6.11 Segment overlay — E-commerce sellers
Why e-commerce sellers are distinct
Section titled “Why e-commerce sellers are distinct”Online sellers — whether marketplace-led (Amazon, Flipkart, Meesho, JioMart, Myntra) or direct-to-consumer (Shopify / WooCommerce store, own app) — have a data shape that bank-statement and GST analysis alone cannot capture well:
- Marketplace settlement cycle — sellers receive payment from marketplace
7 – 30 daysafter order; the bank statement shows lumpy weekly / fortnightly settlements, not per-order receipts. - Return and cancellation rates materially affect net sales — gross marketplace sales of
₹10 lakhmay net to₹7 lakhafter returns. - Multi-marketplace presence — many sellers operate on
3 – 5marketplaces simultaneously; consolidation across platforms is essential. - Ad-spend leverage — marketplace ad spend (Sponsored Ads) drives sales; the ROAS (Return on Ad Spend) is the operating metric.
- Inventory across fulfilment models — FBA / FAssured (marketplace fulfilled) vs SBA (seller fulfilled) — different working-capital dynamics.
- Catalog and category risk — apparel and beauty have higher return rates than electronics; some categories are restricted / heavily regulated.
- Promotional event cycles — Big Billion Days, Diwali, Republic Day, end-of-season — create demand spikes that distort monthly averages.
GST and bank-statement underwriting will materially under-state the cash-flow reality. The platform must integrate marketplace-settlement data to underwrite this segment well.
Standard rules that need adjustment
Section titled “Standard rules that need adjustment”| Standard rule | Default | E-commerce seller overlay |
|---|---|---|
| GST sales vs bank credit | <= 15% divergence for A | <= 25% acceptable; marketplace settlement timing creates legitimate gaps |
| Top buyer concentration | <= 40% | Not applicable — “buyers” are thousands of end customers; concentration measured at marketplace platform level instead |
| Inventory turn | >= 6 for traders | >= 10 for fashion / accessories; >= 6 for electronics; sub-segment specific |
| Monthly volatility | <= 50% CV for A | <= 80% acceptable due to event-driven spikes |
| Bank balance ABB | >= ₹50k for B | >= ₹30k acceptable if marketplace settlement is regular and visible |
E-commerce-specific rules to add
Section titled “E-commerce-specific rules to add”Marketplace platform concentration
Section titled “Marketplace platform concentration”| Rule | ecom_marketplace_concentration |
|---|---|
| Purpose | Single-marketplace sellers have platform-policy risk (account suspension, fee changes) |
| Data source | Borrower-provided marketplace settlement reports |
| Logic | Sales contribution from top marketplace |
| Threshold | <= 70% top platform for A; > 90% REFER |
| Action | Per |
Return / cancellation rate
Section titled “Return / cancellation rate”| Rule | ecom_return_rate |
|---|---|
| Purpose | High return rates indicate product or fit issues; net economics suffer |
| Data source | Marketplace settlement reports (gross sales − returns = net) |
| Logic | Returns / gross sales over 90 days |
| Threshold | <= 15% for electronics A; <= 25% for apparel A; per-category threshold |
| Action | Per |
Account standing
Section titled “Account standing”| Rule | ecom_account_status |
|---|---|
| Purpose | Suspended marketplace accounts have zero future cash flow |
| Data source | Borrower declaration; verifiable via account status check |
| Logic | All claimed marketplace accounts in good standing |
| Threshold | Hard |
| Action | DECLINE if any major account suspended |
Seasonality smoothing
Section titled “Seasonality smoothing”| Rule | ecom_seasonality_smoothed |
|---|---|
| Purpose | Big Billion Days / Diwali sales inflate one month’s GST disproportionately |
| Data source | GST + marketplace settlement |
| Logic | Use rolling 12-month for capacity assessment, not peak month |
| Threshold | Smoothed turnover for capacity |
| Action | Adjusts turnover scorecard |
Ad-spend ROAS
Section titled “Ad-spend ROAS”| Rule | ecom_ad_spend_efficiency |
|---|---|
| Purpose | Seller dependent on inefficient ad spend has compressed margin |
| Data source | Marketplace ad spend data + sales attribution (if shared) |
| Logic | Net margin after ad spend / revenue |
| Threshold | >= 8% net margin after ad-spend for A; < 3% REFER |
| Action | Per |
Category mix
Section titled “Category mix”| Rule | ecom_category_mix |
|---|---|
| Purpose | Some categories (jewellery, certain electronics) have higher fraud / return / regulatory risk |
| Data source | Borrower-declared category mix |
| Logic | Per-category allowed product caps; restricted categories DECLINE |
| Threshold | Per category table |
| Action | Per |
Cash-flow analysis overlays
Section titled “Cash-flow analysis overlays”- DSC threshold:
>= 1.8×standard for established sellers; account for marketplace settlement timing. - Working-capital cycle: Inventory-to-cash for marketplace seller is
marketplace_settlement_days + inventory_days − payable_days. Often30 – 60 days. - Bank balance pattern: Look for
weekly / fortnightly settlement spikes— marker of marketplace seller; absence may indicate the borrower is overstating marketplace revenue.
