COD rules cart value category pincode

COD rules by cart value, category and pincode let you keep cash on delivery where it earns money and restrict it where it does not, instead of switching it off for everyone. Each rule reads facts about an order and returns one of a few outcomes: allow, confirm first, take an advance, or require prepayment. Good rules are simple, disclosed and measured.

This guide builds a rule set step by step. It gives a cost model for cart-value bands, a scoring sheet for categories, a method for pincode bands, a matrix that combines them, ten sample orders run through it, message templates, a rule register and the metrics that tell you whether each rule is working. All rupee figures are assumptions for illustration; replace them with your own contribution, freight and return data.

Quick answer: Set COD rules by cart value, category and pincode in a fixed order: category hard rules first, then the pincode risk band, then the cart value band, then a small adjustment for customer history. Give every rule one of four outcomes (allow, confirm, advance, prepaid only), show restrictions before payment, and review each rule monthly against RTO rate, checkout completion and net contribution.

What a COD rule is, and what it should return

A COD rule is a condition and an outcome. The conditions use information you already hold at checkout: the cart value, the product category, the destination pincode and what you know about the buyer. The outcomes should be few and clearly different, or your support team will not be able to explain them to buyers.

Keep to these four. Every extra outcome multiplies the rules you have to maintain. If your tools cannot enforce all four, start with allow, confirm and prepaid only, and handle the others by hand. Our post on whether to keep, limit or drop COD explains how to decide the overall stance; this guide assumes you have decided to limit. The companion posts on RTO cost calculation and on pincode-level risk supply the break-even rate and the zones that feed the rules below.

Also decide the default outcome for orders that match no rule, such as a pincode with no history; "confirm" is a sensible default for new territory.

Where a rule can actually be enforced

A rule that exists only in a document is a policy; a rule that the checkout applies is a control. Before designing anything, find out at which layer each rule can be enforced.

To see what a configurable checkout can look like, take Shiprocket's checkout product. Its help page, which showed a modification date of 27 July 2026 when read on 14 October 2026, says the checkout supports custom rules for prepaid discount, COD charge, shipping charge, disabling COD, shipping visibility and partial COD. The criteria it lists include product SKU, product tags, product collection, destination pincodes, weight range, invoice value, lane, risk profile and phone number. That vocabulary is a useful checklist for any platform: it shows the conditions the industry treats as normal inputs. This describes one vendor's product and is not a recommendation.

ShopMate provides COD collection tracking, a pincode serviceability check and shipping partners with COD. Where a rule cannot be applied automatically, a manual confirmation step is a workable stand-in, provided you log how often it runs and how long it takes.

Design principles

Keep to six to eight rules, each traced to dated evidence, explainable in one sentence, disclosed before payment, reversible with an owner and a review date, and changed one at a time.

Consumer rules that touch COD restrictions

Restricting COD is a commercial choice, but how you present it is regulated. The Central Consumer Protection Authority lists the Consumer Protection Act and the Consumer Protection (E-Commerce) Rules, 2020 on its Act and Rules page (read on 16 October 2026). The practical lesson for COD rules is to state payment modes and any COD charge before the buyer commits, and not to penalise a buyer through hidden terms.

The Central Consumer Protection Authority's dark-pattern guidelines are the second source. The Press Information Bureau reports that the guidelines specify 13 dark patterns, including drip pricing. The Gazette notification of 30 November 2023, hosted on the CCPA guidelines page, is available as the full guidelines PDF. Its Annexure 1 defines drip pricing to include price elements that are not revealed upfront or are revealed surreptitiously within the user experience, and charging an amount higher than the amount disclosed at checkout. It defines interface interference as design that highlights some information and obscures other relevant information to misdirect a user, and false urgency as falsely stating or implying urgency or scarcity to push an immediate purchase. The Annexure says its illustrations are guidance, not an interpretation of law.

For COD rules, this means: show the COD charge in the price breakdown before the final click, do not hide the prepaid alternative, and do not invent countdowns such as "only 3 minutes to keep COD".

Start with the data you need

Every threshold in this guide should come from your own order export. Before you design a rule, prepare a sheet with one row per shipped order and these columns.

