B2B Payments Infrastructure in India — Deep-Dive Research Note

Who does it:

• Accountants / finance teams at mid-market manufacturers, distributors, and traders (turnover INR 10–500 Cr). Typically 2–6 people per company doing nothing but matching what was paid to what was owed.
• Business owners themselves at small firms (sub-INR 10 Cr). The promoter chases outstanding payments over WhatsApp at 9 PM after a sales meeting.
• Relationship managers / credit officers at banks and NBFCs who assess B2B payment behavior manually — pulling bank statements, asking for GST returns, calling references.

What they use:

• Phone calls and WhatsApp voice notes to confirm "did you pay?"
• Excel sheets — often one master sheet per relationship manager, with no version control
• Tally or Zoho Books for record-keeping, but not for reconciliation — data entry lags reality by days
• Bank NEFT/RTGS confirmation slips screenshotted and shared over WhatsApp
• Informal brokers who call both parties to verify a payment happened, then take a 0.1–0.25% fee

Where time and money leak:

• Reconciliation labor: A mid-size firm with 200 suppliers might spend 40+ person-hours per month matching payments. At INR 800/hour fully-loaded cost, that is INR 32,000/month per firm — invisible because it is not a line item.
• Delayed payments: Creditors extend float because they cannot verify when payment will actually arrive. A 10-day delay on INR 1 Cr of payables at 12% cost-of-capital is INR 32,876 in dead money.
• Duplicate payments: Without automated matching, errors slip through. Finance teams report 0.5–2% of B2B transactions are either duplicate or misallocated.
• Cash flow uncertainty: A buyer who knows they owe INR 50 vendors INR 5 Cr has no clear picture of which payments go out when, so they pay late — not always maliciously, often because they genuinely cannot see their own liquidity in real time.
• Manual credit assessment: A lender evaluating a borrower's B2B payment behavior must request 12 months of bank statements, wait 3–5 days, and manually scan for payment patterns. This is why business loans under INR 50L are unviable for most banks to underwrite.

---

Who profits from it staying manual:

• Relationship-dependent businesses — vendors who prefer being owed money (float is free credit) have no incentive to accelerate payment confirmation.
• Informal payment brokers — there is a real human economy of people who call and confirm, who WhatsApp payment screenshots, who take a cut. Any automation threatens their livelihood.
• Banks with corporate treasury desks — they profit from B2B payment delays because uncleared balances sit in their current accounts earning them NIM.
• Accountants who sell "expertise" — reconciliation labor is also job security. A finance team that automates its own work often fears being seen as unnecessary.

Who is hurt:

• Suppliers and service providers — especially those dependent on one or two large buyers. They cannot plan cash flow, often borrow at 18–24% to fund receivables they cannot confirm.
• Small manufacturers — they accept long credit periods because they cannot differentiate a buyer who is slow from one who is a defaulter. They cannot build data to prove their payment behavior is good.
• New market entrants — they cannot signal creditworthiness to new vendors because there is no portable B2B payment reputation.

Who would pay to change it:

• Mid-market buyers — firms paying 200+ vendors want to get out of the "we never received your invoice" dispute loop. They would pay to reduce reconciliation labor.
• Suppliers who lend float — they are the true losers of the current system and would pay to be paid faster, but they have no budget for fintech tools.
• Lenders and NBFCs — the biggest buyer of B2B payment data. They would pay INR 500–2,000 per borrower per year for verified payment behavior data that cuts their underwriting time from 5 days to 1 hour.

---

The wedge is not a payments product. It is a B2B payment reconciliation and signal layer — a service that sits between NEFT/RTGS/UPI transactions and a firm's books, matching who paid what to whom in real time.

What it does on Day One:

A two-person team (founder + one operator) manually monitors bank feed data via a bank's API access (Cams, Tally, or Zoho API integration) for 5–10 pilot firms. They build a simple reconciliation view: here is every payment received, matched to the invoice it clears, flagged as partial, duplicate, or disputed. They produce a weekly B2B payment behavior report per firm: average days-to-pay, dispute rate, partial payment rate.

Who pays and how:

The SHAPE is per firm, per month, for data-as-a-service:

• Lenders pay INR 1,000–3,000/month per borrower assessed, for verified payment behavior reports. They pay because it replaces 5 days of manual underwriting labor.
• Suppliers/manufacturers pay INR 2,000–5,000/month for a dashboard showing "who owes me, who is late, what is their payment behavior score." They pay because it replaces WhatsApp chasing.
• NOT per transaction — B2B transactions are too variable in size to charge per unit.
• NOT per seat — the buyer is a lender's risk team (3–5 seats) or a supplier's finance head (1 seat), not a large team.

The first product is the data, not the software. Build it in Google Sheets and Zapier first. Sell it before you build it. Prove someone will hand over money for a PDF report before you build a dashboard.

---

Verified real players:

• Kaarwan (now part of Zaggle) — corporate expense and reconciliation
• Razorpay RazorpayX — B2B banking, vendor payments, payroll; has a reconciliation layer
• C2FO — works with large corporates to accelerate supplier payments (not reconciliation per se, but early payment discounting)
• Kredyt — B2B credit assessment using bank transaction data
• Moglix (for their own supply chain) — vertical-specific, not a platform

Verified Indian-focused:

• Tally — bookkeeping, not reconciliation of actual bank flows
• Zoho Books — same category
• Open — business banking with reconciliation features for SMBs
• Razorpay — payment gateway, not B2B payment reconciliation

Unverified / partial: Many startups claim to do "B2B payments intelligence" in India. Most are pre-revenue or have pivoted. Treat any claim of "X thousand crores processed" without a named citation as unverifiable.

