The constant-product design spreads depth across every possible price, including prices the asset will never reach. Most of the capital does nothing.
Concentrated designs let a provider commit capital only within a chosen band. Inside that band the pool is dramatically deeper for the same money; outside it, that provider contributes nothing at all.
What it changes for a trader
A single depth number stops being comparable. Two pools with identical totals can behave completely differently depending on where the liquidity sits. The only meaningful question becomes depth at the price you will trade, which is the framing already required in price impact.
Slippage becomes non-linear in a new way. Inside the band, execution can be excellent. Push price past the edge and depth falls off a cliff rather than tapering, so an order that was cheap can become expensive partway through.
The provider side, and why depth moves
Concentration amplifies fee income and amplifies the exposure in impermanent loss. A provider whose band is left behind by price holds entirely the losing side and earns nothing, so they must actively re-position.
Which produces a behaviour worth knowing: after a large move, liquidity does not simply thin — it is stranded outside the new price, and it stays stranded until providers act. The pool can be simultaneously well-capitalised and shallow where it matters.
Reading it honestly
Ask where the depth is, not how much there is. And treat a total-liquidity ranking across pools of different designs as comparing quantities that are not the same measurement — the same discipline as depth versus volume.
It is also one of the differences that makes cross-chain comparison unreliable, since design mix varies by ecosystem — Solana versus EVM.