Content authored by practitioners with experience at Amazon India, Intuit QuickBooks, and the Tata Group. Meet the team →
Indian subscription merchants in the ₹149 to ₹2,499 ticket band — OTT, B2B SaaS, edtech, hospitality membership, consumer NBFC EMI — pick a recurring-debit rail at contract stage without a quantified per-rail unit-economics model, then absorb a structurally avoidable fee burden when they default to eNACH at scale. The ratio between UPI AutoPay total cost and eNACH total cost at this ticket band is typically of the order of ten to fifteen times in AutoPay's favour for a book with eighteen thousand active subscribers, and the reliability gap on first-attempt success reinforces the cost gap because every eNACH rejection is a fresh fee event.
Build a per-rail unit economics model with three line items per rail — successful-debit cost, failed-debit cost, and platform fee — using contracted rates for UPI AutoPay and the published per-debit sponsor-bank charge for eNACH. Multiply by expected success and rejection rates for the relevant customer segment. Apply the monthly active mandate base. Compute the rail-mix-weighted total against the contracted blended rate and surface the gap. Reconcile the actual fee column on the settlement file against the modelled expectation per debit, by rail. Flag any non-zero network MDR on UPI AutoPay lines and any per-debit rejection charge not classifiable against an NPCI return reason code.
Per-mandate registry by rail with creation date, status, customer segment, ticket size, and contract reference. Per-rail rate sheet — UPI AutoPay gateway platform fee schedule and eNACH per-debit and per-rejection schedule. NPCI return reason code dictionary for eNACH failure classification. Rail-mix model with expected success rate and rejection rate by segment. Settlement file pipeline split by rail. GST 18 percent overlay on fee components. Variance register feeding the subscription unit economics dashboard.
A monthly subscription rail dashboard showing per-rail debit count, success rate, failed-debit fee burden, platform-fee burden, and GST on each, totalled against the modelled unit economics. A per-rail leakage report flagging non-zero network MDR on UPI AutoPay lines and unclassified rejection charges on eNACH lines. A quarterly rail-mix recommendation brief setting target migration from eNACH to UPI AutoPay for ticket bands under ₹15,000 with quantified annual savings, supported by the reconciliation evidence base.
A mid-sized OTT platform headquartered in Mumbai runs a ₹2,499 quarterly plan with 18,000 active subscribers. The finance controller picked eNACH as the recurring rail at launch because the sales engineer at the payment aggregator was more fluent in the NACH product, and the legal team was familiar with NPCI’s mandate format. Twenty-four months later, the controller pulled the per-rail cost line and discovered that eNACH was costing the business roughly ₹3.18 lakh a month in per-debit and per-rejection fees, while a UPI AutoPay-led rail mix would have cost closer to ₹27,000 a month at the same volume. The gap is real, it is structural, and it sits at the centre of every subscription-rail decision in the Indian market today.
This article is for the controller or CFO at an OTT, SaaS, edtech, hospitality membership or consumer NBFC EMI business who is sitting on a recurring book in the ₹149-₹2,499 ticket band and trying to decide whether to migrate to UPI AutoPay, stay on eNACH, or run a mix. The framework here is the per-rail unit economics model, the worked example is an OTT business at the median ticket of the band, and the reconciliation discipline at the end keeps either rail honest.
Quick reference: UPI AutoPay and eNACH at the ₹149-₹2,499 band
| Aspect | UPI AutoPay | eNACH |
|---|---|---|
| Network MDR on bank-account debit | 0% by zero-MDR mandate | n/a (per-debit fee structure, not MDR) |
| Per-successful-debit fee | Gateway platform fee on the recurring execution (percentage of ticket) | Around ₹15 plus 18% GST per debit, sponsor-bank pass-through |
| Per-failed-debit fee | Gateway platform fee on the attempt (varies by contract) | Around ₹15 plus 18% GST per failed debit, on every retry attempt |
| Auto-debit threshold without additional-factor authentication | ₹15,000 per debit | n/a (mandate ceiling set at registration, typically ₹5,000-₹1,00,000) |
| Typical first-attempt success rate at this ticket band | Higher (PSP-side affordances, real-time balance view) | Lower (window-based, sponsor-bank dependent) |
| Customer-side mandate update | Tied to UPI handle, often refreshed at PSP onboarding | Tied to bank account, decays silently on bank switch |
| Regulatory framework | NPCI UPI AutoPay e-mandate framework | NPCI NACH and eNACH operating circulars |
| Legal basis for zero network MDR | Section 269SU Income-tax Act, Section 10A PSS Act 2007 | n/a (eNACH is a per-debit fee rail, not a network-MDR rail) |
Why the rail-mix decision is a CFO-level question
A subscription merchant is not really making one decision when it picks a recurring rail. It is making three decisions simultaneously. The first is the per-debit fee structure — a percentage rail at small ticket sizes (UPI AutoPay) versus an absolute per-debit rail (eNACH). The second is the reliability profile — first-attempt success rate, retry economics, mandate decay over time. The third is the customer experience — onboarding friction at mandate creation, pause-and-resume affordance, dispute resolution path.
