Skip to main content

Compare

One integration, or your own merchant ID

One vendor says an aggregator gets you taking payments this month; the other says direct integration saves real money at scale — and both are right, for different businesses.

Ask us which one fits

Why this choice is confusing, and the question that settles it

The confusion is that both sides are selling something true. The aggregator rep is right that you could be live this month, on one API, with no underwriting file; the bank rep is right that on their schedule you keep more of every transaction. Neither one mentions what their option costs you.

One question settles this more often than any other: would the fee difference on a year of your current volume pay a payments specialist's salary? If it plainly would not, take the aggregator and revisit when it does. That is not a compromise for most businesses, it is the correct answer.

Where they diverge

Where the choice actually bites

These are not feature-list differences. Each one is a place where the decision shows up in somebody's week, months after go-live.

  • Who owns the merchant relationship

    With an aggregator you are a sub-merchant under their licence. If their risk team reclassifies your category, your account can be paused without an acquirer conversation.

  • PCI-DSS scope

    Hosted fields keep you on SAQ A whoever you use. Bank-direct integrations more often route the card number through your servers, which moves you to SAQ D.

  • Reconciliation

    Aggregators hand you a dashboard and a settlement report. Direct means somebody parses the acquirer's daily MIS file and chases the rows that do not match.

  • Failed-payment recovery

    Multi-acquirer retry needs more than one merchant ID. Aggregators sell it as smart routing; building it yourself is a second integration, not a config change.

  • UPI and local rails

    UPI needs a sponsor bank acting as PSP, so you cannot integrate it the way you can a card processor. Most direct setups keep an aggregator for it.

  • Marketplace payouts

    Splitting one payment across vendors means holding their money. Aggregator split products sit inside an escrow arrangement; doing it on your own merchant ID usually needs a nodal account.

  • Disputes and chargebacks

    A dispute arrives on the network's clock either way. The difference is whether you upload evidence to a dashboard or assemble the representment pack yourself.

  • Getting live

    Aggregator KYC is a document upload. Bank underwriting reviews your business model, refund history and category before anyone issues a merchant ID, and it can decline.

  • Who you call at 2am

    Aggregator support is a ticket queue shared with thousands of merchants. A direct acquirer relationship comes with a named contact, once your volume is worth one.

The trade-offs

Four dimensions that decide it

Read these as the honest version of what each vendor left out of their pitch.

TIME TO LAUNCH
Aggregator, by a wide margin
An aggregator integration with hosted checkout, webhooks and refunds is typically 1–2 weeks. A direct merchant ID adds bank underwriting before a line of code helps, and underwriting is not a schedule you control.
COST OF CHANGE
Direct locks you in harder
Swapping aggregators mostly means a new SDK and re-tokenising saved cards through the network. Unwinding a direct integration also means unwinding a banking relationship, so the switching cost is commercial as well as technical.
WHO CAN MAINTAIN IT
Aggregator: the team you already have
Most product teams can keep an aggregator integration healthy. A direct setup needs somebody who reads settlement files, understands 3DS challenge flows, and notices a rising decline rate before finance does.
WHAT BREAKS FIRST
Webhooks, in both cases
The customer who closes the tab immediately after paying is the first bug in every integration ever built. Direct adds a second one: settlement that reconciles to the wrong day and reads as missing money.

Which one fits you

If you are on the fence after reading this, you are an aggregator business. The direct case is rarely ambiguous when it is real.

Choose a payment aggregator if…

  • You are pre-launch, or still proving the product sells at all
  • Your fee saving would not cover a specialist's time, let alone their salary
  • You need UPI, cards and wallets live from one integration
  • Nobody on the team wants to own a daily reconciliation job

Choose direct gateway integration if…

  • Your volume makes basis points a line item finance argues about
  • An aggregator has already frozen or declined you over your category
  • You need an acquirer your aggregator does not support, in a specific market
  • Regulation requires the merchant of record to be you, not a platform
FAQ

Frequently Asked
Questions

Common questions about fees, compliance, settlement, and switching providers later.

No, they do different jobs. An aggregator gives you a merchant account underneath its own licence, so you are onboarded as a sub-merchant and settled from its pool rather than directly by an acquiring bank. A gateway is the technical layer that carries the authorisation. Most well-known providers are both at once, which is why the two words get used interchangeably. The fee gap is real, but you are buying underwriting, settlement, dispute handling and a compliance perimeter, not only an API.

When the fee difference on a year of your current volume comfortably exceeds the engineering and finance time it creates. That time is not a one-off. Somebody has to reconcile settlement files, watch decline rates by issuer, handle representments, and keep the 3DS flow working when the acquirer changes something without telling you. Run the sum against last year's actual processed volume rather than next year's forecast. The forecast is the part that talks people into this too early.

Yes, and it is the usual path. Keep payments behind a thin internal interface from day one, so checkout, refunds and webhook handling are not written against one provider's SDK shape. Card-on-file tokens do not travel between providers, so plan a migration where new customers tokenise on the new provider while the old one carries existing mandates until they lapse. Subscriptions are the painful part of any switch. Start that conversation well before you commit to a date.

Your own merchant ID, if you can get one. Aggregators carry the risk for every merchant under their licence, so their risk teams act quickly and unilaterally, and a category reclassification can pause your settlement while orders are still shipping. A direct acquiring relationship is slower to obtain and comes with underwriting you may not pass, but once granted it is a contract rather than a policy setting somebody can change. Ask any aggregator outright whether your MCC is restricted.

Aggregator integrations, most of the time, and we will say so before you pay for a discovery call. Our Custom Payment Gateway Integration work covers hosted checkout, subscriptions, refunds, webhook idempotency and reconciliation dashboards on providers such as Stripe, Razorpay, PayU and Cashfree. We build direct and multi-acquirer setups too, usually for teams already processing at volume or carrying a category no aggregator will keep. A single-gateway integration is typically 1–2 weeks; subscriptions with multi-currency and refund flows, 4–8 weeks.

Have a project in mind?

Fixed price after a paid discovery — no hourly billing. A real engineer reads every enquiry, and we reply within 24 hours.