Most payment facilitator programs don’t run into trouble at registration. They run into trouble around 9 months later, when the company discovers that the registration was the cheap part.Registration is the visible milestone, so it absorbs the planning. Founders budget for it, build a timeline around it, and treat approval as the moment the business starts. What actually starts is an operating job: underwriting a portfolio whose losses you now own, funding sub-merchants before your acquirer funds you, and defending an approval rate you promised in a sales deck. The question people ask is how to become a payment facilitator. The more useful question is whether you can operate as one at the volume you are forecasting.What follows is the definition the whole decision rests on, then the assumptions that cause the most expensive corrections, each with the correction that would have saved the money.What is a payment facilitator: PayFac vs ISO vs payment processorA payment facilitator is a registered third-party agent that holds a master merchant relationship with an acquirer and boards sub-merchants underneath it. Under Visa's Third Party Agent registration rules, a payment facilitator can sign a merchant acceptance agreement on behalf of the acquirer and receive settlement of transaction proceeds on behalf of its sponsored merchants. Everything attractive about the model follows from those two rights. So does everything expensive about it.These comparisons get blurred in most conversations, and each one hides a different liability.The ISO comparison matters most at the decision stage. An ISO refers merchants to an acquirer, earns residuals, and never touches settlement or loss. A payment facilitator owns onboarding, pricing, and the downside. The choice between PayFac and ISO comes down to how much risk you are prepared to convert into margin, and how much operational capability you can fund before the margin arrives.The processor comparison confuses people for a different reason. A payment processor is a set of rails that authorizes, clears, and settles. Payment facilitation is a commercial and risk position layered on top of those rails. You will still use a processor after you become a payment facilitator, and choosing to become one changes almost nothing about how a transaction is technically processed. It changes who is standing behind it.The best-known payment facilitator companies make a point that is easy to miss. Stripe, Square, Toast, Shopify, and Mindbody all built distribution before they built acquiring. Payments were attached to software that merchants were already using every day. If your plan requires you to acquire merchants first and then sell them payments, you are running an ISO sales motion with PayFac liability bolted onto it, which is the most demanding combination in the model.Why the sponsoring acquirer is the real gateYou cannot register as a payment facilitator yourself. Visa requires an acquirer to sponsor and register you, and Mastercard follows a similar model through a principal member.That makes securing an acquirer the first major hurdle. They assess your financial strength, merchant mix, team experience, and ability to absorb losses. More importantly, they determine which merchants you can onboard, required reserves, market access, and how risk is handled.Once sponsorship is secured, the next steps are more procedural: card-scheme registration, risk and KYB requirements, PCI DSS validation, and deciding how funds move.If you hold funds, additional licensing may apply, such as US money transmitter licenses or European PI/EMI authorization. Direct settlement from the acquirer to sub-merchants can reduce this burden but also limits flexibility.Industry estimates suggest becoming a fully registered PayFac can take 12–24 months and require $500k+ in infrastructure investment.Where the PayFac margin actually goesA PayFac earns the spread between its underlying processing costs and the rate charged to merchants. But gross margin is not net profit. Chargebacks, fraud losses, reserves, risk and support teams, and working-capital needs can quickly reduce it.There is also a scaling issue: Visa generally requires sponsored merchants above $1 million in annual volume to contract directly with the acquirer, subject to exceptions. Strong merchants may therefore outgrow the standard PayFac setup.The economics work best for platforms with an existing merchant base. Payments can increase revenue per customer and switching costs without adding much acquisition expense. If you also need to pay to acquire merchants and absorb their losses, the model becomes much less attractive.What underwriting policy has to cover before you board anyoneInstant onboarding is what sub-merchants are buying, and it only survives contact with a real portfolio if the risk policy is written down and enforced by the system rather than by whoever is on shift. A workable policy covers five things.Tiered boarding by merchant category and expected volume, with automated approval below defined thresholds and manual review above them.Documented know-your-business checks, including beneficial ownership and sanctions screening at boarding and on a recurring schedule afterward.Monitoring tied to the underwriting assumptions: ticket size, velocity, refund ratio, and