Fake COD Orders: How Many Are Fake, What Each Costs, and When Verification Pays

About 8% of COD orders are fake or invalid in Profitbox360's client data. Each one that ships costs ₹210. The standard advice is "add OTP verification." The real math sits close to breakeven — and the deciding variable is not your fake rate. It is how many genuine buyers your verification scares away.

The numbers here come from 50+ Indian D2C brands and roughly 156,000 shipped orders analysed by Profitbox360 between January and July 2026 — COD-heavy brands at ₹1,200–1,800 AOV.

What share of your COD orders is fake — and how do you spot yours?#

About 8% across Profitbox360's client base. That is 440 orders a month at 10,000 orders with 55% COD. Brands under attack see far worse — one brand measured 23%. Your own number takes an afternoon: count last month's COD orders that failed phone verification or went to unreachable buyers.

Fake COD comes in three patterns. Malicious bulk orders: someone burning your shipping budget, clustered in time, pincode or phone pattern. Casual zero-intent orders: a real person who ordered on a whim and never meant to accept. Invalid orders: wrong numbers, unreachable phones, garbage addresses.

The share varies so much between brands because of targeting. Impulse-priced products advertised to cold audiences collect zero-intent orders at several times the base rate. If your RTO on ad orders runs high, expect your fake share above 8% too. The same type of buyer causes both.

What does one fake order cost if it ships?#

₹210. The money goes into two shipping legs, COD handling, packaging that comes back damaged or not at all, and warehouse time on a parcel that earned nothing.

Ship all 440 monthly fakes and the bill is 440 × ₹210 = ₹92,400 a month. That is the ceiling on what any detection system is allowed to cost.

₹210 sits below the ₹275 full-RTO cost in the RTO cost stack. A fake caught at the doorstep skips part of the return leg, and unreachable-buyer parcels come back faster. It is the cheapest kind of RTO. That is what makes it dangerous: each loss is small enough to ignore, so brands ignore it. ₹11 lakh a year.

What does verification cost — and what does it catch?#

Automated verification (OTP or one-tap confirm) costs about ₹2.50 per COD order and catches roughly 75% of fakes. A manual confirmation call costs about ₹8 per order. And every verification layer loses about 3% of genuine COD buyers to friction. That last number decides everything.

The three numbers that run the whole decision, at 5,500 COD orders a month:

Look at rows two and three. The fraud saving and the friction cost are nearly the same number. The "just add OTP" advice never mentions this.

Does verification pay? The per-10,000-orders math#

At an 8% fake rate with 3% friction: no. It loses ₹12,760 a month — ₹69,300 saved, minus ₹13,750 spend, minus ₹68,310 friction. Cut friction to 1% and the same system makes ₹32,780 a month. Friction decides, not fraud.

The rule: verification pays when fake share × catch rate × ₹210 beats verification spend + friction loss × margin. At 3% friction, the breakeven fake share is about 10%. Most healthy brands sit below it. At 1% friction, breakeven drops to about 4.5%. Almost everyone clears it.

So reverse the usual question. Not "should I add OTP?" but "how cheap can I make confirmation?" Then add it. One-tap confirmation beats typed OTP. Verifying only risk-flagged orders beats verifying everyone: spend falls, friction falls, and the catch rate barely moves, because most fakes sit in the risky group anyway. Manual calls at ₹8 do not clear this math (₹44,000 a month on all COD orders). Use them only on high-value flagged orders.

A brand under bulk attack at 23% fake is a different case: 1,265 fakes, ₹1,99,290 recoverable. Verification pays there under any friction assumption. Verify first, optimise later.

RTO still high after OTP? The next four levers#

Because OTP only fixes fakes, and fakes are about 10% of the RTO problem. In Profitbox360's cause data, plain refusal is 45% and address failure is 30%. A brand at 50% RTO with OTP running is not doing verification wrong. It has run out of things verification can fix.

The next levers, in order of cost:

1. Address quality. Address failure causes 30% of RTO. Pincode validation, completeness checks and a delivery-instructions field attack it at near-zero friction. The biggest fix after OTP, and the one no tool vendor talks about.
2. Repeat-refuser history. A buyer who refused their last COD parcel is your most predictable future RTO. Flag past refusals at order time and route those buyers to prepaid-only. The data is already in your order history.
3. Selective COD. Stop offering COD to everyone. Cut it on high-RTO SKU types, flagged pincodes, and orders above your value cap — the cap bands are here.
4. Partial COD. The heaviest lever. It attacks the 45% refusal share directly by collecting money upfront. That math is here.

Verification is the first step of this ladder, not the whole ladder. Brands that stop at OTP have fixed a tenth of the problem.

FAQ#

Is WhatsApp confirmation better than SMS OTP? Yes, on the variable that decides the math: friction. One-tap confirm/cancel loses fewer genuine buyers than a typed code, at similar cost per order.

How do I block repeat COD refusers? Record every RTO against the buyer's phone number. At order time, route flagged numbers to prepaid-only or advance-required. Most brands have the data and have never joined it.

Can I detect fake orders before verification? Partially. Time-clustered orders, repeated phone patterns, mismatched pincode-address pairs and unreachable numbers are all pre-ship signals. Use them to decide who gets verified, not whether to verify.

Should I just turn off COD instead? At ₹1,500 AOV, no. COD carries too much genuine demand. Restrict it selectively before removing it. The gentler tool for shifting the mix is the prepaid discount.