UPI Autopay and Recurring Payments in India 2026: Setup Guide
The Quiet Killer of Indian Subscription Businesses
A gym in Lucknow signs up 400 members. Every month, someone from the front desk calls or messages the ones who have not paid. Some pay, some say they will, some go quiet. Over a year, membership drifts downward — not because people decided to leave, but because nobody actively decided to stay.
This is the default state of recurring revenue in India, and it applies as much to SaaS products and coaching institutes as to gyms. Manual renewal collection turns every billing cycle into a re-sale, and you lose a percentage every time.
Automated mandates change the default. Instead of the customer having to actively pay each month, they have to actively stop. That single inversion is worth more to a subscription business than almost any feature.
I'm Ashish Sharma, founder of Codingclave. We build custom systems with payment flows for Indian businesses. This guide covers how recurring payments actually work here, and — more usefully — what has to be built around them.
The Three Mandate Types
Each suits a different situation. Choosing wrongly costs you setup completions.
UPI Autopay
The customer authorises a recurring debit from their bank account through their UPI app, specifying a maximum amount and frequency. Approval happens in the app they already use every day.
Best for: consumer subscriptions, memberships, small and mid-value recurring amounts.
Why it wins in India: setup completion. A customer who has to enter card details drops off far more often than one who taps approve in their UPI app. For consumer-facing subscriptions this is usually the difference that matters.
Limits: per-transaction ceilings apply, and higher-value debits may require additional authentication. Check current thresholds when designing anything above a modest monthly amount.
Card mandates
Registered against a debit or credit card, following card network and regulatory rules including registration, pre-debit notification and defined limits.
Best for: higher-value subscriptions, B2B, and customers who prefer cards.
Trade-off: more setup friction than UPI for Indian consumers, but a card is more likely to have a balance available than a bank account on the 1st of the month.
e-NACH and e-mandate
Bank account debits registered through the NACH framework, typically authorised with net banking or debit card credentials.
Best for: loan EMIs, insurance premiums, higher-value or long-duration commitments, and B2B collections.
Trade-off: registration is slower and completion rates are lower than UPI, but mandates are durable and suit long-term high-value collection.
Which to offer
Offer UPI Autopay first for consumer subscriptions. Offer cards as an alternative. Use e-NACH for high-value or long-duration commitments.
A single recurring payment option is a self-inflicted wound — the customer whose preferred method is missing simply does not subscribe.
The Rules You Must Design Around
Two constraints shape every implementation in India.
Pre-debit notification. The customer must be notified before the amount is collected. Your gateway usually sends this, but you should know what it says and when it goes out, because a confusing notification triggers mandate cancellations. Some businesses send their own friendlier message alongside.
Mandate limits. The mandate specifies a maximum amount. If your charge exceeds it, the debit fails. This bites when you raise prices or when a customer upgrades — a price increase within an existing mandate ceiling is fine; one above it requires a new mandate. Design your mandate ceiling with headroom at registration, and be explicit with the customer that you are doing so.
Regulatory detail in this area changes, so confirm current thresholds and requirements with your payment provider before building rather than relying on any article, including this one.
What You Actually Have to Build
Integrating the gateway is the small part. This is the rest of it.
Mandate lifecycle management
You need state for every customer: mandate registered, active, paused, expired, revoked, failed. Each state changes what your product should do. A customer whose mandate was revoked last week should not be treated identically to one whose payment failed this morning.
The retry strategy
This is where the money is, and where most implementations are lazy.
Most failures are insufficient balance, and insufficient balance is a timing problem, not a willingness problem. Retrying at 2am the next day hits the same empty account.
What works:
- Retry around salary dates. In India, the last and first few days of the month are materially different from the middle.
- Space retries out. Same-day retries mostly fail again.
- Cap the attempts. Repeated failed debits irritate customers and can affect your standing with the provider.
- Communicate between attempts. A silent retry that fails teaches you nothing and gives the customer no chance to act.
Dunning: the communication sequence
The single highest-return thing to build, and the most commonly skipped.
A failed payment is a customer who probably still wants your service. Treat it as a conversation:
- On failure — a message that assumes good faith. "Your payment for this month did not go through, most likely a balance issue. We will retry on the 5th, or you can pay now here: [link]"
- Before retry — a short heads-up.