Keep one row per shipped order with: order date and delivery outcome (delivered, RTO, cancelled before dispatch), payment mode, cart value, category, destination pincode and zone, buyer ID with earlier-order count, freight billed forward and on return, COD fee billed, and any refusal reason the courier records. Never mix COD and prepaid in one rate. Use a window long enough that most orders have a final outcome. Delhivery's help page on non-delivery reports says a seller can opt for reattempting delivery three times for a forward shipment, after which the shipment moves to RTO, so an order shipped yesterday may not yet have its final status. Exclude the most recent two to three weeks from any rate you compute.

Dimension 1: cart value

Cart value matters at both ends. At the low end, flat costs such as freight, packaging, COD handling and advertising eat the margin, so a small basket cannot survive many returns. At the high end, a returned parcel loses more in absolute terms, may be damaged, and may carry fraud or dispute risk. In between is the range where COD tends to work.

The illustration below uses assumed numbers: cost of goods of 55 per cent of selling price, forward and return freight of ₹70 each, packaging ₹15, handling on a return ₹10, COD fee ₹25, advertising ₹100 per order and stock loss of 8 per cent of cost of goods on returns. Delivered contribution is C = price - cost of goods - 70 - 15 - 25 - 100; loss on a return is L = 70 + 70 + 15 + 10 + stock loss + 100; and the break-even RTO rate is C / (C + L), the break-even idea applied to contribution margin.

Check each row. The fixed costs total 70 + 15 + 25 + 100 = 210. At ₹500: C = 500 - 275 - 210 = 15; stock loss = 0.08 x 275 = 22; L = 70 + 70 + 15 + 10 + 22 + 100 = 287; 15 / (15 + 287) = 5.0 per cent. At ₹1,000: C = 1,000 - 550 - 210 = 240; stock loss 44; L = 309; 240 / 549 = 43.7 per cent. At ₹2,000: C = 2,000 - 1,100 - 210 = 690; stock loss 88; L = 353; 690 / 1,043 = 66.2 per cent.

Under these assumptions, delivered contribution is zero when 0.45 x price = 210, that is at about ₹467. A cart below that loses money even when delivered. So a COD floor around ₹500 is a minimum, and even there the tolerable RTO rate is only 5 per cent. Your own floor depends on your margin and fixed costs: compute it with your numbers.

The top of the range: why a high break-even rate is not the whole story

Now take a ₹5,000 cart with the same assumptions, but with a COD fee that rises with value. Shiprocket's help page on COD charges says that for COD shipments a fixed COD charge or COD percentage of the order value, whichever is higher, is used; its example, which it labels as illustrative, has a fixed charge of ₹36 and a rate of 2.5 per cent on a ₹750 order. Delhivery's rate card page likewise says a COD shipment attracts an additional cost for cash handling. Applying the Shiprocket-style formula as an assumption:

The crossover is 36 / 0.025 = ₹1,440. For the ₹5,000 cart: cost of goods 2,750; COD fee 125; C = 5,000 - 2,750 - 70 - 15 - 125 - 100 = 1,940. If the returned goods go back on the shelf with the usual 8 per cent stock loss (220), L = 70 + 70 + 15 + 10 + 220 + 100 = 485 and the break-even RTO rate is 1,940 / 2,425 = 80.0 per cent, a very tolerant figure. But if the return is damaged or unsellable (the full 2,750 lost), L = 70 + 70 + 15 + 10 + 2,750 + 100 = 3,015 and the break-even falls to 1,940 / 4,955 = 39.2 per cent. One unsellable return wipes out the contribution of about 1.55 delivered orders (3,015 / 1,940).

That is why the top band is governed by confirmation or an advance, not by the break-even rate alone: the average case looks safe, but the bad case is large. Set the top band where one bad return would hurt your month, and confirm identity and intent before you ship.

Dimension 2: product category

Category rules protect you from product types where a refused parcel cannot be resold or where refusal is common for reasons the product cannot fix. Classify your catalogue once, by the traits below, and attach the rule to the class, not to individual products.

A scoring sheet for borderline products

Some products fit more than one class. Score each trait from 0 to 2 (resale after refusal, damage in transit, value relative to your average order, fit uncertainty, freight burden) and add them up: 0 to 3 allow, 4 to 6 confirm, 7 to 10 prepaid only or advance. A 2 on resale after refusal alone should trigger prepaid only, because every return is then a full loss.

