The verdict in three sentences
Fake orders — stolen cards and phantom cash-on-delivery — eat margin faster than any discount. Basic velocity rules (same IP, same phone, too many attempts) block 60 to 75 % of fraud with no data scientist. The real trade-off: too loose and you bleed chargebacks, too strict and you lose ~2 % of sales to false positives.
Three layers of defence
You can moderate by hand, set automatic rules, or plug in an aggregator's scoring. Hand moderation can't hold volume; rules cover 80 % of cases at near-zero cost; external scoring adds a layer for big baskets.
| Approach | Coverage | Cost | Suits |
|---|---|---|---|
| Manual blocking | Low, not scalable | Human time | Very low volume |
| Auto velocity rules | 60-75 % | Near zero | Most stores |
| Aggregator scoring | +10-15 % extra | Per-request fee | Big baskets, cards |
A velocity rule is a cap: no more than 3 payment attempts per card/hour, one OTP-verified phone before shipping, blocking an IP that fires 5 orders in 10 minutes.
The real cost of fraud
The 2026 figures below are order-of-magnitude estimates for Southern African e-commerce. A chargeback isn't just the amount: it adds fixed fees and hurts your rate with the acquirer.
| Indicator | 2026 order of magnitude |
|---|---|
| Card fraud (share of revenue) | 0.8-1.5 % |
| Average chargeback | ZAR 400-1,200 |
| Fixed fee per chargeback | ZAR 100-200 |
| Fake COD (unpaid) | 8-15 % of orders |
| Reduction with phone check + cap | -60 to -75 % |
| False-positive cost | ~2 % of lost sales |
| Phone OTP verification | a few seconds |
Across 1,000 orders, a 12 % fake-COD rate means 120 parcels packed and shipped for nothing: packaging, transport and return, often ZAR 100-250 lost per phantom parcel.
Need a professional website?
Kolonell builds websites that attract clients, optimized for the Sénégalese market. Free quote in 2 minutes.
Mini case study
Thabo sells tech accessories in Johannesburg, 800 orders/month, average basket ZAR 650. His fake-COD rate hits 12 % (96 lost parcels/month) and he takes 6 card chargebacks at ZAR 800 + ZAR 150 in fees.
He enforces phone OTP verification before any COD and a cap of 3 card attempts/hour per device. Fake COD drops to 4 % (32 parcels), meaning 64 parcels saved × ZAR 150 of avoided logistics cost. Chargebacks fall from 6 to 2. Cost: ~2 % false positives, i.e. 16 real orders lost, easily offset by the losses avoided.
FAQ
What is a velocity rule in practice? It's an automatic cap: attempts per card/hour, orders per IP, or one verified phone before shipping. It blocks the burst behaviour typical of fraud.
Does OTP verification scare off honest customers? Very little: OTP takes seconds and reassures. The false-positive cost is around 2 % of sales, to be weighed against 8-15 % of fake COD avoided.
Do you need an aggregator to score cards? Not at first. In-house rules cover 60-75 % of fraud; external scoring adds 10-15 % coverage, useful mainly on big card baskets.
How do you cut fake COD without killing sales? Verify the phone by OTP before packing and offer a small mobile money deposit. This filters phantom orders while keeping COD accessible.
What does a chargeback really cost? The disputed amount (ZAR 400-1,200) plus fixed fees (ZAR 100-200), plus a worse acceptance rate with your acquirer if the ratio climbs.
Let's talk about your project. We implement your velocity rules, OTP verification and an anti-fraud dashboard calibrated to your false-positive rate. WhatsApp +221 77 596 93 33.
Mohamed Bah
Fondateur, Kolonell
Passionate about digital and entrepreneurship in Africa, Mohamed has been helping Sénégalese businesses with their digital transformation since 2020. Founder of Kolonell, he believes every SME deserves a professional and accessible online présence.