chargeback ratio measured against the profile the merchant declared, not against a portfolio average.Concentration limits, so no single sub-merchant or category can take the portfolio down, and reserve rules that scale with risk rather than sitting at one flat rate.A documented exit path covering fund holds, offboarding, and merchant communication, agreed with your acquirer before you need it.This is the point where tooling stops being a convenience. Sub-merchant hierarchies, boarding rules, limits, and offboarding belong in the platform as configuration, which is what sub-merchant and merchant management tooling is for. The practical benefit is that the policy is enforced at the moment of boarding, instead of being reconstructed from logs after a loss.What the payment facilitator tech stack has to do beyond accepting cardsAsk a founder what their payment facilitator technology looks like, and the answer is usually a gateway and a checkout. Those are the parts merchants see on day one. The parts that decide whether the business is operable in year two are less visible.Sub-merchant onboarding, hierarchy, and lifecycle management.A ledger that can survive an audit, with per-sub-merchant balances, fees, reserves, and adjustments.Split settlement and payouts, including rolling reserves and holds.Reconciliation across acquirers, schemes, methods, and currencies.Chargeback and dispute workflow, with evidence collection at sub-merchant level.A merchant-facing portal and reporting good enough that your sub-merchants stop asking your support team for numbers.Routing, cascading, and retry logic across more than one acquiring relationship.Building that in-house takes a cross-functional team 12 to 18 months to reach a credible minimum, based on what we see with companies who try it before coming to us. The estimate covers the build, not the maintenance, and maintenance is the part that compounds: provider APIs change, methods get added, regions bring their own settlement quirks, and none of that work makes your product better.The alternative is to start on white-label payment gateway infrastructure and keep your engineering on the part of the product merchants actually choose you for. The test to apply when evaluating any option is narrow and revealing: can you change routing rules, merchant pricing, and provider mix without an engineering release? If not, every commercial decision you make later will be gated by a sprint.When full registration is the right routeFull PayFac registration is usually not the best starting point. Most companies should compare three routes first:Sponsored or managed PayFacA registered provider handles network and regulatory responsibilities, while you manage onboarding, pricing, and merchant relationships. It is faster to launch, but you give up some margin and flexibility.HybridYou control the merchant experience, underwriting, and commercial terms while using a partner for settlement and licensing. This preserves more flexibility if you plan to register later.Full registrationFull registration makes more sense at scale, when fixed costs such as scheme fees, PCI compliance, licensing, and compliance teams are spread across a large merchant base.If you start with a sponsored or referral model, choose infrastructure that can support your future setup too. A white-label platform for ISOs and MSPs should therefore be evaluated against long-term PayFac requirements, not just today's needs.The key point: the license does not have to come first. Build the operating capability first, then move toward registration when the economics justify it.How to sequence the buildIf you are working out how to become a payment facilitator this year, the order below reduces the cost of being wrong at each stage.Decide whether payments attach to distribution you already own. If not, model merchant acquisition cost before anything else.Choose the route: referral, sponsored, hybrid, or full registration, based on volume you can evidence rather than volume you project.Select the sponsoring acquirer on category permissions, reserve terms, and market coverage, not on price.Write the underwriting policy before you write the onboarding flow, and make sure the platform can enforce it.Stand up the operating layer: ledger, settlement, reconciliation, disputes, reporting, and routing.Add the second acquiring relationship before scale, so approval rate is a lever you hold rather than a number you report.Register when volume and portfolio quality justify the fixed cost.Becoming a payment facilitator is a solved problem procedurally. The rules are published, the sequence is documented, and the vendors who help with each step are easy to find. What is not solved, and what separates the programs that scale from the ones that quietly convert back to referral models, is whether the company can run a merchant portfolio at the approval rate, risk profile, and cost base it promised. Build that capability first, and the registration becomes a formality. Build it second, and the registration becomes an expensive way to discover what was missing.If you are at the model-choice stage, the most useful conversation is about what your first hundred sub-merchants will require operationally. That is where the real budget sits. No#paymentsDenys KyrychenkoCo-founder & CEOCorefy14 Aug, 2026