The three decisions interact. A rail with a very low per-successful-debit cost but a high failure rate can still cost more than a rail with a higher per-debit cost and a lower failure rate, because every failure is a fresh fee event. A rail with a benign per-debit cost at low tickets can become punitive at high tickets because the per-debit fee does not scale. A rail with friction at mandate creation can produce a smaller active book even if it is cheaper per debit, because conversion at the checkout falls.
For the ₹149-₹2,499 ticket band — the band that captures most consumer OTT, SaaS, edtech subscriptions and the smaller NBFC consumer EMI products — UPI AutoPay wins on the first decision (per-debit fee structure) and on the second decision (reliability), and is roughly at parity or slightly better on the third decision (onboarding friction has fallen materially as PSPs have built dedicated AutoPay flows). For larger ticket bands above ₹15,000, where the additional-factor-authentication threshold bites, eNACH remains the structural choice; that is a different article.
What does UPI AutoPay actually cost at this ticket band?
The network MDR on a UPI AutoPay debit is zero. This is the same zero-MDR regime that covers all UPI peer-to-merchant bank-account transactions, set out in Section 269SU of the Income-tax Act and Section 10A of the Payment and Settlement Systems Act 2007. The mandate ceiling for an auto-debit without additional-factor authentication at every debit is currently ₹15,000 per debit attempt. Within that ceiling, the customer’s PSP executes the debit headlessly on the due date.
The merchant does, however, pay a gateway platform fee on the recurring execution. The platform fee is the aggregator’s billed charge for running the mandate workflow — token storage, periodic-debit triggering, retry handling, mandate-status synchronisation with the customer’s PSP. The platform fee is typically a small percentage of the ticket or a small absolute number, depending on the contract. It is not network MDR; it is a separate line. The structural point for reconciliation is that any non-zero network-MDR component on a UPI AutoPay debit line is leakage by definition, because the network MDR is zero by law. The reconciliation discipline section below treats this as the single most important deterministic flag for the AutoPay rail.
What does eNACH actually cost at this ticket band?
eNACH carries a per-debit charge that is a flat absolute number, not a percentage of the ticket. The published range across aggregators is around ₹15 plus 18% GST per debit attempt, with some variation by sponsor-bank pair and by aggregator tier. The same charge applies to successful debits and to failed debits, which means every rejection is a fresh fee event. The retry economics compound the cost because a typical merchant configures a two-cycle retry policy; a 7% first-attempt rejection rate with a 60% second-cycle persistence produces a multiplier on the rejection fee of roughly 1.6 times the headline rejection rate.
The mandate ceiling on eNACH is set at registration and can be substantially higher than the ₹15,000 UPI AutoPay threshold, which is why eNACH remains the structural choice for institutional collections and high-ticket NBFC EMI books. For the ₹149-₹2,499 ticket band that this article addresses, the ceiling is irrelevant; what matters is the per-debit absolute fee, which becomes a much larger fraction of the ticket as the ticket shrinks. At ₹199, a ₹15-plus-GST debit fee is approximately 9% of the ticket. At ₹499, it is roughly 3.5%. At ₹2,499, it is around 0.7%. UPI AutoPay’s gateway platform fee at all three points is materially smaller, because it scales with the ticket and the base rate is contracted at a level below the absolute eNACH fee.