Mixed carts

Decide explicitly how mixed carts behave. The simple option is to apply the strictest outcome among the items to the whole order. The alternative is to split the order, which adds shipments and freight. Document the choice, or two team members will answer differently, and tell the buyer why when a single item changes the payment options for the whole cart.

Dimension 3: pincode risk band

Use bands derived from your own zone analysis rather than individual pincodes. Our companion post on pincode-level risk explains the method in full; the short version follows.

Serviceability comes first

Courier coverage is a separate check from risk. Delhivery's help page says a non-serviceable pincode produces a message that the pin code is not serviceable, and your pincode check should apply that before any COD rule. A pincode can be serviceable for prepaid and for COD or only for one of them, so test the payment mode too.

Choose a grouping that has enough orders

The six-digit PIN has structure: the first digit indicates the postal zone, the second the sub-zone, and the first three together the sorting district. Grouping by the first three digits gives districts with more orders in each group than single pincodes. Couriers use their own zones too: Delhivery's rate card page describes zones A to F, from local within-city delivery and metro-to-metro to special zones such as Jammu, Himachal Pradesh and the North East, and Kashmir, Manipur, Ladakh and the Andaman and Nicobar Islands. Choose one grouping and stay with it.

Turn a rate into a band

A band needs a rate and the confidence you have in it. Illustration with assumed numbers: a zone shows 12 returns on 40 shipped COD orders, a rate of 30 per cent. A 95 per cent Wilson score interval for 12 of 40 runs from about 18.1 to 45.4 per cent, so the true rate could plausibly be anywhere in that range. Compare this with a zone at 3 returns on 60 orders (5 per cent), whose interval runs from about 1.7 to 13.7 per cent.

Under that definition the 5 per cent zone, with an upper end of 13.7 per cent, is green. The 30 per cent zone, with an interval of 18.1 to 45.4 per cent, straddles 25 per cent, so it is amber, not red: you do not yet know that it is bad enough to ban COD. A zone would be red only if even the lower end sat above 25 per cent. The 30-order minimum and the 25 per cent break-even rate are assumptions; use your own.

Separate courier delays from buyer behaviour

If a zone is amber because of courier delays, a different courier may fix it more cheaply than a rule; see our guide to delivery to all pincodes for partner coverage.

Dimension 4: customer history

History is a modifier, not a primary dimension. A buyer with accepted past parcels deserves a looser outcome and a buyer with a past refusal a stricter one. Keep the definitions short and countable.

Steps run in the order prepaid only, advance, confirm, allow. Decide beforehand whether history may loosen a red pincode band or a category hard rule. We recommend it may not: a good buyer in a red zone should still be offered prepaid, and a made-to-order item is unsellable whoever refuses it.

Customer data is personal data. India's Digital Personal Data Protection Act, 2023 applies to it. Under Rule 3 of the Digital Personal Data Protection Rules, 2025 (Gazette CG-DL-E-14112025-267650 of 14 November 2025, read on 16 October 2026), a notice must be understandable on its own, in clear and plain language, with an itemised description of the personal data and the specified purpose of processing; Rule 1(4) brings Rule 3 into force eighteen months after publication, which is 14 May 2027. Keep the history you hold to what the rule needs, and keep your privacy notice accurate. Couriers sell risk scores too: Delhivery's page on its RTO Predictor says it produces a score between 0.0 and 1.0 from consignee details, product, shipment value and payment method, shows it as high, medium or low risk, and charges ₹2.49 plus GST for low and high risk COD orders (as read on 14 October 2026; the page was last updated on 3 April 2025). Whether a paid score adds value beyond your own history is something to test, not assume.

Setting COD rules by cart value, category and pincode: the order of evaluation

Evaluate rules in a fixed order so the result is predictable.

1. Serviceability. If the pincode is not serviceable for COD, stop; offer prepaid if it is serviceable for that.
2. Category hard rules. If any item is in a prepaid-only class, the result is prepaid only.
3. Matrix lookup. Take the outcome for the pincode band and the cart value band.
4. History modifier. Move one step looser or stricter, but never loosen a red band or a category hard rule.
5. Category floor. If the category class suggests confirm or advance, take the stricter of that and the result so far.
6. Default. If nothing matched, apply the default outcome.

The matrix below is an illustrative starting point built from the assumptions above; the cart bands follow the break-even analysis, and the pincode bands follow the zone method.

