Every online store in Bangladesh eventually faces the same operational question: how does a confirmed order become a parcel, and how does the delivery result come back? The answer can be a person copying addresses into a merchant panel or a system that routes each order to the right courier, keeps the status in sync and reconciles the cash. The difference is not only convenience. It changes the error rate, the speed of month-end reconciliation and who owns the customer experience when a parcel is late.
The options below are operating models, not brands. Most stores end up with a combination, but one model usually dominates. Compare them on the work you will do every day, not on the demo.
| Approach | Fits | Setup | Per-order work | Status quality | COD reconciliation | Main weakness |
|---|---|---|---|---|---|---|
| Manual panel entry | One courier, low volume | Low | High | Manual | Manual | Copy errors, no history |
| CSV batch | Regular dispatch batches | Low | Medium | Delayed | Spreadsheet | Stale files, duplicates |
| Direct API | One courier, growing volume | Medium | Low | Good with webhooks | Built once | Courier concentration |
| Aggregator | Several couriers, coverage gaps | Medium | Low | Provider-dependent | Usually included | Intermediary dependency |
| Platform app | Standard store and flow | Low | Low | Varies | Varies | Limited exception handling |
| Custom layer | Complex rules, many channels | High | Low | Designed | Designed | Build and maintain |
Manual merchant panel entry
Staff open the courier panel, type the recipient, address, weight and cash-on-delivery amount, then copy the tracking number back to the order. It works at very low volume and teaches the team how the courier behaves. It fails when two people enter the same order, when addresses are long, or when the panel changes without notice. Keep a daily check of parcels created against orders marked ready to ship.
CSV batches and scheduled imports
Orders are exported, cleaned in a spreadsheet and uploaded to the courier; statuses come back as a file. This suits a store with predictable cut-off times and no development capacity. The weaknesses are stale status, duplicate rows and silent rejection of individual rows. Validate row counts before and after upload, and keep the file as evidence for the reconciliation.
Direct API connector to one courier
A developer connects the store to one courier's API. Order creation, label printing, cancellation and tracking become part of the order screen. This is the simplest real integration and often the best first step. The risk is concentration: if coverage or service fails in a district, the store has no fallback. Keep a manual export path for the day the API is unavailable.
Multi-courier aggregator
An aggregator provides one API and a panel across several couriers. Coverage improves and the store can route by destination, weight or price. The trade-off is another dependency: the aggregator's status mapping, settlement cycle and support become part of your operation. Ask which couriers are direct integrations and which are manual behind the scenes.
Store platform app
Marketplace apps install quickly and cover standard flows. They are a reasonable start for a small store on a mainstream platform. Read the limits before relying on one: which couriers, which statuses, whether it handles returns, whether it exports data and what happens to historical orders if you uninstall it.
Custom orchestration layer
A custom layer owns the order-to-parcel workflow and treats couriers as pluggable providers. It is justified when delivery rules are a differentiator, when several channels share stock, or when reconciliation is still manual. It is not justified for a store with one courier and a small daily volume. Budget for monitoring, maintenance and a named owner after launch.
Criteria that decide the choice
- Coverage: which districts and remote areas each courier actually serves.
- API maturity: sandbox, webhooks, idempotency and stable identifiers.
- COD remittance: collection, fees, adjustments and the date cash reaches your account.
- Returns: reverse pickup, inspection, restocking and the cost of a failed delivery.
- Address quality: validation, landmarks and the handling of Bangla addresses.
- Exception ownership: who calls the customer when a parcel is stuck.
- Data portability: can you export every parcel, status and settlement line?
Trade-offs that are easy to miss
Status names differ between couriers, so "delivered" may mean different things. Webhooks can arrive late, out of order or not at all, which makes a polling fallback necessary. Cash-on-delivery reconciliation is harder than order creation because fees, partial returns and remittance dates rarely line up. A multi-courier setup doubles the exception paths even as it improves coverage.
Platform apps are easy to install and hard to leave. Custom builds solve the unusual case but need a product owner after the project ends. Address quality drives failed deliveries at least as much as courier choice, so a clean address form is part of the integration, not a separate concern.
Choose each approach when
- Manual: one courier, a handful of orders a day, and no development capacity.
- CSV: predictable batches and a team that already works from spreadsheets.
- Direct API: one courier covers most destinations and volume is growing.
- Aggregator: coverage gaps, seasonal peaks or a need to compare courier performance.
- Platform app: a standard store and a standard delivery workflow.
- Custom: delivery rules or reconciliation are a competitive advantage.
A decision path
Count orders, couriers and exceptions for the last month. List the manual steps and the errors they create. Decide whether delivery is a differentiator or a utility. Require a sandbox and a pilot with real addresses. Agree who owns the exception queue and write the exit terms before signing. A proposal that cannot answer those questions is a brochure, not an integration plan.
Red flags in a proposal
Unlimited courier coverage without named APIs, no sandbox, no cash-on-delivery reconciliation, no exception owner, per-order fees with no ceiling and no data export. These gaps usually appear after go-live, when the store is already dependent on the service.
We build order, payment and delivery workflows under Commerce Systems. The pricing page explains how integration scope is separated from platform work. For the budget view of the same decision, read courier integration cost for online stores in Bangladesh.
Sources
- Shopify, FulfillmentOrder API reference, accessed 17 November 2026.
- WooCommerce, REST API documentation: Orders, accessed 17 November 2026.
- Meta, WhatsApp Cloud API documentation, accessed 17 November 2026.
- Bangladesh Bank, Mobile Financial Services statistics, accessed 17 November 2026.
Responses (0)
No comments yet. Add the first one below.
Leave a response
Your email address will not be published. Required fields are marked *.