Shipping production acquiring in one quarter
Most acquiring failures show up two weeks after launch. Here is what we build to survive that second week — and why that is the bar we hold ourselves to.
Most acquiring failures do not show up at launch. They show up two weeks after launch — chargebacks the team cannot answer, settlements that do not reconcile, gateway downtime the monitoring missed. The first week looks great. The second week tells you whether you built it right.
What we always build
- Idempotent webhook handling on every state transition. Replays cannot double-charge or skip events.
- Settlement pipeline that writes a ledger the finance team can audit, not a black-box.
- Chargeback workflow connected to support — disputes get answered, not dropped.
- PCI-DSS scope assessment from day one, not after a breach.
What we never skip
Sandbox parity, dual-region writes, runbook with on-call rota by week three. The temptation to ship without these is real because they do not earn praise on launch day. They earn praise in the second week, when the alert fires at 2 AM and the on-call engineer can fix it without paging the founder.
How we charge for it
Milestone billing, fixed price per milestone, full handover at the end with a 30-day support window. We do not staff-augment, we do not bill by the hour. If you need bodies in seats, we will refer you elsewhere.