How to charge customers automatically every month in India (2026): UPI Autopay, e-mandates and the rules

A business coach I know spent the first week of every month doing the same soul-destroying job. He would open his phone, scroll through a list of members, and send the same WhatsApp message forty times: "Hi sir, your monthly fee is due, please pay on this link." Some paid the same day. Most paid after two or three reminders. A handful vanished every month, not because they were unhappy, but because life got busy and paying again slipped their mind. He was losing money and hours to a problem that had nothing to do with the quality of his coaching.

If you sell anything that repeats, a membership, a subscription box, tuition, a software plan, monthly society maintenance, you know this pain. The good news is that in India, in 2026, it is completely solved. The customer can approve the payment once, and the money then comes to you automatically every month until they decide to stop. No links, no reminders, no chasing.

But "recurring payments" in India sits on top of a set of rules that confuse a lot of owners, and a few sharp numbers that, if you get them wrong, will quietly sink your subscription business. This post is the plain-English version of everything you need: how auto-debit actually works here, the three ways to do it, the RBI rules that matter (and the ones you can ignore because your tool handles them), what it really costs, and a full worked example with the maths. By the end you will know exactly how to stop chasing and start being paid on time, every time.

What "recurring payment" actually means (and why it is not a payment link)

Let us kill the confusion first, because most owners mix up two very different things.

A payment link (or a QR code, or a "pay now" button) is a one-time request. You send it, the customer decides to pay or not, and next month you send it again. Every cycle depends on the customer taking action, and every cycle, some of them don't. This is why manual monthly collection leaks so badly: your revenue is only as reliable as your reminders.

A recurring payment, also called auto-debit, flips the whole thing around. The customer gives you permission once, up front, to collect a fixed amount at a fixed frequency, say 999 rupees on the 1st of every month. After that one approval, the money moves to you automatically on each due date, with no action from either side, until the customer cancels. In India, that standing permission is called an e-mandate.

The difference sounds small. In practice it is the difference between a business that chases money and a business where money arrives on its own. When I have moved clients from links to auto-debit, the change in on-time collection and cash-flow predictability is usually larger than anything a marketing campaign would have done. Your revenue stops depending on the customer remembering and starts depending on them actively deciding to leave, which most people, most of the time, don't.

The three rails: UPI Autopay, e-NACH and cards

An e-mandate can run over three different "rails". You do not have to understand the plumbing, but you should know which one fits your business, because it changes the customer experience and your cost.

UPI Autopay. The customer sets up the mandate right inside their normal UPI app (PhonePe, Google Pay, Paytm and the rest). They see the amount and frequency, tap approve, and enter their UPI PIN once. That's it. Future debits happen automatically, and the app shows them the next debit date and reminds them before it happens. Because nearly every Indian consumer already has a UPI app, the setup friction is the lowest of the three, and because it runs on NPCI's Autopay rails, the per-debit cost is also the lowest. For most consumer subscriptions in India today, this is the default choice.

e-NACH (electronic National Automated Clearing House). Here the money is auto-debited straight from the customer's bank account. The customer authorises it once, usually through net banking, a debit card or Aadhaar. Setup is a bit heavier and can take a day or two to activate, but e-NACH shines for higher-value and business-to-business recurring payments, loan or EMI collection, expensive software billed to a company, where the amounts are large and the slightly clunkier setup is worth it.

Card standing instructions. The mandate sits on the customer's debit or credit card. After India's card-security changes, this route is fiddlier for domestic use (cards have to be securely tokenised, and there are extra checks), so its two real strengths today are customers who simply prefer cards, and international customers, where UPI and NACH do not reach.

Here is the honest comparison:

You do not have to pick only one. The same subscriptions dashboard from a gateway like Razorpay supports all three, and you can let the customer choose the one they are comfortable with, most Indian consumers will pick UPI Autopay on their own.

The simplest rule of thumb: offer UPI Autopay first to Indian customers, keep e-NACH for big-ticket or business billing, and add cards for anyone paying from abroad.