The matrix is not a recommendation for your store: its values come from invented margins, so recompute the cart bands first. Write the matrix, the modifiers and the order of evaluation on a single page, with the date of the data behind each threshold.

Worked example: ten orders through the rules

The orders below are invented. Each is evaluated in the order above.

Orders 5 and 8 show why you should settle the exceptions in advance. If the team has to argue about them at the time, the buyer waits and the outcome varies by person. Order 10 shows the category floor: when the category already calls for confirmation, a repeat buyer does not escape it.

Template: a one-page rule register

Keep one register for all rules. Each row should be complete enough that a new team member can explain and, if needed, switch off the rule.

Explaining the rule to the buyer

A restriction without an explanation feels arbitrary. Give a short, neutral message where the buyer first meets it: on the pincode check, in the cart and at the payment step. The templates below can be adapted; keep your own wording consistent with your return policy.

• P1, prepaid only: "Cash on delivery is not available for this order. Pay online by UPI, card or net banking to continue."
• C2, confirm: "Cash on delivery is available. We will message you to confirm your order before dispatch. Please reply within 24 hours."
• A3, advance: "Pay a small advance now and the balance on delivery. The advance is shown in your order total."

Avoid wording that blames or pressures. Do not add false countdowns, and do not hide a fee until the last click. For prepaid, UPI is the usual option; Wikipedia summarises UPI, and the Press Information Bureau's release of 15 September 2026 states that merchant payments up to ₹2,000 remain free of merchant discount rate with a nominal rate above that. Your gateway's own fees are separate, so read them in your gateway dashboard before you promote a payment method. Our guide to Razorpay setup, at the Razorpay setup walkthrough, covers the payment side.

Designing the confirmation step

"Confirm" is the outcome most often designed badly. A good confirmation step has a deadline, a clear action and an owner.

A workable schedule, with assumed timings: send the confirmation message with an order summary and a one-word reply at hour 0 and hold the order; send a reminder or call at hour 6; at hour 24 send a final reminder with a prepaid link, then cancel with a notice and release the stock. Set the timings so that the delay costs you less than the returns you avoid, and tell the buyer the deadline in the first message. Capture replies as evidence. Do not treat silence as consent: if you want to cancel after a deadline, say so in the message and in your terms. If you rely on a messaging channel such as WhatsApp for confirmations, message only buyers who have opted in and follow the channel's own policy.

The courier can add a second check at the door. Delhivery's NDR page describes a manual retrigger of IVR verification for consignee-refused shipments that were marked cancelled without code or OTP validation, and its SmartNDR page says the service can resolve non-delivery reports through WhatsApp and AI calls, and that it does not automatically return shipments except when the consignee confirms cancellation through an OTP-verified response. These are courier-side services on Delhivery One, priced separately. Shiprocket's help page on what an NDR is lists reattempt, contacting the buyer and starting an RTO as the seller's options once an undelivered report arrives.

Advance payments and how couriers see them

An advance combines prepaid and COD in one order, which complicates reconciliation. Delhivery's help centre explains that, for Shopify orders, its platform maps a partially paid financial status to the COD payment mode, so the courier collects the balance; other channels differ. Shiprocket's checkout page, cited above, lists a partial COD setting where the seller sets the percentage allowed against chosen criteria, with an option to include shipping charges or not. Confirm how your store passes the balance due to the courier, since the COD amount must equal the unpaid balance. You can collect the advance with a payment link: Razorpay documents payment links that can be shared with customers.

Sizing the advance

One way to size an advance is to cover the loss on an ordinary return: for the ₹5,000 cart above, L = ₹485, about 10 per cent of the cart. It does not cover an unsellable return (₹3,015), so for fragile or made-to-order goods choose prepaid only instead.

Decide the refund rule for an advance when the buyer refuses the parcel: is it kept, partly kept or refunded? State it before the buyer pays, in line with your return policy and consumer law; the Department of Consumer Affairs publishes the Consumer Protection Act and Rules through its website.

Illustrative scenario: an apparel seller in Indore

The figures below are invented. An apparel store ships 600 COD orders a month with an average cart of ₹1,000, so the earlier ₹1,000 economics apply: C = ₹240, L = ₹309. Its overall COD RTO is 30 per cent. Zone analysis shows that 200 of the 600 orders fall into amber or red zones, or into the under-₹500 band, with a combined RTO of 45 per cent.