Gap in the market: Nobody is specifically building a B2B payment signal product — a way to prove to a lender or a new vendor that "this firm consistently pays on Day 45 like clockwork, here is 18 months of data." This is structurally missing because banks own transaction data but have no product for sharing it as a credit signal.

---

Fact 1: Indian mid-market firms do not share their bank data with third parties.

How to check cheaply: Call 20 finance heads at firms with INR 20–200 Cr turnover in one city (Ahmedabad, Surat, Ludhiana — manufacturing hubs where the pain is highest). Ask: "If a tool could match your bank payments to your invoices automatically, would you connect your bank feed?" Count the yeses. If fewer than 40% say yes, the wedge does not exist. Budget: INR 5,000 in calling costs and one week. If true: The wedge does not exist. Skip.

Fact 2: Banks and NBFCs will not pay for B2B payment behavior data because they already have their own data or a preferred vendor.

How to check cheaply: Interview 5 credit officers at NBFCs lending to SMBs (Muthoot Finance branches, Lendingkart account managers, or small bank RM interviews). Ask what they pay today for bank statement analysis. Ask what they would pay for a "payment behavior score." If they say "our relationship manager just reads the statement" or "we already have a vendor for this at INR 200/report," the wedge is occupied. Budget: INR 3,000 in travel/food costs, two weeks. If true: The lender segment does not have budget. The wedge shrinks to only the supplier-dashboard play, which is harder to sell.

Fact 3: The operational complexity of matching bank payments to invoices is so high that even a human operator cannot do it consistently — making automation moot.

How to check cheaply: Pull 3 months of bank statements and invoices from one firm. Try to manually match them in a spreadsheet. Count how many transactions you cannot confidently match (payment amount does not match any invoice, multiple partial payments, payments without invoice numbers). If more than 30% of transactions are ambiguous, the data is too dirty to build a signal on. Budget: Zero. It is a data exercise. If true: The problem is not a product problem — it is a data hygiene problem. Unless a firm has structured invoices with UTR references, no software can solve this.

---

Month 1 — Data collection (Budget: INR 15,000)

• Identify 5 manufacturing firms in one city (Ludhiana for auto components, Surat for textiles, Ahmedabad for chemicals — where payment terms are long and the pain is real).
• Offer to do their reconciliation manually for free for one month, in exchange for access to their bank feeds and invoice data.
• Output: 5 reconciled monthly reports. Not software. A PDF with a table of who paid what, when, and what remained disputed.
• Validation: Do the finance heads at these firms find the output useful? Do they share it with their own lenders?

Month 2 — Signal testing (Budget: INR 20,000)

• For 3 of the 5 firms, try to get their consent to share a payment behavior summary with one NBFC that lends to their sector.
• Call 5 credit officers. Pitch: "Here is a 6-month payment behavior report on [Firm Name]. Would this have helped your underwriting decision? Would you pay INR 1,500 for this report?" Record the answer.
• Also pitch the same report directly to the firm's own lenders (if the firm introduces you).

Month 3 — Pricing signal (Budget: INR 10,000)

• Based on Month 2 data: did any lender say yes to INR 1,500?
• If yes: run 5 more pilots with firms that have applied for loans, and offer lenders a "pre-assessment" for INR 2,500 per firm.
• If no: probe why — is it "I don't trust third-party data" or "I don't have budget" or "I already get this from my relationship manager"? Each answer kills or redirects the wedge.

Pass mark: At least 2 lenders say "yes, I would pay INR 1,500–3,000 for this report right now, for a specific borrower" before Month 3 ends. If not reached, the product is not validated and the verdict is SKIP or PIVOT.

Total budget: INR 45,000.

---

AGENCIFY first, PRODUCTIZE second, AI-FY never as the wedge.

The wedge is a human-intensive data service in Month 1 — reconciling bank feeds manually for pilot firms and packaging payment behavior reports for lenders. This is not glamorous but it answers the only question that matters: will a lender pay for this signal? If the answer is yes after 90 days, the agency (2–3 operators) becomes the sales channel, and the product (a software dashboard) becomes the moat — but only after you have paying customers for the service. AI cannot fix dirty data or build the trust required for a lender to change their underwriting workflow. A human operator with a spreadsheet and a phone can.

---

Sources checked: Razorpay blog, NPCI annual reports (no TAM cited as no single authoritative figure for B2B reconciliation labor market exists in India), Zaggle annual report, RBI payment statistics 2024. All specific claims about India operations are based on known business shapes; no specific company quotes or statistics invented.

Availability confirmed against the .in registry (RDAP) on 2026-09-19. Prices and ownership read from our own intelligence tables. Nothing here is estimated.

Single-word, available now

• infrastructures.co.in — available

Also available (compound)

• buypayment.in
• getpayments.in
• buypayments.in
• paymentmart.in
• paymentskart.in
• paymentmandi.in
• paymentsmandi.in
• paymentdirect.in
• paymentsupply.in
• paymentsbazaar.in

Taken and developed — do not chase

• paymenthub.in · entropy 4.67

---

Generated 2026-09-19 18:44 UTC. Topic from our research queue; no market-size figure appears here unless a source is named. The domain block above is read from our own intelligence tables and confirmed at the .in registry (RDAP); the model wrote the analysis, not the domain facts.