The RBI rules you must know (and the ones your tool handles for you)

Auto-debit in India is regulated, for a good reason, nobody wants a business quietly pulling money from their account without control. Since 2019 the Reserve Bank of India has required an e-mandate for recurring debits. Over the years the rules were spread across many separate circulars, and in April 2026 the RBI cleaned this up into a single Digital Payments E-mandate Framework, 2026, notified on 21 April 2026, which repealed the old scattered rules and put everything in one place.

You will not deal with the RBI directly. Your payment gateway already holds the approvals and its subscriptions product enforces these rules for you. But you must understand a handful of them, because they shape how you price and design your plans.

1. The 15,000-rupee line is the most important number in this whole post. Any single auto-debit up to 15,000 rupees goes through with no OTP each time, the customer authenticates only once, when the mandate is first created. Any debit above 15,000 rupees forces the customer to approve every single debit with an OTP (an "additional factor of authentication"). That extra OTP every month defeats the entire purpose of "automatic". Remember this line; the next section is built around it.

2. A few categories get a higher 1 lakh limit, but probably not yours. In 2024 the RBI raised the no-OTP ceiling to 1,00,000 rupees for specific categories, mainly mutual fund SIPs, insurance premiums and credit card bill payments. A normal subscription business, a gym, a coaching class, a SaaS tool, a D2C box, does not qualify for the higher limit. For you, 15,000 rupees is the line that matters.

3. The customer gets a 24-hour heads-up before every debit. The bank must send a pre-transaction notification at least 24 hours before each auto-debit, showing the merchant name, the amount and the date, with an easy way to opt out of that particular charge. Your gateway triggers this automatically. It is a good thing: it builds trust and cuts "why did you take my money?" disputes.

4. The customer can pause or cancel any time, and it must be easy. Registering, changing or cancelling a mandate needs the customer's authentication, and they can view, pause or cancel their mandates whenever they want. NPCI has even added a way to see and manage all your UPI Autopay mandates across apps in one place. Banks also cannot charge the customer for using the e-mandate facility.

The point of all four rules is the same: the customer stays in control. As a business, you meet every one of them just by using a proper gateway and switching its subscriptions product on. What you cannot outsource is understanding the 15,000-rupee line, so let us use it.

Why the 15,000-rupee line should shape your pricing

Because any auto-debit above 15,000 rupees drags an OTP into every cycle, you almost never want a single recurring charge to cross it. If it does, "automatic" quietly becomes "manual with extra steps", and your churn goes up for a reason most owners never diagnose.

For the vast majority of subscription businesses, this is a non-issue. A gym at 1,500 rupees a month, a coaching plan at 3,000, a subscription box at 799, a SaaS tool at 499 or 2,499, all sit comfortably under 15,000, so auto-debit is genuinely hands-off for the customer.

The care is needed only when the amount is large. Say you sell a 40,000-rupee course, or an annual plan you want billed yearly. You have three clean choices:

1. Split it into smaller monthly debits that each stay under 15,000, for example four monthly charges of 10,000. Each debit is OTP-free, and the customer feels the smaller number.
2. Collect it as a one-time payment up front using a normal payment link or checkout (not a mandate at all). A one-time payment has no 15,000 cap, the customer just authenticates that single payment. This is often the right call for an annual plan: charge the year in one go, at a discount, as a one-time payment, and skip the mandate entirely.
3. Use e-NACH for genuinely large B2B EMIs, where the customer expects to authenticate larger debits and the ticket size justifies it.

The mistake to avoid is setting up a 25,000-rupee monthly auto-debit and then wondering why people drop off. They are being asked for an OTP every single month, which feels exactly like paying by hand, so many simply stop. Design around the line, and the line stops being a problem.

The number nobody warns you about: failed debits

Here is the part most "just add auto-debit" advice skips, and it is the part that separates a business that collects reliably from one that thinks it will and then doesn't.

A real share of auto-debits fail, and the number-one reason is boring: no money in the account on the day you charge. A mandate is permission to collect, not a promise the funds are there. And this happens at genuine scale. In one recent month, more than 20 million UPI Autopay mandates were revoked specifically because customers did not have enough balance, and UPI Autopay is now processing close to a billion recurring debits a month, around 5 percent of all UPI volume. Insufficient balance is simply a fact of life, not an edge case.

