Two buyers run the same offer at the same cost and one prints while the other bleeds. The difference is approve rate — the number that quietly multiplies everything downstream.
Ask a buyer how a campaign is doing and you'll hear about CPL, CR and volume. Ask why two buyers running the identical offer at the identical cost end up with opposite P&Ls, and the honest answer is almost always the same: approve rate. It's the least-watched number on the dashboard and the one that decides whether you get paid.
Approve rate is approved conversions divided by the ones that resolved — approved ÷ (approved + rejected). Pending holds are excluded until they settle, because counting them early flatters the number and then punishes you when they turn to trash. It's a settlement metric, not a delivery metric.
Effective revenue is a chain of multiplications: prelander CR × offer CR × approve rate × payout. Because approve rate sits at the end, a swing there scales everything before it. Lift approve from 30% to 40% and you haven't gained ten percent — you've gained a third of your paid volume, on the exact same ad spend.
Approve rate is mostly decided before delivery. Fraud filtered upstream never lands as a rejection; a phone that an operator can actually reach converts; a lead whose geo matches the offer clears review; and delivering while the partner's call centre is staffed beats dumping leads at 3am. The Mask, HLR validation and time-of-day routing are approve-rate levers, not just fraud tools.
A blended approve rate hides the story. The same offer can approve at 45% from one source and 12% from another; one partner may trash what another accepts. Break it down both ways and you stop optimising clicks you'll never get paid for, and start shifting volume to where it settles.
CPL tells you what a lead cost. Approve rate tells you whether it was worth buying.