact segment closed at, the counter-offers candidates are declining, and the leading indicators, live postings and the movement in the senior salary band, that run ahead of the survey. Where the segment is re-pricing, weight it further: GCC-specific increments have run hotter than the national number, with Zinnov putting average GCC increments near 9.9%. The output is a working range for this role, this quarter, in this segment: the published band as a floor, corrected upward by evidence.

Most comp tracking stops at “offer made.” The number worth watching sits one step later: how often your offers get signed. A benchmarked band can look defensible on paper and still lose people at the table. When a well-benchmarked segment starts posting a low accept rate, that's the band and the closed offer telling you they've drifted apart: the published figure says one thing, and the candidates you actually want are answering with another. Read the accept rate in your re-pricing segments and it will flag the drift before your next budget cycle does.

A published band is a base-salary number. The offer that closes a senior candidate is rarely a base-salary decision. At the leadership level, variable pay and equity carry a growing share of the package. Mercer reports the wider India market shifting toward skills-based, more transparent pay frameworks, and Deloitte's executive rewards work puts CXO pay up 7–11%, with a large slice of senior earnings performance-linked rather than fixed. Budget only against base and you'll lose candidates on the parts of the offer the band never showed you.

So the correction runs on two axes. The first is level: how far the live market has moved past the published midpoint. The second is mix: how base, variable and equity combine to make an offer competitive at this seniority. A committee that adjusts the base but ignores the mix has corrected half the problem. The closed offer speaks to both, which is the final reason it should anchor the budget.

For a compensation committee, this resolves into one procedural change. Treat the published band as the floor the budget is built up from. The survey enters the room as evidence. The verdict is the corrected range: the published midpoint, aged forward by live signal, with the mix specified for the level.

In practice that means three standing questions before a senior band is signed off. What did our last comparable mandates actually close at, and how far above the published midpoint? Is this segment re-pricing, and if it's AI, data or a scarce IT skill, it almost certainly is? And is the number a base figure or a total-package figure, mix included? A committee that asks those three questions is budgeting against what candidates are signing. A committee that reads the band as the answer is budgeting against a market that has already moved, and, as the offers keep landing above the range, will lose the hire it most wanted to make, slowly, one great candidate at a time.