Wholesale & Distribution

Primary sales: Brand sells to its direct customers — super stockists or CNF (Carry & Forward) agents. This is the brand's recognized revenue. Tracked in brand ERP (SAP). Secondary sales: Super stockist / CNF sells to distributors. This is the closest proxy to market demand. Tracked in DMS. Tertiary sales: Distributor sells to retailers (kiranas). This is actual sell-out — what reaches the trade. Hardest to measure accurately; estimated via DMS secondary data or direct retailer surveys. Why it matters: Brands want to align primary with secondary+tertiary — over-filling the channel (stuffing) is a red flag in audits. Real challenge: India has 1M+ distributors, many on paper-based systems — DMS adoption is still < 40% making accurate secondary data a major data engineering problem.

Kirana credit scoring for B2B BNPL uses a mix of traditional and alternative data: 1) Traditional signals: CIBIL/Equifax score (available for ~40% of kiranas), GST return filing history (GSTIN), bank statement analysis via Account Aggregator. 2) Platform signals: Order history on the platform — order frequency, GMV, return rate, payment history for past invoices. 3) Behavioural signals: App usage patterns, response time to delivery confirmations, referral network. 4) Bureau + alternative model: Udaan and similar platforms build proprietary ML credit models combining all signals. Since many kirana owners lack formal credit history, OkCredit/Khatabook ledger data is valuable — showing cash flow, payment behaviour with local suppliers. Credit limit is dynamic — increases as payment history builds. Collections: UPI autopay mandate on due date; field collection agent for large defaults.

E-invoicing (IRN): Mandatory under GST for businesses above ₹5 Cr turnover. Every B2B invoice must be reported to the Invoice Registration Portal (IRP — run by NIC/GSTN). IRP validates invoice, generates IRN (Invoice Reference Number) and QR code. Invoice is only legally valid after IRN generation. In DMS/ERP: Invoice raised in system → API call to IRP → IRN returned → printed on invoice PDF. Failure to have IRN means buyer cannot claim GST Input Tax Credit (ITC). e-Way Bill: Required for goods movement above ₹50,000 in value across states (or within state >10 km in some states). Generated on the GST e-Way Bill portal. Contains: supplier, recipient, goods, vehicle number, transporter. Driver must carry or show it during transit; intercepted by tax authorities otherwise. In TMS/DMS: Automatically generated when dispatch is created for qualifying orders via the NIC API.

Design for real-time secondary sell-out visibility: 1) Data collection layer: Distributor DMS (Bizom, custom) captures every secondary invoice in real-time. SFA app captures van sales and retailer orders at POS. Challenge: many distributors use Tally — build a Tally plugin or daily file export integration. 2) Event streaming: Each secondary invoice is an event pushed to Kafka. Brand's analytics platform consumes the stream — sub-minute latency. 3) Data normalisation: Distributor SKU codes ≠ brand SKU codes. A master data layer maps distributor's local codes to brand's EAN/SKU codes. 4) Brand analytics platform: Aggregated dashboard — sell-out by SKU, region, distributor, outlet type. Compare against primary sales (shipments) to calculate pipeline inventory (primary - secondary). 5) Alerting: Detect fast-moving SKUs going out-of-stock (secondary > primary replenishment). Trigger auto-replenishment order to distributor. Scale: 3,000 distributors × 500 invoices/day = 1.5M events/day. Kafka + Flink stream processing, Redshift for analytics.

Route-to-Market (RTM) is the strategy that defines how a brand gets its products from factory to consumer — which channels, which distribution tier, direct vs indirect. Types: 1) Traditional General Trade: Brand → Stockist → Distributor → Kirana. 2) Modern Trade: Brand → Key Account Manager → Supermarket (BigBazaar, Reliance Smart). 3) D2C: Brand direct to consumer via own website. 4) B2B marketplace: Via Udaan/JioMart. Technology enables RTM via: Beat optimisation — algorithm determines optimal number of outlets per beat, visit frequency by outlet category (A/B/C outlets). Geographic whitespace identification — GIS mapping of covered vs uncovered outlets. Outlet classification — ML clustering of outlets by revenue potential, channel type, urban/rural. Productivity analytics — calls made, orders hit rate, strike rate per salesman. Ideal call calculation — which outlets deserve weekly vs fortnightly visits. Companies like Nielsen, IRI sell RTM analytics services; tech platforms like Bizom, FieldAssist embed this into SFA.