Can a second-tier global PSP like Paysafecard Business or CryptoProcessing

What you’re seeing isn’t just a drop, it’s a slide into the basement when the street still smells like success elsewhere. I’ve run the unit economics on half a dozen PSP stacks in LatAm and every point below 90 % on PIX is bleeding margin faster than a cashier counting fiat on a Saturday night. That 4 % gap you’re staring at—87 % vs 91 %—isn’t a rounding error; it’s the difference between a tidy 2.6 % blended cost at scale and a haircut that leaves your NGR looking like it was processed through a paper shredder. Let me map the trade-offs for you in three layers. First layer: the auth delta itself. If your blended approval today sits at 87 % on a second-tier PSP and you’re lured by that promised 2-3 pp fee cut (say 3 % all-in vs AstroPay’s 2.7 %), then you’re accepting a 0.3 pp hit in rate versus the fee gain. So far so simple—until you convert those raw numbers into GGR. On a 1 M USD monthly GGR stack, a 4 % approval slide equates to roughly 40 k USD in lost deposits overnight. Even if your PSP’s blended cost is literally free, that 40 k figure doesn’t care about marketing spin. Second layer: the hidden acceleration. Lower approval drags your CAC upward because you have to buy more traffic to make up volume; then your rolling reserves swell because FTDs pile up and your acquirer starts eyeing your MID with suspicion. I’ve seen cases where a 3 pp CPA increase cancels out any theoretical fee advantage inside four weeks—before you even account for KYC backlog and rising chargeback ratios. The vendors who’ll still undercut AstroPay at this failure rate aren’t sleeping soundly; they’re betting on tomorrow’s volume to erase today’s losses and praying the BNPL macro doesn’t flip negative. Third layer: the local acquiring premium versus the PIX economy. AstroPay PIX at 2.6 % through EBANX on 1 M USD volume is brutal for competitors who can’t access Banco Inter’s plumbing. But if you’re paying 7-8 % to a local acquirer just to stay in the game, the equation flips fast. Take Brazil, add 10 % auth loss on the local side, tack on another 20 % KYC churn because the acquirer flags the MID, and your blended cost suddenly screams past 5 %—even after you swap the second-tier PSP for a slightly cheaper toy. That’s why most operators who flirt with the 87 % cliff end up swallowing the full 2.7 % sooner rather than later; the math punishes hesitation. Bottom line: a second-tier PSP promising 2-3 pp savings while your approval drifts below 90 % is less a saviour and more a headwind dressed as tailwind. Run the unit economics on your own FTD curve, your own CAC stack, your own reserve waterfall, and then decide whether you’re betting against an algorithm or against gravity. Because gravity always wins.

Lost me on the FTD curve part — what even *is* an FTD curve in plain English? Like, does it shoot up when my KYC team is slow, or is it more about how many first-time depositors vanish without spending a dime?

FTD curve is just the mountain of first-time depositors who show up, throw a few chips on the table, and then vanish into the ether without ever lighting a fire under their account. Picture it like this: you run a campaign in the favelas of São Paulo, buy 100 k USD in traffic, and 70 % of those players hit “deposit” on Monday. Of those 70 k deposits, 40 % never finish KYC because your team is swamped, another 20 % bounces the email like it’s spam, and the remaining 15 % click “withdraw” the same afternoon without playing a single hand—just testing the pipe. By Friday your FTD pile is 45 k USD instead of the 30 k you budgeted. That gap—between money that walked in the door and money that actually stuck—is your FTD curve screaming at you every week. In Harry’s math, if your PSP approvals are already at 87 %, each extra 5 % of FTDs buried in KYC delays eats another slice out of your blended margin before you even blink.

Ey, that FTD curve hit different after Chris’ favela picture—suddenly my dashboard looks less like “nice charts” and more like a crime scene. 😬 So if 15 % of those São Paulo deposits ghost before KYC even wraps up, the rolling reserve math doesn’t care how cheap the second-tier PSP says it is; the game’s already over by Tuesday. Plus, Lucy’s question still nags: when does the approval slide stop being “meh, we’ll fix it later” and start looking like a one-way ticket to margin zero? If AstroPay’s 91 % at 2.6 % is the line in the sand, what’s the actual ceiling—90.5 %, 89 %, or is there a sliding scale where every extra half-point of loss wipes out the fee cut before I’ve finished my morning coffee?