UPI AutoPay, e-mandates, and the real cost of recurring payments in India

Recurring billing looks like a solved problem. Charge the customer the same amount every month, send a receipt, move on with your life. In India it is not a solved problem, it's a quietly hard one, and the choices you make about how to collect that recurring payment have a direct, measurable effect on your revenue. Get it wrong and you'll watch perfectly happy customers churn for reasons that have nothing to do with your product.

So let's talk about why recurring payments here are different, what the real options actually are, and how to think about the trade-offs.

Why recurring is hard here in the first place

In a lot of markets, recurring billing rests on a card kept on file. You store the card, you charge it monthly, and the success rate is high enough that nobody thinks twice. India broke that assumption on purpose.

The rules around card-on-file and recurring mandates got tightened to protect consumers: extra authentication, mandate registration, pre-debit notifications. The intent was good. The side effect is that the naive "just charge the card" approach is far less reliable here than founders coming from other markets expect. International cards in particular fail at rates that will quietly bleed a subscription business dry.

The three ways to collect recurring money

In practice, recurring collection in India comes down to three mechanisms, each with a different reliability and reach profile.

1. UPI AutoPay

UPI AutoPay is the recurring mandate built on top of UPI. The customer approves a mandate once inside their UPI app, whether that's PhonePe, Google Pay, Paytm or BHIM, and future debits up to the agreed cap go through with very little friction. It's the RBI-approved mechanism for recurring UPI, and for most domestic consumer and SMB billing, it's the highest-converting option you have.

• Reach. UPI is already how a huge share of Indians pay, so there's no new behaviour to teach anyone.
• Friction. One mandate approval up front, then debits that mostly just work.
• Reliability. The highest success rates of the three for domestic recurring billing.

2. Card-on-file mandates (e-mandates)

Cards still matter, especially for higher-ticket B2B subscriptions and customers who simply prefer them. With a registered e-mandate, recurring card charges work fine, but you're living inside the mandate rules: registration up front, pre-debit notifications, and the authentication requirements that make Indian card recurring stricter than what you might be used to.

3. Net-banking e-mandates (NACH)

For bigger or longer commitments, net-banking mandates on the NACH rails come into play. Setup is heavier and slower, but for enterprise contracts or high-value recurring debits they're dependable and well understood by finance teams, which counts for a lot.

The cost you can't see on the pricing page

When teams compare payment options, they look at the transaction fee. That's the visible cost. The invisible cost, which is usually way bigger, is failed-payment churn.

Here's how it works. A recurring charge fails. The customer didn't decide to leave, their card just didn't go through. But now your system has to notice, retry, notify them, and recover the payment before they forget about you entirely. Every step in that chain leaks customers. The lower your payment success rate, the more people fall through, and the leak never stops.

Nobody churns on purpose because a card failed. They churn because the recovery was slow, silent, or never happened at all.

This is why payment success rate is a retention metric, not just a finance one. A two-point bump in recurring success rate can easily outweigh a meaningful difference in transaction fees, because you're keeping customers you'd otherwise have lost without ever knowing it.

Dunning: the unglamorous art of getting paid

"Dunning" is the process of recovering a failed payment: the retries, the reminders, the grace period. It's deeply unglamorous, and it's where a surprising amount of revenue gets won or lost. A good dunning flow does a few things automatically:

• Retries the charge on a schedule tuned to when money is actually likely to be there, not just blindly the next morning.
• Notifies the customer on a channel they actually read, which in India usually means WhatsApp, not email.
• Holds service gracefully during a grace period instead of cutting access the second a charge fails.
• Escalates to a human or a clear in-app prompt before giving up on the relationship for good.

Almost none of this should be something you build by hand. It's exactly the kind of multi-step, multi-system process that automation exists for: payment status from the gateway, a WhatsApp nudge, a CRM update, a retry on a schedule, all wired together once and left to run.

How to think about the choice

If you're billing Indian consumers or SMBs monthly, default to UPI AutoPay and keep cards as a secondary option for the people who prefer them. If you're billing enterprises on annual or high-value contracts, e-mandates and NACH earn their place. Whatever the mix ends up being, instrument your success rate and build a real dunning flow, because the payment you fail to collect is the cheapest customer you'll ever lose.

Adoment connects Razorpay, Cashfree and PayU natively, so the events that drive all of this, a successful debit, a failed charge, a mandate getting revoked, can kick off a workflow directly. Which means the recovery flow that saves a failed payment doesn't have to be some engineering project. You wire it up once and then you stop worrying about it.