Selling through Instagram messages works when you have a few orders a day. Then it eats all your time, because every order is typed by hand again and again.
The fix is one store link that captures the order once, confirms it, and sends it straight to the delivery company. That is what I built WovenDZ to do.
This post shows where a DM seller's hours really go, why returned parcels decide who survives, and the build order I followed. If you sell online in Algeria, the lessons apply even if you never use WovenDZ.
Why do so many Algerian sellers sell through Instagram?
Because that is where the buyers are, and there is no big marketplace doing it for them. Jumia, the pan-African online marketplace, announced on 10 February 2026 that it was leaving Algeria (Ecofin, February 2026).
So sellers do it themselves, in direct messages. A customer comments "prix?", the seller replies, they exchange many messages about size, colour and wilaya. Somewhere in that thread is a phone number the seller copies into a notebook.
Then the seller calls a delivery company, reads the address out loud, and writes a tracking number next to the name. It works at a few orders a day. It does not scale.
Where does a DM seller's time really go?
Not to selling. To copying.
Count the places the same customer details get written down by hand:
- In the message thread.
- In the notebook.
- In the delivery company's dashboard.
- On the parcel label.
- In whatever list the owner uses to check if the week made money.
That is five chances to swap two digits of a phone number. And with cash on delivery, a wrong phone number is not a small typo. It is a parcel that ships, fails, and comes back at your cost.
It is like a school where the teacher copies every student's name by hand into five different registers every morning. The teaching never starts.
What is RTO, and why does it decide if a store survives?
RTO means return to origin: a parcel that goes out, is refused or can't be delivered, and comes back unpaid.
In a market where people pay by card first, a failed delivery is an annoyance. With cash on delivery, the seller has already paid to send the parcel, often pays again for it to come back, and earns nothing.
Do this calculation with your own courier's price list.
Say your profit on a product is small, and one failed parcel costs you two delivery fees. Then one return can wipe out the profit of one or more good orders.
That is why cutting returns is often worth more than buying more traffic.
What drives returns is mostly order quality. An order from a serious buyer who confirmed the address usually arrives. A one word comment at midnight often does not.
How is a real store different from selling in DMs?
| Selling in Instagram DMs | A store link with tracking | |
|---|---|---|
| How the order is taken | Free text spread over many messages | Fields: product, size, name, phone, wilaya, commune, address |
| Phone number | Retyped by hand several times | Typed once, by the buyer |
| Confirmation | The seller calls, sometimes | A confirmation before the parcel is created |
| Handing over to the courier | Retyped into the courier's dashboard | Parcel created automatically through the courier's API (how two apps talk) |
| Tracking | A number written next to a name | Attached to the order, visible to the buyer |
| Your limit | How fast you can type | Your stock |
| What you see each week | A notebook and a feeling | Sales, returns and delivery cost per order |
What do Algerian sellers need that foreign platforms ignore?
1. Cash on delivery as the default, not an exception. It is how Algerian buyers pay, and for good reasons. I explain them in why Algerians pay cash on delivery.
2. Direct links to local couriers. Several Algerian couriers can be connected by code. The open source CourierDZ library creates parcels with Yalidine, ZR Express (Procolis), Maystro and Ecotrack-based couriers, and tracks them with most (CourierDZ). Most sellers do not know this is possible.
3. A phone first, Arabic first design. The seller manages the shop from a mid range Android, often on a weak connection. Right to left Arabic is the main layout, not a translation added at the end.
4. Trust in the platform itself. Sellers have seen platforms take a cut and then disappear. If the seller feels the platform stands between them and their customer, they go back to DMs.
How does WovenDZ handle this?
I built WovenDZ around those four needs. Here is how it works today, from the WovenDZ site in October 2026:
| Feature | How it works |
|---|---|
| Commission | 0% on sales. The seller keeps 100% |
| Price | Monthly plans at 1,500, 2,500 or 4,000 DA, with a 14 day free trial and no automatic renewal |
| Store link | Each seller gets a link with their shop name, to share in the Instagram bio |
| Payment | Cash on delivery by default, online payment with Chargily as an option |
| Delivery | Through Yalidine, ZR Express, Maystro or Noest |
I am the person who built it, so read this table as a description, not a neutral review. The build lessons below apply to any store.
In what order should you build an Algerian online store?
This is the order I would follow again, and close to the one I did follow.
- Make the order structured before making it pretty. The first feature is a checkout that captures every field from the buyer's own thumbs. I explain why short forms convert better in the paradox of choice at checkout.
- Confirm before you ship. Create the parcel only after the buyer confirms. It turns a guaranteed loss into no cost.
- Connect one courier properly before adding a second. Pick the one with the best documented API.
- Give the seller their own link, not a listing lost inside a marketplace.
- Make stock safe at checkout. Reduce stock in the same step that saves the order, so two buyers can't both get the last item.
- Show the numbers that decide the business: return rate, delivery cost per order, money left after courier fees.
- Build for a cheap phone first, in Arabic. If it is slow on a mid range Android, it does not work.
What mistakes do new online sellers make?
- Buying ads before fixing order quality. More traffic into a leaky process means more returns.
- Shipping without confirming. Every unconfirmed order is a parcel that may come back unpaid.
- Never measuring the return rate. Most sellers have never seen it as a number. Seeing it changes behaviour.
- Leaving Instagram entirely. Keep it for attention. Just end conversations with a link.
- Ignoring the legal side. Selling goods online needs a commercial register and a .com.dz site under Law 18-05, as I explain in the cash on delivery guide.
What should you do this week?
- Today: count last month's orders, delivered parcels and returned parcels. Work out your return rate.
- This week: write down every place you copy customer details by hand.
- This week: start confirming every order by call or WhatsApp before shipping.
- This month: move from DMs to a store link with a short order form.
- Then: connect your courier so parcels are created without retyping.
Building your own store from scratch instead? Read what a first store or app really costs in Algeria before asking for quotes.