Before the rules, that 200-order segment produces 90 returns and 110 deliveries: 110 x 240 - 90 x 309 = 26,400 - 27,810 = minus ₹1,410. The segment loses money.

The store introduces prepaid-only for the segment. Suppose half the buyers leave and half pay online, and the prepaid orders return at 6 per cent. Then 100 orders remain: 6 returns and 94 deliveries, giving 94 x 240 - 6 x 309 = 22,560 - 1,854 = ₹20,706. The change is 20,706 - (-1,410) = ₹22,116 a month, even though half the segment's orders disappeared. The result depends on retention: each kept order is worth 0.94 x 240 - 0.06 x 309 = ₹207.06, so keeping only 60 orders still gives about ₹12,424, and keeping 140 gives about ₹28,988.

The store should still watch the COD orders it left untouched for spill-over, since buyers who meet a restriction may abandon other carts too.

Illustrative scenario: a gifting seller in Jaipur

Again with invented numbers: a seller of personalised photo frames sells at ₹1,200 with cost of goods of ₹600, forward and return freight of ₹70 each, packaging ₹15, advertising ₹100 and a COD fee of ₹25. Delivered contribution is C = 1,200 - 600 - 70 - 15 - 25 - 100 = ₹390. A refused frame cannot be resold, so a return loses L = 70 + 70 + 15 + 10 + 600 + 100 = ₹865, and the break-even RTO rate is 390 / (390 + 865) = 31.1 per cent.

The seller currently receives 100 COD orders a month with 22 returns, so net contribution is 78 x 390 - 22 x 865 = 30,420 - 19,030 = ₹11,390. Under a prepaid-only rule, suppose 70 buyers pay online and 3 parcels fail for reasons such as wrong addresses. With prepaid contribution of about ₹391 (a 2 per cent gateway fee of ₹24 replaces the ₹25 COD fee), the net is 67 x 391 - 3 x 865 = 26,197 - 2,595 = ₹23,602, a gain of ₹12,212 a month.

The made-to-order trait makes every refusal a full loss, so say on each product page that personalised items are prepaid only.

Monitoring metrics for each rule

A worked net-effect calculation for one rule

Use the ₹1,000 economics again (C = ₹240, L = ₹309) with invented results for a confirm rule that applies to 180 COD orders a month. COD RTO in that segment falls from 25 per cent to 20 per cent, so 180 x 0.05 = 9 fewer returns. Treat each avoided return as becoming a delivery: the gain is C + L = 240 + 309 = ₹549, so 9 x 549 = ₹4,941.

Now the costs. Suppose 4 per cent of the 180 buyers, 7.2 orders, abandon instead of confirming. Each lost order had an expected value of 0.75 x 240 - 0.25 x 309 = 180 - 77.25 = ₹102.75, so the loss is 7.2 x 102.75 = ₹739.80. If confirming costs ₹8 of staff time per order, that is 180 x 8 = ₹1,440. Net effect: 4,941 - 739.80 - 1,440 = ₹2,761.20 a month. The numbers are assumptions; the method is what to reuse. If your confirmation costs ₹30 an order, the same rule costs 5,400 and loses money, so the rule must change or be automated.

Testing a rule before you roll it out

Treat each rule as an experiment. Use an A/B test where traffic allows: apply the rule to half of qualifying orders and leave the rest as a control for the same period. For small stores, a time-based approach can work if nothing else changes, but a seasonal event can distort it. Judge by net contribution per 100 checkouts after returns have settled, not by the prepaid share or the number of orders alone. Wikipedia's article on sample size determination explains why small differences need many orders to detect. If a rule affects only 30 orders a month, you will not be able to measure its effect in a month; accept a wider review period or use the break-even logic instead of a test.

Cash flow effects of the rules

Keep festival and sale periods separate in your analysis, and decide in advance whether rules relax during them. Define whether bands use the cart value before or after coupons, and use the same definition everywhere.

