On most lanes the last mile is a cost line. On the China–Bangladesh parcel lane it is a commercial variable: payment is collected at the doorstep, most orders are paid in cash on delivery, and a failed attempt sends the goods back with freight both ways unpaid. This guide covers the delivery end of the lane: how cash-on-delivery works, what it does to your cash cycle, how Bangladeshi courier networks differ, why address data decides more than carrier choice, and what to fix before the first parcel leaves Shenzhen.
In a card-first market, a delivery failure is an inconvenience: the customer is charged, the parcel is re-attempted, and the cost lands as a small operational drag. In a cash-on-delivery market the same failure is a revenue event. If the consignee is not home, refuses the parcel or cannot be reached, no money changes hands, the goods travel back through the network, and you absorb outbound freight, return handling and the risk that returned stock never sells again at full price.
That asymmetry is why the delivery promise belongs in product design rather than in shipping settings. Bangladeshi shoppers compare a cross-border seller against local merchants delivering inside a day or two, and share delivery experiences widely — one bad attempt costs the repeat purchase too. Sellers who treat the last mile as procurement optimise the per-parcel rate and lose on delivered orders, which is the number that matters.
COD is the mainstream payment method in Bangladeshi e-commerce, driven by limited card penetration and a simple trust logic: see the goods, then pay. Wallet and digital payment adoption is growing, but assume COD dominates your early order mix unless your own data says otherwise.
The consequence is structural: your courier is also your payment collector, which makes it a financial counterparty. Three policies follow. Cap COD order value for first-time customers rather than accepting unlimited exposure on an unverified address. Offer a small prepaid incentive, because moving even a minority of volume to prepaid improves both cash timing and refusal rates. And require a confirmation contact for high-value or first-time orders, since fake orders are a real cost in COD markets and a short call removes most of them before dispatch.
The most common modelling error on this lane is booking a COD order as revenue at dispatch. It is not. Money leaves your account first — goods, origin handling, freight, duty and VAT at clearance — and arrives later, after the courier delivers, collects and remits. Remittance timing varies by partner and agreement, and is commonly settled on a periodic cycle rather than per parcel; confirm the actual schedule with your provider rather than planning against an assumption.
| Outcome | What happens | Cost you carry | Cash timing |
|---|---|---|---|
| Delivered, collected, first attempt | Consignee present, pays rider, parcel handed over | Planned cost per delivered order | Collected at delivery, remitted on the agreed cycle |
| Delivered after re-attempt | First attempt fails, rider returns on another day | Extra attempt or re-delivery handling; longer promise | Collected later; remittance cycle shifts accordingly |
| Refused at the door (RTO) | Consignee declines; parcel enters the return flow | Outbound freight plus return leg, handling, restock or write-off | No cash in; the original outlay stays unrecovered |
| Undeliverable / bad address | Address or contact unusable; parcel held then returned | Holding storage, return leg, total loss of the sale | No cash in; highest true cost per order |
| Lost in the network | Parcel not traceable after a scan gap | Goods plus freight; claim depends on declared value and policy | Claim settlement, not customer payment |
A materially higher return-to-origin rate than you would accept in a prepaid market does not just add return freight; it converts inventory into non-selling stock, distorts your demand forecast and inflates the real cost per delivered order. Track four numbers from day one: delivery success rate, RTO rate, average days from dispatch to cash, and cost per delivered order rather than per shipped order.
Carrier choice on this lane is not a choice between equivalent providers. Bangladesh has several structurally different networks, and the right answer is usually a mix rather than a single vendor.
| Network type | Coverage shape | Strength | Limitation | Suits |
|---|---|---|---|---|
| Nationwide courier / parcel brands | Broad district coverage through owned or franchised hubs | Reach beyond the big cities; established COD handling | Slower in dense Dhaka than city-focused rivals; service quality varies by franchise | National coverage, district-town and rural orders |
| E-commerce-focused delivery networks | Dense in Dhaka and major cities, thinner at the tails | Fast metro delivery, good tracking, merchant-friendly integration | Coverage drops outside the urban core; peak capacity can tighten | Dhaka and Chattogram volume, marketplace-style promises |
| Marketplace-owned logistics | Built around a single platform's order flow | Tight integration with that marketplace's checkout and SLA | Limited or no use for orders sold elsewhere | Sellers whose Bangladesh volume runs through one large marketplace |
| Postal and traditional cargo networks | Widest nominal reach, including remote upazilas | Universal service obligation; useful for the long tail | Longer transit, weaker tracking granularity, slower COD remittance | Low-urgency, remote-destination parcels |
The practical approach is to map your own order distribution and assign by destination: a metro-optimised network for the city share, a national network for district towns, and a fallback for remote orders you would otherwise decline. Ask each candidate for thana- or area-level serviceability lists and their published attempt and remittance policy in writing before you compare rates.
