COD ecommerce in Algeria

How does a COD order move from click to cash in Algeria?

An order arrives from your store, from Instagram DMs, or from a form that writes a row into a Google Sheet. It carries a name, a phone number starting 05, 06 or 07, a wilaya, a commune, and a product. Nothing has been paid. You message or call the customer to check that the order is real and the address is usable. Once they agree, you create a parcel with a carrier: Yalidine, ZR Express, Maystro or Noest. The carrier collects it from your address, or you drop it at their agency. A driver delivers it to the customer's door, or the customer collects it from a stopdesk agency, and hands over cash. The carrier holds that cash and pays you on its remittance cycle, typically weekly, minus the delivery fee and any return fees you have accumulated. Between the click and the payout there are five points where the order can die, and each one is worth naming rather than lumping into a single vague "conversion problem."

Why does confirmation exist as a separate step?

Because nothing was paid, an Algerian COD order is a statement of interest, not a commitment. A customer can order twice by accident, mistype a digit in their phone number, give a commune the carrier does not serve, or simply change their mind before the driver arrives. Confirmation is where you turn interest into an agreement: you restate the product, the total price including delivery, and the address, and you get a yes in writing. Everything skipped at this step comes back later as a return. A parcel that ships unconfirmed still costs the outbound fee, the return fee, and the days that stock spent sitting in a van instead of being available to sell to a customer who genuinely wanted it. Stores that started requiring a written yes before shipping typically see their return rate drop within the first month, simply because the population of shipped orders changes: fewer of them were ever accidental or half-hearted to begin with.

What does a COD order actually cost you?

Yalidine, ZR Express, Maystro and Noest all price by destination wilaya and by whether the parcel goes to a stopdesk agency or to a door. Southern wilayas such as Tamanrasset, Illizi and Adrar cost more and take longer than Alger, Blida or Oran, since fewer trucks run those routes and each one carries fewer parcels per kilometer covered. Read your own carrier's fee table before you set a delivered price, because the outbound fee is charged whether or not the customer takes the parcel, and a return on a southern wilaya order costs meaningfully more than a return on a parcel that never left Alger.

Where does a COD store lose money?

In four places, roughly in order of size for a typical store. Leads that ask a question on WhatsApp and never get an answer, because you were asleep or dealing with someone else at the moment they wrote in. Orders that sit unconfirmed for two days and go cold, because nobody followed up before the customer's intent faded or they ordered the same thing somewhere else. Confirmed parcels that fail on the first attempt because the phone is switched off or the address is a landmark rather than a street. And returns you never diagnose, so the same commune or the same product keeps failing month after month without anyone noticing the pattern. Each of these is countable if you write the numbers down: leads answered, orders confirmed, parcels shipped, parcels delivered, cash collected. Most stores track only the last number and treat everything above it as noise. That is exactly why the leak stays where it is: nobody is looking at the stage where it actually happens, so the fix gets aimed at the wrong part of the funnel.

What should you fix first?

Fix the order of operations before anything else. Nothing ships until you hold a written yes from the customer covering the product, the total with delivery, and the full address including wilaya and commune. Then set a deadline: an order with no reply after 48 hours gets one follow-up message, and after that it is closed rather than shipped on a hope that it will sort itself out later. Then, once a week, split your returns into two piles. Refused at the door is one pile. Never reached is the other. Refused is a confirmation and pricing problem, usually meaning the customer was surprised by something at delivery time. Never reached is a phone and address problem, usually meaning the driver could not physically execute the delivery. Take the bigger pile, change one thing about how that field is collected or that message is worded, and count again the following week. Two weeks of that tells you more about your specific store than any general advice about cash on delivery, because it is built from your own orders rather than an average across a market as varied as this one.