Marketplace settlement data integration
Section titled “Marketplace settlement data integration”The transformative data source for this segment is marketplace settlement reports. Each major marketplace provides:
- Amazon Seller Central: settlement reports (CSV / Excel) per settlement cycle.
- Flipkart Seller Hub: settlement reports per cycle.
- Meesho Supplier Panel: settlement / payment reports.
- Myntra Partner Portal: similar.
- Custom D2C (Shopify / WooCommerce): payment gateway settlement (Razorpay / Cashfree settlement reports).
These reports contain:
- Order count, gross sales, returns, commission, shipping, GST.
- Final net settlement to seller’s bank account.
- Cycle-by-cycle history.
The platform can:
- Ingest via API where the marketplace offers seller-data API (Amazon SP-API, etc.).
- Ingest via borrower upload of CSV / Excel.
- Parse and reconcile against borrower’s claimed turnover and bank inflows.
This data set is gold for underwriting this segment. Without it, the platform is underwriting blind to 60 – 80% of the real signal.
Inventory financing variant
Section titled “Inventory financing variant”For sellers carrying significant own-inventory (especially apparel and electronics), an inventory-backed line can be more economically efficient than pure WC. The line draws limit against verified inventory value; the asset itself serves as soft security.
Implementation depends on:
- Inventory verification (warehouse audit, photo, third-party inventory management).
- Marketplace FBA inventory data (Amazon FBA reports show inventory at Amazon fulfilment centres — useful proxy).
- Insurance on inventory.
Pricing and exposure
Section titled “Pricing and exposure”| Sub-segment | Pricing band |
|---|---|
Multi-marketplace seller with 2+ years history | Standard A |
| Single-marketplace concentration | Standard A + 50 – 100 bps |
| D2C own-store with stable repeat customers | Standard A − 25 bps |
New seller (< 12 months on marketplace) | C grade if eligible at all |
Ticket grid: usually capped lower than goods-distribution SMEs because of inventory turnover volatility, until track record builds.
Common fraud / pitfall patterns
Section titled “Common fraud / pitfall patterns”| Pattern | Detection |
|---|---|
| Account suspension undisclosed | Verify account standing |
| Inflated marketplace turnover (fake reports) | Cross-check with marketplace API if accessible; reconcile with bank settlements |
| High return rate undisclosed | Settlement report shows gross vs net |
| Multi-account violation (multiple seller accounts circumventing marketplace caps) | Pattern detection across borrower’s PAN |
| Inventory financed at multiple lenders (stacking) | Bureau + commercial bureau monitoring |
| Inventory shrinkage (FBA reconciliation gaps) | FBA reconciliation reports |
Data sources to prioritise
Section titled “Data sources to prioritise”- Marketplace settlement reports (essential).
- Bank statements with settlement-pattern recognition.
- GST 1 / 3B.
- D2C payment gateway settlement (for own-store sellers).
- Marketplace account-standing verification (where feasible).
- Borrower’s category mix declaration with verification via settlement reports.
Co-lending implications
Section titled “Co-lending implications”E-commerce seller portfolios are:
- Generally PSL-eligible if MSME-registered.
- Higher operational complexity for partner banks — marketplace settlement data is less familiar territory for traditional banks.
- Suitable for specialised partner programmes with banks that have an e-commerce / digital-distributor focus.
Most major Indian banks now have e-commerce seller programmes; co-lending pools specific to this segment are increasingly common.
Sources
Section titled “Sources”- Marketplace seller portals — Amazon Seller Central, Flipkart Seller Hub, Meesho Supplier Panel, etc.
- Amazon Selling Partner API (SP-API) documentation —
developer-docs.amazon.com/sp-api. - Standard SME WC underwriting (see 6. Underwriting).