The consultancies are selling you a benchmark. You need a decision.

The compensation report on your desk was built from data gathered months before it published, then cleaned, aggregated, and smoothed into a tidy percentile band. For a slow-moving function that’s fine. For a BFSI centre hiring a scarce risk-and-quant leader in a market that reprices every quarter, it’s a photograph of a market that has already moved. You’re negotiating against today’s scarcity with yesterday’s average, and the distance between those two numbers is the whole game.

And the lag isn’t a rounding error. It compounds fastest in exactly the roles you’re leaning on the number to price. A generalist band barely drifts in 6 months. A niche band, say a model-risk lead who can actually sign off on an internal ratings approach, can jump in a single quarter because 3 centres decided at once that they needed the same person. The survey averages that surge back down toward the middle, so the one figure you trusted is most wrong in the one place it had to be right.

The band also flattens the signal you needed. It hands you the middle of a distribution when your problem is the tail: the three people who can genuinely do the job, and what it will take to move one of them. A percentile can’t see the tail. It was engineered to smooth it away.

That matters because a senior hire is a population of one. You already know the person’s name, their current employer, and roughly what it would take to get them to answer the call. The 50th percentile has nothing to say about that human. It was assembled by stripping out the very specifics that decide whether your offer lands: the counter their boss will make, the equity they haven’t vested, the reason they’d walk. You can’t price an individual with an average built to erase individuals.

A benchmark like ‘the 75th percentile for this role is X’ describes the market. The decision is a separate question: given who we’re chasing, who else is chasing them, what their current package actually vests to, and how badly we need this seat filled this quarter, what do we put on the table? That answer lives in live placement data, in the published band versus the closed offer, in knowing the specific person’s specific alternatives.

That’s the line the big firms won’t cross, and it’s deliberate. Handing you a decision means owning the outcome. A benchmark is unfalsifiable; advice is accountable. So you’re sold the safe half and left to improvise the half that mattered, usually at the moment you can least afford to guess.

There’s a commercial logic under this worth saying out loud. A benchmark scales beautifully. You run the survey once, sell the same percentile table to hundreds of clients, and not one of them can call you wrong, because all you ever did was describe the room. A decision refuses to scale. It gets made one seat at a time, by someone close enough to the live search to know what’s actually closing this week, and it can be proven wrong in public within the month. One of those is a printing press at zero liability. The other is hard, specific, exposed work.

Watch what that incentive does to the advice as the stakes climb. For a common role the guidance is crisp, because the data is thick and the downside of a miss is small. For the rare, expensive hire, the exact seat where you’d happily pay for a real answer, the language goes soft: the range widens, the caveats multiply, and you get told to triangulate against your peer set. The confidence drains out at the precise moment you needed it. Read that softness for what it is, a firm managing its own exposure while the call stays parked on your desk.

So ask for the decision itself: the number to put on the table for this person, the shape of it across base, variable and equity, the point past which you should walk away, and the honest read on what their current employer will counter with. Every one of those is knowable from live search data. A percentile table carries none of them.

The test for whether comp intelligence is worth paying for is simple: does it change what you do on Monday? A report that tells you the band and stops has told you something true and left you exactly where you started. Intelligence that’s worth the money names the number for this hire, in this market, this quarter, and stands behind it.

We’d rather be the firm that says ‘pay this, here’s why, here’s what we’re seeing close’ and be held to it, than the one that hands you a distribution and calls the guessing your job. The band is the input. The decision is the product. Stop paying for the input as though it were the answer.