A week after accepting the “best” swap quote, the damage rarely looks dramatic. It looks like 0.7% fewer tokens than the screen suggested, a failed transaction that still cost gas, and a screenshot nobody wants to use in the approval thread. On a $25,000 conversion, that kind of loose execution can turn into $175 of avoidable value loss before anyone counts the hour spent reconstructing what happened.
The mistake is treating the displayed output as a price. It is really a conditional route: valid at this block, for this size, with these liquidity pools, token approvals, gas assumptions, and a market that has not moved yet. That distinction is the small but important question worth settling before choosing an aggregator.
Judge the execution envelope, not the headline quote
Paraswap is useful when the job is to find and execute a route across fragmented on-chain liquidity, especially when a direct pool is not obviously the cheapest path. But “it found the highest output” is not a sufficient reason to approve a trade. The defensible reason is that the acceptable worst-case result, after gas and price movement, beats the alternatives.
For a practical orientation to the product itself, keep a Paraswap practical first guide nearby; the point here is how to decide whether its quote deserves to be sent.
Start with the minimum received field. If you are swapping 10 ETH and the quoted return is 30,000 USDC, a 0.5% tolerance permits a result as low as 29,850 USDC. That is not a harmless setting: it gives the route up to 150 USDC of room to deteriorate. For a liquid, non-urgent trade, that may be more permission than necessary. For a volatile token or a large order that moves through several pools, too little tolerance may simply produce a revert.
- Compare net output: subtract the estimated network cost from each option in the asset you actually care about.
- Read the route: more hops and obscure intermediate tokens can be justified, but they create more places for liquidity to change.
- Match tolerance to urgency: use the smallest setting that is realistic for the pair and current conditions.
- Split size when impact is visible: two smaller trades may cost more gas but preserve more than they cost.
The approval conversation becomes much simpler when framed this way: not “which interface promised the most,” but “what is our maximum acceptable execution cost, and does this route stay inside it?” That is the part that holds up after the transaction is no longer a quote.