How to implement a loyalty program: from approved design to live scans

Everything downstream — scan verification, fraud detection, secondary-sales visibility, anti-counterfeit checks — depends on each unit carrying a unique, non-guessable code. Three application routes, in rising order of integration effort:

• Pre-printed serialised labels, applied manually or by applicator on the packing line. Fastest to start (2–3 weeks), costs ₹0.15–0.40 per code applied, and lets you begin with just your top SKUs. The weakness is application discipline — labels applied crooked, doubled or skipped need line-level QC sampling.
• Inline digital printing — variable-data QR merged into the existing label, pouch or carton artwork at the print vendor. Cheapest at national scale (₹0.10–0.20 per code) and tamper-consistent, but requires artwork revision, printer coordination and 3–6 weeks of lead time per SKU family. Cylinder-printed flexible packaging needs a hybrid: static artwork gravure-printed, variable QR inkjetted in a reserved white window.
• Concealed codes — under scratch panels, inside lids, on the inner flange of a wire coil, behind the label. Mandatory in high-fraud categories, because a visible outer code can be photographed and claimed by anyone in the supply chain before the genuine buyer ever opens the pack. Many brands run a dual-code design: outer code for authentication and track-and-trace, inner code for the reward.

Practical sequencing: serialise the two or three fastest-moving SKUs first — they generate 60–70% of scans and let the program go live while slower SKUs follow. Plan for the un-coded pipeline: stock already in the channel takes 4–8 weeks to sell through, so the field script must explain why some packs earn and some don't during the transition. Cost the whole exercise honestly with the QR cost estimator — code generation, printing, application, and wastage of 2–4% on line rejects.

Every serialised SKU needs a point value, and the mapping is strategy expressed as arithmetic. Work from dealer price and strategic role:

• Traffic / commodity SKUs (the 1L economy paint pack, the standard house-wire coil): 0.5–1% of unit value. On a ₹1,050 coil that is ₹5–10 — enough to make scanning habitual, cheap enough to sustain at volume.
• Premium / focus SKUs: 1.5–3% or a 2–3x multiplier. A ₹4,200 premium FR-LSH coil at ₹60–100 per scan tells the electrician exactly which product the brand wants specified.
• Launch SKUs: boosted rates for 90 days with an automatic sunset date configured upfront, so the boost dies without a second decision.

Keep the mapping as a version-controlled master with effective-from dates, owned by trade marketing but changeable without a development ticket. In copper-linked or price-revised categories, unit prices move quarterly; a points table frozen in code becomes wrong within one cycle. Worked check before freezing: expected monthly scans per SKU × point value, summed across the catalogue, must land within the always-on share of budget — typically 55–60% of a total spend held at 1–2% of secondary revenue (thin-margin categories) or 2.5–4% (high-margin). If the model breaches, cut commodity rates before premium ones.

KYC is where programs quietly lose a third of their audience. The stack that works for trade users:

• Mobile OTP as identity — the phone number is the account. One number, one account, enforced.
• PAN capture with real-time validation. Non-negotiable, because Section 194R aggregation is per PAN: once cumulative benefits cross ₹20,000 in a financial year, 10% TDS applies, and without PAN the brand cannot track the threshold or file correctly. Validate the PAN-to-name match at entry, not at payout time when a mismatch blocks money and creates a helpdesk ticket. Explain it in one honest vernacular line — "sarkari tax niyam ke liye" — because "why does the company want my PAN" is the single most common enrolment objection in every category.
• Progressive KYC. Let users register with mobile alone and start scanning; hold redemptions above ₹500–1,000 until PAN and bank/UPI verification complete. This converts KYC from an entry barrier into a motivated task — the user finishes it because money is waiting.
• UPI verification by penny-drop and name-matching against the PAN, which doubles as the first fraud control against mule handles.

Go-live is not the end of implementation; it is the start of hypercare — four to six weeks of deliberately over-resourced operation. The rhythm: a daily 30-minute stand-up (trade marketing, platform success manager, finance, pilot RSMs) reviewing five numbers — enrolments, first-scan activation rate, payout success rate, tickets by type, fraud flags. A 24-hour SLA on any blocked payout. Weekly structured field feedback: what are counters actually saying, which objection is trending. A fortnightly steering review empowered to change point values, KYC steps or comms without a committee cycle.

Watch fraud patterns from the first week, not the first month: dealer bulk-scanning shows up as one device claiming 50 codes in an evening from a single location; code harvesting shows up as scans of units that primary data says are still in a warehouse. Act on the first confirmed case visibly but proportionately — a warning and clawback communicated through the sales team teaches the channel the platform is watching, without turning launch month into a policing exercise.

Exit hypercare on evidence, not on the calendar: payout success stable above 98%, ticket volume flattened, activation trending up week-on-week, and the fraud-flag rate below 2–3% of scans. Then hand over to the steady-state cadence described in our guide to loyalty program KPIs — and start planning the scale-up from pilot districts using the pilot-to-national decision gates.