Every order moved from COD to prepaid arrives in your bank on the gateway's settlement schedule instead of the courier's COD remittance. Delhivery's help page says it transfers the COD amount within 48 hours under conditions, including a positive wallet balance, while Shiprocket's support page describes a D+8 working-day processing time, so the effect depends on your partner. Our COD remittance guide shows how to calculate the working-capital impact. Track days to cash by rule outcome, since a rule that pushes buyers to prepaid improves cash even if the order count is flat.

Common mistakes and how to avoid them

Checklist: launching COD rules

• Compute C, L and break-even RTO rate for each cart band and product class.
• Classify the catalogue into category classes and decide mixed-cart handling.
• Assign pincode bands from zone analysis with confidence ranges and a minimum order count.
• Write the matrix, modifiers, default outcome and order of evaluation on one page.
• Draft buyer-facing messages for each outcome and show them before payment.
• Decide advance refund terms and test the payment-link or advance process on one real order before launch.
• Set up the metrics table and a control group.
• Set a review date and a stop-loss for every rule (for example, pause a rule if net contribution per 100 checkouts is below control for two consecutive weeks).

Checklist: team readiness

• Support agents can explain each outcome in one sentence.
• Someone owns confirmation messages every working day, including weekends if you ship then.
• Staff know whether they may override a rule, and every override is logged with a reason.
• Held orders appear in one list, with the deadline for each.
• Your privacy notice mentions that order history is used to decide payment options.

Frequently asked questions

What cart value should I set as the minimum for COD?

Calculate it, because it depends on your margin and fixed costs per order. Subtract cost of goods, freight, packaging, COD fee and advertising from the cart value to see delivered contribution, then check the break-even RTO rate. In our assumed example the contribution turned positive only above about ₹467, and the tolerable return rate at ₹500 was just 5 per cent.

Can I block COD for certain product categories?

Yes, and it is common for custom, personalised or perishable items, where a refused parcel cannot be resold. Classify the catalogue once, apply the rule to the class and say so on the product page. For mixed carts, decide whether the strictest item governs the whole order and document that choice for your team.

Is it fair to restrict COD by pincode?

It can be, if it rests on data and is disclosed. Use zones with enough orders, show a clear message at the pincode check and offer a prepaid alternative. Avoid blanket state-level blocks without evidence.

Should a COD fee be charged to the buyer?

That is a commercial choice. Couriers charge sellers a COD handling cost, as Delhivery's rate card page and Shiprocket's help page describe. If you pass it on, show it in the price breakdown before the final payment click, because a charge revealed late can raise drip-pricing concerns. Compare the fee with how much the average buyer values the convenience.

How do I handle COD advance payments for the courier?

The courier should collect only the unpaid balance. Check how your store passes this to your shipping tool; Delhivery's help page says that for Shopify orders a partially paid status maps to COD. Reconcile the advance in your payment gateway and the balance in the courier's remittance, and state the refund rule before the buyer pays.

How often should I review COD rules?

Review pincode bands and metrics monthly, and cart bands whenever freight, margins or ad costs change materially. Use a rolling window of about 90 days with a cut-off old enough for final outcomes, and change a rule only when the evidence clearly crosses your threshold. Keep a log of changes and the reasons for each one.

What if my store cannot enforce rules automatically?

Use a manual confirmation step as a stand-in: hold qualifying COD orders, message or call the buyer and dispatch only after confirmation. Track how many orders it covers, how long it takes and what it costs per order.

How many COD orders make a rule reliable?

There is no fixed number, but small counts mislead: 12 returns on 40 orders gave a plausible rate from about 18 to 45 per cent. As a working minimum, many sellers wait for roughly 30 shipped COD orders per band and treat anything below as undetermined.

Set COD rules by cart value, category and pincode with ShopMate

COD rules by cart value, category and pincode work best when orders, payments and shipping sit in one system. ShopMate provides COD collection tracking, a pincode serviceability check, 40+ shipping partners and analytics you can export to CSV for the metrics above. Book a demo, read about us or see the pricing page. For wider selling logistics, see our all-India selling playbook.

Disclaimer: This article is for general information only and is not legal, tax or financial advice. ShopMate is the publisher of this guide and has a commercial interest in the platform it describes. All rupee figures, thresholds and order examples are assumptions or inventions for illustration, not benchmarks. Courier and vendor terms are described as read on 14 October 2026 and change often. Rules, thresholds, fees and platform terms change, and your situation may differ. Consult a qualified chartered accountant or legal professional before making decisions.