Fraud Prevention

How to Prevent Fake COD Orders in Pakistan — Complete 2026 Guide

By OrderWitness Team July 22, 2026 8 min read

If you sell on cash-on-delivery in Pakistan, fake orders are not a possibility — they are a certainty. Every seller who ships COD will encounter customers who place orders with no intention of paying, or who simply deny placing the order when the courier arrives.

The question is not whether it will happen to you. The question is how much it costs you per month, and what you can do about it.

25–40%
of COD orders are fake or refused in Pakistan
PKR 1,200
average loss per returned COD order
3–7 days
inventory locked in transit per fake order

A seller shipping 500 orders per month with a 30% fake rate loses PKR 180,000 every single month to fake orders — before accounting for inventory damage on returns. This guide covers every method available to reduce that number.

The Problem is Worse for Daraz Sellers

If you are a Daraz seller, fake orders hit you harder than direct sellers. Daraz's buyer-friendly return policy makes it trivially easy for customers to refuse delivery with no explanation required. Your Daraz seller account metrics suffer — late dispatch rates, return rates, and performance scores — even when the fault lies entirely with a customer who had no intention of paying.

Daraz sellers running their own COD operations outside the platform face the same issue compounded: no buyer protections, no dispute resolution, and no evidence if a customer denies placing the order. This is exactly the problem OrderWitness solves — giving every Daraz seller and independent COD seller a timestamped, cryptographically-signed record of every order confirmation before dispatch.

Daraz seller tip: Use OrderWitness for orders taken directly (Facebook, Instagram, WhatsApp) alongside your Daraz store. Your Daraz orders have Daraz's infrastructure — your direct orders have none. That gap is where most losses happen.

Why Fake COD Orders Are So Common in Pakistan

Understanding the root cause helps you pick the right solution. Fake COD orders in Pakistan happen for several distinct reasons:

Method 1: Customer Confirmation Before Dispatch (Most Effective)

The single most effective method for eliminating fake orders is getting the customer to actively confirm the order before you ship it. A customer who placed an order on a whim will often not bother to confirm — and a customer who planned to refuse at the door cannot later claim they never ordered once they have already clicked YES.

Confirmation works on two levels:

  1. Filtering: Fake orders self-cancel when customers ignore or reject the confirmation.
  2. Evidence: If a customer confirms but later denies at the door, you have timestamped proof of their confirmation.

WhatsApp Confirmation

Send a WhatsApp message to the customer's number immediately after the order is placed. The message should contain the order details and two buttons: Confirm Order and Cancel Order. When a customer taps a button, the response is recorded with a timestamp.

Using the official WhatsApp Business Cloud API (via Meta) is strongly recommended over unofficial WhatsApp automation tools. Unofficial tools risk your number being banned and do not generate legally acceptable evidence.

Email Confirmation

Send a branded email with a confirmation link. Less effective than WhatsApp in Pakistan (many customers do not use email), but useful as a backup channel for customers who do not respond on WhatsApp.

Key insight: In our data, sellers using WhatsApp confirmation see 60–75% of orders confirmed within 2 hours. Orders with no response after 48 hours can be safely cancelled — saving the shipping cost entirely.

Method 2: AI Fraud Risk Scoring

Not every fake order is obvious. AI fraud scoring analyses every order automatically when it is placed and assigns a risk score from 0 to 100 based on multiple signals:

Risk SignalWhat It DetectsWeight
Repeat rejection historyPhone number has refused past orders from any sellerHigh
Order velocitySame number placed multiple orders in 24 hoursHigh
Address anomalyAddress is incomplete, a known commercial building, or previously flaggedMedium
Value outlierOrder value is unusually high for a first order from this numberMedium
Blacklisted numberNumber appears on a known fraud listCritical
Area rejection rateHistorical refusal rate is high for this delivery areaLow

Orders scoring above your configured threshold (typically 70+) are automatically placed on hold for manual review before dispatch. This prevents your highest-risk orders from ever leaving the warehouse without human approval.

Method 3: Phone Number Blacklisting

Every time a customer refuses a delivery, their phone number should be added to a blacklist. Future orders from that number are automatically flagged or rejected.

An effective blacklist has two tiers:

Important: Do not blacklist customers who cancelled before dispatch — only those who refused at the door or were unreachable. Cancellations before shipping cost you nothing and should not be penalised.

Method 4: Prepaid Conversion for High-Risk Orders

For orders with a fraud risk score between 50 and 70 — too risky to ship confidently but not clearly fake — offer the customer a prepaid option. Send a message explaining that prepaid orders get priority dispatch and faster delivery.

Legitimate customers will often agree to prepaid, especially if you sweeten the offer with a small discount. Fraudulent customers will not pay upfront. This single filter can reduce losses on borderline orders by 40–60%.

Method 5: Address Verification

Require customers to confirm their delivery address as part of the confirmation step. An incorrect or incomplete address is a strong fraud signal. This is especially important for orders to industrial areas, commercial zones, or apartment buildings where individual unit numbers are often missing.

Some platforms offer Pakistan Post address validation or Google Maps pin-drop verification during checkout. Even asking for a Google Maps pin in the confirmation message filters out many fake orders.

What Does NOT Work

Experience across thousands of sellers shows that some commonly tried methods are largely ineffective:

Combining Methods: A Practical Framework

The most effective approach combines all methods in a tiered system:

  1. Order placed → Instant fraud score calculated (0–100)
  2. Score 0–30 (Low risk) → Send WhatsApp confirmation, ship on response
  3. Score 31–60 (Medium risk) → Send WhatsApp confirmation + offer prepaid discount
  4. Score 61–80 (High risk) → Hold for manual review, call if valuable order
  5. Score 81–100 (Critical) → Auto-cancel, log to blacklist, do not ship

Sellers using this framework report reducing fake order rates from 30–40% down to 8–12% within the first month.

The Cost of Doing Nothing

Let us put a number on inaction. If you ship 200 orders per month at an average COD value of PKR 2,500, and your current fake order rate is 30%:

A proper confirmation and fraud scoring system costs a fraction of that — and eliminates the majority of losses within the first billing cycle.

Start Preventing Fake Orders Today

OrderWitness combines WhatsApp confirmation, AI fraud scoring, phone blacklisting, and cryptographic evidence PDFs in a single platform. Free plan includes 50 orders per month — no credit card required.

Try OrderWitness Free →