Competitive with what?
That's the question the phrase is engineered to stop you from asking. It sounds like a promise. It contains no information. And the ambiguity isn't laziness — it's the whole mechanism.
The Frame Is The Product
"Competitive salary" means: we pay roughly what other companies pay people with your title.
Read that again, because there's a substitution buried in it. The thing being measured is no longer the value you produce. It's what similar people currently tolerate.
Those are unrelated quantities. One is about your output. The other is about the negotiating history of strangers who share your job title.
And the second one is bounded in a way the first isn't. Market rate is an average of past outcomes. It's a lagging indicator built from everyone who took the offer — including everyone who was underpaid, desperate, bad at negotiating, or in a cheaper city. You're being benchmarked against a distribution that includes every person who settled.
Accept the frame and you've agreed that the ceiling on your pay is set by other people's compromises.
It Caps The Top Much Harder Than It Lifts The Bottom
Here's the asymmetry that makes it a trap rather than just a shrug.
If you're below market, "competitive" is a real argument in your favor and you should absolutely use it. Bring the data, get the raise. It works.
But if you're substantially above average in output — the person who ships what three people ship, who fixes the thing nobody else can — market rate becomes the thing arguing against you. "We pay competitively" now means "we pay you what we'd pay the median person in your seat," and your leverage evaporates into a benchmark.
The framing is a ratchet that only turns one direction. It pulls the underpaid up toward the median and pins the high performers down at it. Which is exactly what a compensation band is designed to do — bands exist to control variance, and you are the variance.
That's not a moral failing of HR. It's the function.
You Cannot Capture Value Inside A Band
This is where it connects to the thing I actually care about.
I don't create value I can't capture. A salary band is a hard, explicit, written-down cap on capture. It says: no matter what you produce, the number lives between X and Y.
So the productive question is never "am I being paid competitively." It's:
What's the spread between what I generate and what I'm paid, and is it getting bigger or smaller?
Because if you're getting better — and you should be getting better — and your compensation is anchored to a median that moves at inflation, the spread widens every single year. Your improvement accrues to the employer by default. You get a 3% adjustment and they get the compounding.
That's not theft. It's just the deal, running as written. It only becomes a problem when you don't know it's the deal.
What To Use Instead
Three replacements, in ascending order of leverage.
1. Price on value, not title. "Competitive for a senior engineer" is a title question. "What's the revenue or cost impact of the system I own" is a value question. You won't capture all of it — nobody does — but arguing about the right quantity beats winning an argument about the wrong one.
2. Get the number from outcomes, not surveys. The only real market data is offers you've actually been given. Salary aggregators are self-reported, stale, and heavily weighted toward whoever bothered to fill in the form. An offer in hand is worth more than every percentile chart ever published, and it's the only thing that reliably moves an existing employer.
3. Build a second income line that has no band at all. This is the actual answer and the reason the other two are a sideshow. Employment income is capped by definition — there is a band, someone wrote it, it exists in a spreadsheet you'll never see. Income from something you own has no band. It has a floor of zero and no ceiling, and which of those you get is determined by you rather than by a compensation committee.
Either way... one of these is a negotiation and one is an exit. I'm running both, and only one of them is on a deadline.
The Golden Handcuffs Version
Watch what "competitive" does over a decade.
Each year they move you enough that leaving costs something. Not enough to close the spread — enough to make the walk-away math annoying. Match, vesting, a title bump, an adjustment slightly above inflation. Every one of those is real money and every one of those raises the price of the door.
The handcuffs work because they're genuinely golden. If they chafed you'd take them off. "Competitive" is the lubricant: it keeps the arrangement feeling fair enough that you never do the full calculation.
And the full calculation is not complicated. It's: what do I produce, what do I keep, and is the gap closing.
Run it once a year. Most people never run it once.
Ask A Better Question
Next time you see "competitive salary," translate it: we pay approximately what other people accepted.
Then ask the question it was built to deflect — what's the spread, and which direction is it moving?
If the spread is widening, no raise fixes it. A raise is a one-time adjustment inside a structure that produces the gap by design. You don't out-negotiate a structure.
You replace it, on a timeline, with something that has no band.