UPI AutoPay vs eNACH Cost Comparator

UPI AutoPay beats eNACH on the two dimensions that matter for subscription economics: cost basis and success rate. The bank-account UPI rail underlying AutoPay carries zero network MDR (mandated since 1 January 2020 under Section 10A of the Payment and Settlement Systems Act and Section 269SU of the Income-tax Act 1961). Gateway platform fees on AutoPay are typically 0.50 to 0.80 percent of the debit, materially below the eNACH per-debit fee of roughly Rs 15 plus GST regardless of ticket. For tickets at or below Rs 1,000 — where Rs 15 represents 1.5 percent or more of the debit — UPI AutoPay is structurally cheaper. UPI AutoPay also runs at 94 to 97 percent success on first attempt versus eNACH at 85 to 90 percent (bank-mandate processing depends on physical signature-equivalent verification, batch debit windows, and a longer rejection-and-retry cycle that compounds cost). The cross-over is ticket size: above Rs 15,000 UPI AutoPay needs additional-factor authentication on every debit (NPCI mandate), which reintroduces a friction step that eNACH does not have once the mandate is registered. For tickets between Rs 500 and Rs 15,000 — the OTT, edtech, SaaS and lending-EMI band — UPI AutoPay is the cheaper rail. For tickets above Rs 15,000 or for institutional debits where the customer prefers a bank-mandate paper trail, eNACH remains the right rail.

Under the current NPCI UPI AutoPay framework, the additional-factor authentication (AFA) thresholds are Rs 15,000 for default mandates and Rs 100,000 for specific categories (mutual fund SIPs, insurance premiums, broker margin, government, education) under the enhanced cap framework. Below the applicable threshold, the debit happens silently on the mandate authority without re-prompting the customer; above the threshold, the customer must approve each debit via UPI PIN at the time of debit. The practical effect on subscription merchants is that an OTT subscription at Rs 199 to Rs 999, an edtech course at Rs 1,499 to Rs 4,999, or an SME SaaS plan at Rs 2,499 to Rs 9,999 all sit comfortably under the Rs 15,000 default threshold and run silently. A digital lending EMI above Rs 15,000, a high-end insurance premium, or a B2B SaaS plan in the Rs 25,000 range starts requiring AFA on every collection, which materially degrades the success rate (every debit becomes an active customer touchpoint) and tilts the economics back toward eNACH. The comparator flags this when ticket size crosses Rs 15,000.

eNACH success rate hovers in the 85 to 90 percent range against UPI AutoPay's 94 to 97 percent for three structural reasons. First, eNACH is a batch rail: debits are submitted to the destination bank in NPCI clearing windows, processed at the destination bank's end-of-day or next morning, and any failure surfaces with a 24 to 48 hour lag — versus UPI AutoPay's real-time response and immediate retry. Second, eNACH rejection codes are heavily concentrated in insufficient balance (the most common code, accounting for 60 to 70 percent of rejections), account dormant, signature mismatch in the bank's mandate registry, and account closed — categories that require customer outreach to resolve and a fresh debit attempt in the next cycle. Third, the eNACH mandate registry at the destination bank is a separate write that occasionally diverges from the bank's core CASA system after KYC re-verification, account migration, or branch consolidation, producing rejections that the customer cannot self-diagnose. UPI AutoPay's mandate sits within the UPI rail and shares the customer's UPI handle authority, which avoids the bank-mandate-registry drift entirely. The success-rate delta typically costs 2 to 4 percent of recurring GMV in retry overhead, dunning cost, and involuntary churn — a figure the comparator quantifies against the chosen rejection-retry cost setting.

Yes, and many subscription businesses do. The two common patterns are: rail-by-ticket (UPI AutoPay default for tickets under Rs 15,000, eNACH for above) and rail-by-segment (UPI AutoPay for D2C consumer tickets, eNACH for SMB and institutional). A third less common pattern is rail-by-fallback: try UPI AutoPay first; if the customer's bank is not on the UPI AutoPay live list or the mandate registration fails, fall back to eNACH at onboarding. Running both rails adds reconciliation complexity — you have two separate settlement files (UPI AutoPay flows through the payment aggregator and into the merchant nodal account; eNACH flows through the sponsor bank's NACH window into the operating account), two separate mandate management surfaces, and two separate retry schedules. The reconciliation discipline is to tag every subscriber to a primary rail at onboarding, treat any rail switch as a re-mandate event with audit trail, and reconcile rail-level success and cost separately every month. The comparator is built for the single-rail comparison; for a mixed-rail business, run it twice (once on each rail's volume and ticket profile) and aggregate.

Recurring debits sit in a separate corner of the MDR reconciliation from one-time card and UPI volume, and the bookkeeping conventions diverge across gateways. Three lines must be reconciled per recurring debit. First, the base network MDR or per-debit fee — zero on UPI AutoPay's bank-account rail, the per-debit eNACH fee on the eNACH rail. Second, the recurring add-on, which is the gateway's subscription module fee (commonly the 0.99 percent recurring add-on on Razorpay or a comparable line on PayU and Cashfree) charged on top of base MDR for the subscription orchestration layer — mandate creation, scheduling, dunning, retry. Third, GST at 18 percent on the MDR plus the recurring add-on combined. The MDR Effective-Rate Calculator and MDR Leakage Flag Checker on this site cover the per-network reconciliation; this comparator handles the rail selection decision that precedes it. If you have a recurring business at scale, the audit trigger is the same 0.15 percentage point variance threshold against the contracted recurring rate — but applied to the recurring subset of GMV separately, not folded into the blended one-time GMV rate. Folding recurring into the blended rate is one of the most common ways recurring add-ons go unnoticed.