Cash on delivery still accounts for somewhere between 40 and 57 percent of e-commerce transactions in Egypt depending on the category, and most founders treat that number as a payments preference to accommodate rather than a cost centre to manage. It's both, and the cost side is bigger than the payment-processing fee it's usually compared against.
The actual damage isn't the cash — it's the return-to-origin rate that comes with it. Industry estimates for Egypt put COD return-to-origin anywhere from the high teens to over a third of orders, depending on category and how the merchant runs checkout. Every one of those orders was picked, packed, and sent on a courier round trip that the business paid for twice, on a sale that never happened.
This is a trust problem wearing a logistics costume, not a delivery problem. A customer who orders on impulse at a price nobody said out loud, with no deposit and no verification step, hasn't actually committed to anything — cash on delivery, structurally, lets a customer place an order the same way they'd add something to a wish list, and the business only discovers the difference at the doorstep.
Full prepayment fixes the return rate and breaks the sale for a large share of the market that genuinely doesn't trust entering card details on a site it's never bought from before — so the fix isn't eliminating COD, which would trade one problem for a worse one, it's adding friction at exactly the point that filters intent without filtering the sale itself.
The interventions that actually move the number are specific: phone verification before the order is confirmed, a WhatsApp message that gets a real reply rather than a silent delivery, and a small deposit — even a token amount — that turns an impulse click into a decision the customer has to actively make twice. Each of these adds seconds to the process and each one has been shown to meaningfully cut return-to-origin, because they all do the same thing: they ask the customer to confirm before the business commits real cost.
The businesses treating this well don't chase a single fix — they layer two or three of these checks based on order value, tightening verification as the ticket size rises rather than applying the same friction to every order regardless of what's actually at risk. A five-dollar item can absorb a bad return. A hundred-dollar one shouldn't ship without someone actually confirming it's wanted.