In Türkiye, cash on delivery is not a habit — it is an expectation. Buyers want to see the goods before handing over a card, and no e-commerce operation can afford to refuse. Nobody argues with the demand. What goes unexamined, and costs real money, is when the payment is taken.
In the classic flow the courier arrives, rings, waits, the recipient comes down, the card terminal comes out, the transaction runs, a receipt is printed, a signature is collected. Two minutes if it goes well. Eight if it does not — the recipient is upstairs, is hunting for cash, the terminal will not connect, the card limit is short. And this repeats dozens of times a day.
What the number looks like
Take a courier making 90 deliveries a day, with 40% of them cash on delivery — 36 stops. Assume each one loses an average of 90 extra seconds to payment.
36 stops × 90 seconds = 54 minutes a day. Roughly a seventh of the shift is spent standing at a door waiting for a payment.
The cost is not only labour. Those 54 minutes reduce how many deliveries fit into the day, push the last drops into evening traffic, and generate "the courier was late" complaints. One operational friction bills you from three separate directions.
Why a faster terminal is not the answer
The first instinct is to speed up the device: newer terminal, better mobile connectivity, shorter flow. That buys something, but in the wrong place — most of the lost time happens before the transaction, not inside it. The recipient coming to the door, finding a card, deciding which card, counting cash. Even if the terminal were instantaneous, those minutes remain.
The right question is not "how do we take payment faster?" It is "why is payment being taken at the door at all?"
Moving payment ahead of delivery
In the logistics configuration of our operations platform, the flow works like this: 30 to 60 minutes before the courier reaches that stop, a payment link or QR goes to the recipient over WhatsApp. The message reads like a delivery notification rather than a collection notice — "your parcel is on its way, you can pay now if you prefer, or at the door."
If the recipient pays, two things change at once:
- Dwell time at the door collapses. The courier hands over the parcel, takes the proof photo, leaves. There is no payment step.
- The delivery can complete even if nobody is home. This second one turned out to be worth more than the first.
The real win: eliminating repeat attempts
With an unpaid parcel, if the recipient is out there is nothing to be done. The parcel goes back and ships again tomorrow. A second attempt very nearly doubles the cost of the first — same route, same fuel, same courier minutes, plus warehouse handling.
If payment was taken beforehand, that parcel is simply a paid parcel. The recipient can say "leave it at the door", "leave it with my neighbour" or "leave it with the building manager" — and because the instruction sits in the WhatsApp thread, responsibility is not ambiguous. The delivery closes on the first attempt.
Worth getting right: a "leave it at the door" instruction must be written and clearly attributable. We keep it as the recipient's own WhatsApp message and attach the proof-of-delivery photo to the same thread. If a dispute surfaces later, there is nothing left to argue about.
Why WhatsApp rather than SMS
This choice is about conversion, not aesthetics. Links in SMS do not get tapped — and users are right not to trust them, because SMS is the main channel for exactly this kind of fraud. On WhatsApp the message arrives from a business account, the conversation history is visible, and the recipient can see who is writing.
The second reason is that WhatsApp is two-way. The recipient can reply "I won't be home today, can we do tomorrow?". In the same thread you can correct directions, share a location pin, and complete the building and door number. Over SMS every one of those becomes a call-centre interaction.
Three things to solve before you ship this
- Timing. Too early and the recipient forgets; too late and the courier is already at the door. A 30–60 minute window worked well for us, but it varies by city, route and courier type — it should be derived from the arrival estimate rather than fixed.
- Cancellations and refunds. If a delivery fails after payment, the refund flow has to be automatic. A refund process run by hand spends the trust you just earned within the first week.
- Reconciliation. Pre-paid and door-paid orders must land in the same report. If finance is collecting from two places, you give back in accounting what you gained in operations.
The one metric to watch
The number to track is not "prepayment rate". The right metric is first-attempt delivery rate. If prepayment climbs while first-attempt delivery stays flat, you have changed the payment channel and not the operation.
Second, watch average dwell time per stop, split between COD and non-COD stops. The gap closing is the signal that this is actually working.
What would this be worth in your operation?
Tell us your daily parcel volume, your cash-on-delivery share and your average dwell time, and we will work it through with you. Starting with one pilot region is the fastest way to see the result without changing the whole operation.