Outsources Regulated Work

Three Questions to Ask Before Your Fintech Outsources Regulated Work

Posted by

There is a point in most fintech growth stories where the operations team stops being able to absorb the volume, and somebody raises the idea of using an external partner. It is rarely a controversial suggestion. What is controversial — or ought to be — is how quickly the conversation moves from whether to do it to who to sign with.

The decision deserves more scrutiny than it usually gets, not because outsourcing regulated work is unwise, but because the failures are predictable and nearly all of them trace back to questions nobody asked at the start.

Three are worth putting on the table before a shortlist exists.

1. What exactly are we moving, and what can never move with it?

The useful distinction here is not difficulty. It is accountability.

Preparation travels; determination does not

A great deal of regulated work is preparation: gathering documents, verifying identity, assembling evidence for a dispute, drafting a response for approval, chasing a third party for information. This work is detailed, it is time-consuming, and it transfers to an external team without much difficulty provided the process is written down.

What does not transfer is ownership of the decision. The determination on whether a transaction is suspicious and should be reported, the final outcome of a complaint, the decision to grant forbearance or to close a customer relationship — these commit the licensed entity, and the regulator expects a named person inside that entity to own them.

Precision matters here, because the shorthand version circulating in most boardrooms is simply inaccurate. The rule is not that regulated work must stay onshore. Supervisors in several jurisdictions explicitly contemplate a third party handling regulated processes, including complaint handling, provided the firm remains answerable for the outcome. Published guidance on offshore fintech support models in the Philippines tends to draw the same line: the work can be done anywhere, the accountability cannot be delegated with it.

Write two lists before you write a brief

So write two lists before you write a brief. What the external team will handle, itemised by the kind of case and the task involved, not by naming a department. And what remains with a named individual at your firm, with the authority limits spelled out.

“Every firm can tell you what their provider will be doing. Far fewer can tell you what they have deliberately kept,” said John Maczynski, chief executive of PITON-Global, which shortlists and vets Philippine outsourcing providers on behalf of financial firms. “Write the second list first. It is much shorter, and it is the one somebody will eventually audit.”

Programmes that skip this step do not discover the gap at launch. They discover it in month six, when a case is settled on terms nobody at the firm would have agreed to, and the trail of who signed it off runs cold.

Read: Personal Loan Market Growth and Lending Trends

2. Has this partner actually done our kind of work — this year?

“Fintech experience” is close to meaningless as a selection criterion. Nearly every provider of any size has some.

Why the sector label misleads

The question that separates candidates is narrower: are they running programmes that look like yours, right now, under the same regulatory regime, at roughly your scale? Three years of card disputes teaches a team which evidence actually carries a representment and how an issuer tends to respond when pushed. Move that team to a lending book and they meet affordability assessments, forbearance conversations and customers in genuine financial difficulty. The sector label is identical. The job is not.

This matters more in 2026 than it did five years ago, for an unflattering reason. The spread of AI tooling has made it easy for any provider to claim capability it has not built. A demonstration proves a system exists; it does not prove anybody on the floor knows how to use it on your contact mix, or what happens when it returns an answer that is confidently wrong.

Three questions that settle it quickly

None of these costs a partner anything in client confidentiality:

  • How many programmes of this type are live this month, and at what size?
  • Who would actually manage ours, and what else are they managing?
  • What does your quality review sample look like by contact type, and why those proportions?

The third is the most revealing. A flat review percentage across every contact type means nobody has weighted review by consequence, which tells you something about how the operation thinks.

“The useful question is who would actually be assigned to this account, and what else is already on their plate,” Maczynski said. “Capability sits in named people with calendars, not in a company history.”

3. What happens when something goes wrong at three in the morning?

Most due diligence concentrates on steady-state performance. Most of the damage in an outsourced programme happens in the exceptions.

Rehearse the escalation path

A payment rail fails. A regulator asks a question with a short deadline. A fraud pattern appears that nobody has seen before. A journalist calls. At that moment the useful question is not whether your partner is competent but whether the path from their floor to your decision-maker is short, named and rehearsed.

Ask for the escalation design in writing. Who is called, in what order, with what information attached, and how quickly does it reach somebody at your firm who can authorise a response? Then ask when that path was last tested with a live incident, and what broke. Every genuine test surfaces something; a partner who reports that nothing went wrong either has not tested it or is not telling you.

And how would we leave?

The second half of this question is the one nobody enjoys raising during a courtship. If the arrangement ends — because you outgrow it, because service slips, because the partner is acquired — what actually comes back to you, and in what condition?

Ask who owns the process documentation written during the engagement, how customer records are returned and verified as deleted, how long the partner is obliged to keep operating while you transition, and whether they will support a handover to a competitor. A firm that has thought carefully about its own exit terms has usually thought carefully about the rest of the relationship too.

The pattern behind all three

Each of these questions is really the same question asked from a different side. Outsourcing moves the execution of work. It does not move responsibility for the outcome, the knowledge required to do it well, or the obligation to tell somebody when circumstances change.

Where the partner is wired into the business — looped into releases, present when something breaks, receiving policy changes directly — the arrangement generally delivers what was promised. Firms that treat it as a function handed over at the end of a procurement process tend to spend their second year explaining to somebody why a decision was made that nobody can account for.

The difference is visible long before a contract is signed. It is visible in which questions get asked first.