Worked example: 18,000 OTT subscribers at ₹2,499
Take the mid-band Mumbai OTT business introduced at the top. The plan is ₹2,499 per quarter; the active subscriber base is 18,000; the quarterly GMV is ₹4.5 crore. For modelling purposes, assume the merchant runs a single debit per subscriber per quarter on the recurring rail.
On the eNACH rail, the per-debit fee is around ₹15 plus 18% GST, which works out to ₹17.70 per debit. Multiplied by 18,000 active mandates with one debit per cycle, the per-cycle fee is approximately ₹3.18 lakh. That number ignores the first-attempt rejection rate, which would push the effective cost higher because every retry is a fresh fee. At an industry-typical 7% first-attempt rejection rate with a two-cycle retry policy, the effective rejection multiplier on the fee count is roughly 1.6 times the base rejection rate, which adds approximately ₹35,000 to the cycle cost. Round the total to ₹3.5 lakh per cycle as a defensible upper bound.
On the UPI AutoPay rail, the gateway platform fee on a recurring debit is typically a percentage of the ticket. At a contracted platform fee of around 0.6% on the recurring transaction, the per-cycle fee on a ₹4.5 crore quarterly GMV is approximately ₹27,000. The network MDR is zero. The first-attempt success rate on UPI AutoPay at this ticket band is meaningfully higher than on eNACH, so the failed-debit fee adjustment is small — pencil it at a few thousand rupees per cycle and the total stays well under ₹35,000.
The ratio is approximately 11.8 times — eNACH costs the business roughly twelve times what UPI AutoPay costs at the same volume, at the same ticket band, on the same active book. On a quarterly basis the absolute gap is around ₹2.9 lakh; on an annualised basis it is around ₹11.6 lakh. That is the size of the migration prize for this book.
The exact ratio is sensitive to three inputs. The contracted UPI AutoPay platform fee — at 0.6%, the ratio is 11.8 times; at 1.0%, it falls to roughly 7 times. The per-debit eNACH fee — at ₹15, the ratio is 11.8 times; at ₹10, it is roughly 8 times. The active mandate base — the ratio is invariant to base, but the absolute saving scales linearly with it. A merchant should run this model at its own contracted rates and at its own active book before committing to a migration decision.
UPI AutoPay vs eNACH cost comparator
Model your own subscription book by ticket size, active mandate count, and rejection-rate assumptions. The comparator returns the per-cycle and annualised cost on each rail, the migration-savings number you can defend to a CFO, and the rejection-rate breakeven where the rails cross over.
Open the tool →Where does the reliability gap actually come from?
The cost gap is roughly twelve times. The reliability gap is the second leg of the argument, and it amplifies the cost gap because every failed eNACH debit is a fresh fee event.
UPI AutoPay debits execute through the customer’s PSP application. The PSP has a real-time view of the linked-account balance, which means it can fail a debit cleanly before it hits the destination bank if the balance is short. The customer typically gets an inline notification — a “your auto-debit failed, please top up” prompt that lands on the PSP home screen — and can act on it within hours. The merchant sees the rejection within the same daily settlement window and can retry on the same day. The rejection economics are governed by the same UPI dispute framework that the PSP and the issuing bank already operate at scale for one-time UPI transactions.
eNACH debits execute through the NACH messaging cycle. The sponsor bank pushes the debit instruction to the destination bank in a batch window; the destination bank applies the debit on its end and returns either a success or a rejection with a return reason code drawn from the published NPCI taxonomy. The customer sees the rejection only via SMS or statement, not through a transactional interface that prompts immediate action. The merchant retries on the next scheduled cycle, which can be days later. Mandate decay — the silent inactivation of a mandate when a customer closes the linked account or switches banks — is harder to detect on the eNACH rail because there is no PSP-side handle synchronisation; the merchant discovers it only at the next failed cycle.
The first-attempt success rate gap is therefore real and structural. For a consumer subscription book in the ₹149-₹2,499 band, UPI AutoPay’s first-attempt success rate is typically several percentage points higher than eNACH’s, and the recovery on retry is faster. The reconciliation discipline section below covers how to measure this in the merchant’s own settlement file, not from industry averages.
Per-mandate reconciliation discipline
The reconciliation discipline that keeps both rails honest runs two parallel tracks against a single per-mandate registry.