Bangladeshi addressing is descriptive rather than postal. A usable address runs division, district, thana or upazila, then area or village, road, house or holding, plus a landmark. Postcodes exist but consumers rarely supply them, and street-level addressing is inconsistent outside planned districts. The functional key is the phone number, not the address string.
Design checkout around that reality. Collect district and thana or upazila as structured fields rather than free text, require a landmark line, ask for an alternate contact number, and offer a delivery window or an alternative recipient such as a workplace. Keep the number prominent on the label, because the rider will call before attempting in most cases. Better data on the same network reliably produces a higher success rate than switching carriers, and it costs nothing beyond form design.
National delivery promises hide enormous variation. Plan by city tier, and quote the tail rather than the metro when you publish a promise.
| Destination tier | Service reality | What to plan for |
|---|---|---|
| Dhaka metro | Dense network coverage, fastest delivery, most carrier competition | Traffic-driven variability; cut-off discipline decides next-day promises |
| Chattogram | Strong second market with its own catchment and port-city economy | Slightly longer metro delivery; useful test market before national rollout |
| Other divisional cities | Good coverage from national networks, Dhaka-first hubs add a leg | An extra day or two in transit; fewer re-attempt options |
| District towns | Coverage depends on the specific network's hub footprint | Confirm serviceability per district; longer holding periods before RTO |
| Rural and upazila destinations | Tail of the network, often through franchise partners | Slower delivery, slower COD collection, higher refusal and RTO risk |
Seasonal peaks matter as much as geography. Order surges around Eid, Durga Puja, Pohela Boishakh and the winter wedding and shopping season tighten last-mile capacity at exactly the moment you are pushing volume; build the calendar into your dispatch plan and your published promise rather than discovering it in a queue.
Two operational settings quietly determine your RTO rate. The first is the attempt policy: how many delivery attempts a network makes, how long a parcel is held, and what triggers a return. A network that returns quickly protects its own throughput and inflates your losses; a network that holds longer gives the consignee a chance to collect but ties up stock. Get the policy in writing and match it to your product value.
The second is the cut-off. A published cut-off only helps if it holds during peaks and if the warehouse on your side actually meets it. Once both are set, state the delivery promise in business days by destination tier, and make it visible at checkout — a promise the customer can see reduces both refusals and support tickets, while an optimistic one produces the refusals you were trying to avoid.
A Bangladesh-bound parcel is sorted, bagged, re-sorted and carried by bike or van, often in monsoon conditions. Packaging should be specified for that journey rather than for the warehouse shelf: right-sized cartons that survive stacking, water-resistant outer wrapping in the wet season, tamper-evident closure, and a label that prints the order identifier and the exact COD amount. A mismatch between the amount on the label and the amount quoted at checkout is one of the most avoidable causes of doorstep refusal.
Keep unit-level barcodes on every parcel so scan events exist at each handover, and avoid markings that advertise value. Fragile and liquid contents need inner protection, not a sticker.
Return-to-origin is mostly decided before the parcel moves. The levers, in the order they usually pay back:
MJDT Global (明杰供应链) treats the Bangladesh lane as one chain rather than a set of separate purchases: consolidation at origin in Shenzhen, the mode matched to the order profile, customs clearance prepared before departure, handover into the delivery network, and COD settlement tracked against a defined remittance cycle. Because the delivery leg is where the money is collected, we set the attempt policy, serviceability list and remittance schedule with the seller before volume starts, rather than discovering them in the first month's returns.
Bangladesh programmes usually run alongside our other South Asia and regional lanes; see our freight forwarding services and the global network for the full route map.
Courier coverage, remittance schedules and marketplace logistics in Bangladesh change frequently. Confirm serviceability, attempt policy and settlement terms with your partner before committing volume.
Send your monthly parcel volume, destination spread by city or district, average order value and whether COD or prepaid dominates. We'll come back within 24 hours with a delivery setup view — network split, attempt and remittance terms, expected cash cycle and the RTO controls worth putting in before your first dispatch.
Get a Bangladesh Delivery Quote →