Crowdnode works onchain, so what it does is activate the Masternodes using the funds that each user makes available, it retains a share of the masternode compensations for its service and divides the rest among the users who participate (simplified explanation obviously). Furthermore, it now also offers the possibility of participating in the activation of a masternode in a trustless form:
https://knowledge.crowdnode.io/en/collections/3620166-crowdnode-trustless
Liquidity pools instead have different and more complex compensation mechanisms. You can find more information about this in the Maya protocol documentation for example:
https://docs.mayaprotocol.com/white-paper/readme-1-1/introduction