Other, smaller reasons a debit fails: an expired or reissued card, a customer who paused the mandate, or a temporary bank-side glitch.

So you must build for failure from day one. The good news is that recovering these payments is mostly automatic if you set it up. This retry-and-remind process has a name, dunning, and here is a sensible setup:

• Charge on a smart date. Avoid the very end of the month, when balances are lowest. The 1st to the 5th, just after most salaries land, tends to succeed more often than the 28th to the 31st.
• Retry automatically. If a debit fails, don't give up, retry it a few times over the next several days. Balances change; a charge that fails on Tuesday often succeeds on Friday.
• Nudge, don't nag. Alongside the retries, send one polite message: "We couldn't collect your payment, please keep balance ready, we'll try again in 2 days." Helpful, not accusing.
• Only then treat them as lapsed. After several failed attempts and reminders, move the customer to a paused or expired state, and make winning them back a separate, warmer conversation.

Every good subscriptions tool can run this whole cycle for you, but you have to switch it on and set the rules (how many retries, how many days apart, what the messages say). Treating a single failed debit as a lost customer is one of the most expensive mistakes I see, you are throwing away revenue you could easily have recovered with two automatic retries.

What it actually costs, and the tools that do it for you

You should not build any of this yourself, and you don't need to. Every serious Indian gateway has a subscriptions product:

• Razorpay Subscriptions covers UPI Autopay, e-NACH and cards under one dashboard, and automates the mandate setup, the 24-hour notices and the retries.
• Cashfree and PayU offer comparable recurring-payment products.
• On top of a gateway, if you have many plans, coupons, upgrades, trials and GST invoices to manage, you can add a billing layer like Zoho Billing or Chargebee that handles the invoicing and plan logic. You do not need this on day one.

The cost is refreshingly simple. On the standard plans, recurring payments cost about the same as a normal one-time payment: roughly 2 percent per successful charge, plus 18 percent GST on that fee, with no setup fee and no annual maintenance charge on the standard tiers. Crucially, you pay only when a charge actually succeeds. UPI Autopay generally carries the lowest processing cost of the three rails, another reason to prefer it for smaller amounts.

One current sweetener worth checking: Razorpay has been running an offer where new merchants who complete KYC and activate on or after 1 July 2026 get zero platform fees on domestic payments up to 5 lakh rupees of volume or 90 days, whichever comes first. Offers like this change, so always confirm the live numbers on the provider's own pricing page before you launch. (If you want the full picture of one-time gateway fees too, I've written a companion guide on how to accept online payments in India and the fees nobody explains.)

A worked example: three businesses, three right answers

Let me make all of this concrete with three composite businesses, the kinds I actually work with.

Priya runs a yoga studio. She charges 1,500 rupees a month. This is a small, recurring, consumer amount, textbook UPI Autopay. A new member scans a QR or taps a link, sees "1,500 rupees monthly", approves with their UPI PIN once, and is done. On the 2nd of each month (just after salaries, deliberately not the 31st), the gateway auto-debits 1,500. The fee is about 2 percent plus GST, so roughly 35 rupees, and Priya keeps about 1,465 rupees. When a debit fails for low balance, her dunning rules retry twice over four days and send one reminder; most of those recover. Priya has stopped sending 60 WhatsApp reminders a month. That reclaimed week is now spent teaching and growing.

Arjun runs a SaaS tool for small retailers at 499 rupees a month. Same rail, same logic, UPI Autopay, well under 15,000. His only extra care is onboarding: he lets people try the tool first, then sets up the mandate at the point they subscribe, so the auto-debit begins only when they have decided to stay. His fee per charge is about 12 rupees, and because collection is automatic, his monthly revenue is finally predictable enough to plan hiring against. (If you are pricing a SaaS product, it's worth reading the pricing mistakes I see Indian founders make alongside this.)

