Build or Integrate: The Economics for a Fintech

A financial technology firm adding crypto capability faces a build or integrate decision across several layers, and the economics point clearly in most of them. The sectors below all reach the same infrastructure eventually, and a regulated crypto payment provider with fiat settlement is a reasonable reference for what that infrastructure has to do.

Liquidity: integrate

Sourcing conversion at competitive spreads requires inventory carried at risk or relationships with those who have it.

Building it before having volume spends scarce engineering on a solved problem at worse pricing than a partner offers.

Revisit when volume makes the spread material, which is later than most firms expect.

Chain infrastructure: integrate

Nodes across many networks, reorganisation handling, deposit monitoring. Operationally heavy, entirely undifferentiated, and a source of outages that are yours to explain and not yours to fix.

Compliance screening: integrate the data, own the decisions

Sanctions and risk data are bought.

Who to onboard, what to escalate and what to report are decisions that sit with whoever holds the regulatory obligation, and that is a question of licensing rather than of technology. The institutional version of this runs through a provider serving funds and family offices, where the audit requirements are different from the start.

The ledger: own

The record of what each user is owed.

Owning it keeps the product flexible and the partner replaceable. Firms that let a partner hold the ledger discover that switching means migrating balances rather than changing an integration.

The user experience: own

Where differentiation actually lives, and the only layer where building produces a competitive advantage.

The licensing question underneath

Whether the firm needs its own authorisation or can operate under a partner’s.

Holding crypto for users is custody. Exchanging on their behalf is another permission. Introducing users to a provider who does both may leave the perimeter with the provider.

It turns on who has the contractual relationship with the user and who controls the assets, not on the architecture.

A legal view before building costs far less than restructuring afterwards.

The term firms under-negotiate

Notice period.

Ninety days on a partner embedded in the payment flow is not enough to migrate. Firms negotiate the rate and accept the rest as standard, and the rest is where the operational risk sits. Whatever you conclude here, the balance you actually trade belongs at a crypto exchange with published fees rather than wherever the interface was friendliest.