- After the final failure — a clear statement of what happens and how to fix it.
WhatsApp substantially outperforms email for this in India, because it gets read. These are utility messages tied to a specific account, so they are among the cheapest you can send — see our WhatsApp API pricing comparison, and PayPerWA if you want to run the sequence without a monthly platform fee sitting on top of it.
Businesses that build a proper dunning sequence recover a large share of failed collections. Businesses that retry silently and cancel do not.
Grace periods and access control
Decide what happens between failure and resolution. Immediate cut-off maximises collection pressure and damages goodwill. An indefinite grace period trains people not to pay.
A short, clearly communicated grace period with degrading access usually works best. Whatever you choose, make sure the customer knows the rule in advance.
Reconciliation
Your records must match the gateway's settlements. Every collected amount, every failure, every refund, every settlement, matched daily.
Skipping this feels fine for six months and then becomes a very unpleasant audit. Build it early — it is far cheaper than reconstructing a year of transactions.
Proration and plan changes
Upgrades, downgrades, and mid-cycle changes. Decide the policy first, then implement it. This is a surprisingly large source of complexity and customer confusion, and simple rules beat clever ones.
Build or Buy
Buy for almost everyone. Your payment gateway's subscription product, or a dedicated billing platform, handles mandate registration, notifications, retries and invoicing. That is a great deal of detailed work that adds nothing to your product.
Our comparison of Razorpay, Stripe, PayU and Cashfree covers the Indian gateway landscape, and SaaS billing in India covers the subscription layer specifically.
Build only the unusual part. If your pricing model is genuinely non-standard — usage-based with unusual metering, complex multi-party splits, or an industry-specific structure — build that logic and let the platform handle mandates and collection.
Build the dunning yourself regardless. Standard dunning is generic and cold. Your version, over WhatsApp, in your voice, timed to your customers' salary cycle, will recover more. This is the highest-return custom work in the whole area.
Sector Notes
SaaS. UPI Autopay for smaller plans, cards and e-NACH for larger ones. Annual plans reduce failure exposure entirely, so price them attractively.
Coaching institutes and schools. Fee collection is the perennial administrative burden. Mandates plus WhatsApp reminders convert it from a chase into a process. Fee reminders are utility messages, so the cost is negligible relative to what they collect — see school management software and WhatsApp for coaching institutes.
Gyms and fitness. Classic manual-renewal churn, and one of the clearest cases for mandates. See gym and fitness management software.
Housing societies. Monthly maintenance collection is exactly this problem at scale, with a committee chasing defaulters. See society management software.
Healthcare and diagnostics. Package and plan payments in instalments, where clear communication matters more than aggressive collection.
D2C subscriptions. Replenishment products where the mandate and the delivery cycle must stay in step. A debit that fires when a delivery is paused is a support ticket you created.
A Sensible Build Order
Phase 1 — one mandate type, done properly. UPI Autopay for consumer, or e-NACH for high-value. Registration, collection, and a clean record of mandate state.
Phase 2 — failure handling. Retry logic timed sensibly, and the dunning sequence over WhatsApp. Expect this phase to produce the biggest revenue improvement of the whole project.
Phase 3 — reconciliation. Daily matching against gateway settlements, and a report that surfaces discrepancies.
Phase 4 — second mandate type. Add cards or e-NACH so you stop losing customers whose preferred method is missing.
Phase 5 — the edge cases. Proration, plan changes, pauses, refunds. Real, but not worth delaying launch for.
Notice that dunning comes before adding a second payment method. Recovering the customers you already signed up is cheaper than acquiring new ones, and it is usually the faster win.
Related Reading
- SaaS billing in India: Stripe and Razorpay
- Razorpay vs Stripe vs PayU vs Cashfree
- Fintech app development guide
- Razorpay integration
- GST-compliant software guide
Founder note: if you run a subscription business and cannot state your involuntary churn rate — customers lost to failed payments rather than to a decision — that number is almost certainly larger than you think, and it is the cheapest revenue you will ever recover. WhatsApp me on +91 92771 84741.