For the UPI AutoPay track, the registry holds the mandate identifier, the customer reference, the linked UPI handle, the ticket size, the mandate-creation date, and the current status. The settlement file pipeline lands the per-debit execution record with the platform fee broken out. The reconciliation rule is to compute the expected platform fee at the contracted rate against the actual fee column, transaction by transaction. Any non-zero network-MDR component on a UPI AutoPay line is a deterministic leakage flag — the network MDR is zero by current law, and a non-zero value is either a billing error or a misclassified instrument. This is the same leakage pattern set out in the merchant-fees cluster cornerstone article on zero-MDR UPI and RuPay debit leakage, applied to the recurring rail.
For the eNACH track, the registry holds the mandate identifier, the customer reference, the linked bank account, the ticket size, the mandate-creation date, the sponsor-bank reference, and the current status. The settlement file pipeline lands the per-debit attempt with the success or rejection outcome, the return reason code if rejected, and the per-debit fee. The reconciliation rule is to classify every rejection against the published NPCI return reason code dictionary, attribute the attempt to a retry cycle, and total the per-debit fee plus GST. The variance against contract is the gap between the modelled per-debit fee at the contracted sponsor-bank schedule and the actual fee column. This pattern is the same as the broader stacked-fee architecture set out in the cluster article on recurring add-on and eNACH mandate-rejection fees, focused here on the AutoPay-versus-eNACH choice rather than the card-on-file stack.
Both tracks feed a monthly subscription unit-economics dashboard with five line items per rail: active mandate count, debit count, first-attempt success rate, fee burden, and effective cost per active mandate. The dashboard is the artefact the CFO uses to make the rail-mix decision quarter on quarter. The per-mandate registry is the artefact the auditor uses to verify the dashboard. The reconciliation discipline is what keeps both honest, and what surfaces the leakage in either rail before it becomes a year-end surprise.
A second discipline that pairs with the per-rail reconciliation is the per-instrument MDR audit on the broader gateway book, which is the framing the cluster article on Razorpay MDR reconciliation develops in full. The subscription rail decision sits inside that larger reconciliation programme as one slice; the principles — separate the network MDR from the platform fee, separate the base MDR from any add-on, separate the GST line — are the same across both.
What does the rail-mix decision look like in practice?
For a new subscription book at the ₹149-₹2,499 ticket band launching today, the default rail mix should be UPI AutoPay as the primary rail, with eNACH offered as a secondary rail to customers who explicitly prefer bank-account debit through the NACH framework. The migration story for an existing book — a merchant carrying twenty-four months of eNACH mandates and an active customer base — is more nuanced. The migration is customer-by-customer because each customer has to re-authenticate the mandate on the new rail; the merchant cannot move the mandate silently.
The economically sensible sequence is to default new acquisitions to UPI AutoPay, to surface the rail-switch option at every customer touchpoint (renewal reminders, support interactions, app updates) for the existing book, and to monitor the mix-shift quarter on quarter. The migration savings compound as the eNACH book runs off and the UPI AutoPay book grows. For the OTT business in the worked example, a thirty-month migration window with a steady mix shift produces a cumulative saving of several tens of lakhs, which is a defensible board-level number.
The rail-mix decision is reversible at low cost. The reconciliation discipline is the part that has to land first, because without it the merchant cannot measure either rail’s actual cost in its own book and cannot defend the migration decision with evidence.
Continue reading in the merchant-fees cluster
- Recurring add-on and eNACH mandate-rejection fees: stacked costs for subscription merchants — the broader card-on-file plus eNACH stack, the 0.99% recurring add-on architecture, and the per-mandate reconciliation discipline that surfaces both rails together.
- Razorpay MDR reconciliation: per-instrument audit path — the per-instrument MDR audit framing that the subscription rail decision sits inside.
- MDR charged on zero-MDR UPI / RuPay debit: the most common leakage pattern — the deterministic leakage flag for any non-zero network MDR on a zero-MDR instrument, applied here to the UPI AutoPay recurring rail.
- Merchant fees cluster hub — the full set of patterns, money page links, and the leakage taxonomy.
External authority: NPCI — UPI AutoPay product overview and circulars, for the operative UPI AutoPay e-mandate framework, the ₹15,000 additional-factor-authentication threshold, and the published e-mandate processing flows that govern the recurring-debit lifecycle and dispute classification.