Meera sells a premium coaching programme at 40,000 rupees, and wants to offer instalments. Here the 15,000 line bites. A single 40,000-rupee monthly mandate would force an OTP every month, painful. So she has two good options. Option one: split it into four monthly debits of 10,000 each, every debit OTP-free, and the smaller number is easier to say yes to. Option two: collect the full 40,000 as a one-time payment up front (a normal checkout, no mandate, no 15,000 cap), perhaps with a small discount for paying in full. For the customers who want to pay upfront, option two is cleaner; for those who need instalments, option one keeps it automatic. If Meera were instead billing a company a large monthly software fee, she'd reach for e-NACH instead. Same business owner, different products, different right answers, and the 15,000 line is what decides.

Notice the pattern. For ordinary monthly amounts, the decision is easy and the answer is almost always UPI Autopay. The only real thinking happens when the number gets big, and then it is about staying under 15,000 or switching to a one-time charge, not about giving up on automation.

How to set it up, step by step (non-technical)

You can have this working in a few days. Here is the order I use with clients:

1. Pick one gateway. Razorpay, Cashfree or PayU. Complete the KYC (PAN, GST if you have it, bank account, business proof). This is the same account you'd use for one-time payments.
2. Turn on the subscriptions / recurring product inside that gateway's dashboard. This is a switch, not a development project.
3. Create your plans. Name, amount, frequency (monthly, quarterly, yearly). Keep each recurring charge under 15,000 rupees. For anything larger, plan a split or a one-time upfront charge as above.
4. Set the charge date and dunning rules. Choose a smart date (early month), and switch on automatic retries with a sensible gap and a polite reminder message. Do not skip this step, it is where recovered revenue comes from.
5. Put the subscribe button where customers decide. On your website, app or a hosted checkout page. Show the amount, frequency and an easy cancel path clearly, honesty here increases signups.
6. Test with a real, small mandate on your own UPI app before you go live. Watch the mandate get created, the 24-hour notice arrive, and the first debit succeed. Then open the gates.

If you already have an app or want one, this recurring flow can be built right into it, so members subscribe, pause and manage everything inside your own branded experience rather than a third-party page. That is the kind of thing we wire up correctly from day one at the Roy Digital App Studio.

The mistakes to avoid, and the honest trade-offs

No method is free of friction. Be clear-eyed about these:

• Mandate setup adds one step at signup. Approving an auto-debit is slightly more effort than a single tap-to-pay, and a few customers hesitate. Keep the amount and terms crystal clear, and the drop-off is small and well worth the collection you gain.
• Failures are normal; ignoring them is the mistake. As covered above, plan the retries and reminders. A business that treats first-failure as final leaks money quietly.
• Don't cross 15,000 on a recurring charge without a reason. If you do, know you are trading away "automatic". Usually you should split or take it as a one-time payment instead.
• Make cancelling easy inside your own product. If a customer can't find your cancel button, they'll cancel the mandate at their bank instead, and you lose the chance to offer a pause or a discount to keep them. Easy cancellation actually raises signups, because people commit more freely to something they know they can leave.
• Keep refund terms simple and visible. Cancelling stops future debits; it does not auto-refund past ones. State your refund policy plainly, before purchase. Surprises here destroy trust fast.
• Confirm live fees and offers before launch. Percentages, GST and promotional offers change. Read the current pricing page of your chosen gateway rather than trusting any number, including the ones here, forever.

The bottom line

The coach who spent the first week of every month begging for payments didn't have a marketing problem or a product problem. He had a collection problem, and in 2026 India it is entirely solved. Let customers approve once, over UPI Autopay for normal amounts, e-NACH for big or B2B ones, cards for international, keep each recurring charge under 15,000 rupees so it stays truly automatic, switch on retries so failed debits recover themselves, and let your gateway handle the RBI rules quietly in the background.

Do that, and your revenue stops depending on your reminders and starts arriving on its own. You get back the hours you were spending chasing, and, just as valuable, you get a predictable monthly number you can actually build a business on.

If you'd like help setting recurring payments up correctly, choosing the right rail for what you sell, staying the right side of the 15,000 line, and getting the retry logic sane, or building it cleanly into your own app or website, that is exactly the kind of thing we sort out at the start of a project. And if you'd rather learn to build these flows yourself with no code, the live no-code build sessions walk through payments hands-on.