Office Address
18/09, Block B, Bosila City R/A
Mohammadpur, Dhaka - 1207
Phone Number
+880 9638 667701
+880 1711 871722
18/09, Block B, Bosila City R/A
Mohammadpur, Dhaka - 1207
+880 9638 667701
+880 1711 871722
by Shah Nawaze Pavel
E-commerce inventory management in Bangladesh is the discipline of knowing exactly which products and variants are available, where they are stored, what has been committed to open orders, and when more stock must be purchased. A reliable process connects purchasing, receiving, sales, returns, courier outcomes, and financial reporting. It prevents a seller from accepting an order for an unavailable item while also reducing cash trapped in products that do not sell.
Inventory management is the controlled process of planning, buying, receiving, storing, selling, returning, and counting physical goods. In e-commerce, it also includes reserving stock for orders, synchronizing availability across sales channels, and releasing or adjusting that stock when an order is cancelled, delivered, returned, or damaged.
“Stock on hand” alone is not enough. An operations team needs to distinguish physical stock, sellable stock, reserved stock, damaged stock, and incoming stock. Suppose a shelf contains 25 units, eight are reserved for confirmed orders, two are damaged, and ten are due from a supplier. The business does not have 35 units ready to sell. Its currently available quantity is 15, while incoming quantity should be tracked separately until receiving and quality checks are complete.
This distinction matters particularly for Bangladesh businesses that take orders from a website, Facebook, phone calls, showrooms, and resellers at the same time. A number written in a notebook or an isolated spreadsheet may be correct at 10:00 a.m. and wrong after a busy hour of Messenger orders. A dependable inventory record must change whenever the real business event changes.
Inventory is both an operational resource and invested cash. Too little stock causes cancelled orders, delayed confirmations, disappointed customers, and wasted advertising spend. Too much stock consumes working capital, occupies warehouse space, increases handling, and exposes products to expiry, damage, season changes, or declining demand.
Accurate stock also affects teams beyond the warehouse. Customer service needs a truthful availability answer. Marketing should not promote a nearly exhausted variation without a plan. Finance needs purchase cost and ending inventory to calculate margins. Procurement needs demand and lead-time information. Management needs to know which items create revenue and which merely occupy shelves.
The goal is not to maximize inventory. It is to maintain enough sellable stock to fulfil expected demand at an acceptable service level while keeping investment and risk under control.
Every product should have a stable name, category, unit of measure, purchase cost, selling price, supplier reference, and tax treatment where relevant. Products with size, colour, capacity, flavour, pack size, or other choices need separate variants. Decide which fields are authoritative and restrict casual editing. Duplicate products with slightly different spellings create reporting errors and make stock counts unreliable.
A stock keeping unit, or SKU, is an internal identifier. It should remain stable even if the public product title changes. A simple format might combine category, product, colour, and size, such as TSH-CORE-BLK-M. The exact pattern matters less than uniqueness, consistency, and readability. Never let two sellable variants share a SKU.
Before goods arrive, record what was ordered, from which supplier, at what cost, for which location, and on what expected date. Expected stock helps procurement planning, but it should not normally become sellable stock. Supplier shortages, substitutions, and damaged units can make the received quantity different from the ordered quantity.
At receiving, compare the supplier document with the purchase order. Count units, inspect quality, identify the correct SKU, and record shortages or excess. Acceptable units move into sellable stock; rejected or damaged units move into a separate condition or return-to-supplier process. The receiving record should identify the user, time, location, quantity, and source document.
Use consistent warehouse, zone, rack, shelf, or bin references. Fast-selling products should be easy to pick without creating congestion. Similar-looking variants should not be mixed loosely. A system quantity is useful only when staff can find the same units physically.
Define the exact order status that reserves stock. Reserving immediately at checkout can protect against overselling, but unpaid or fake orders may hold stock unnecessarily. Reserving only after confirmation reduces false reservations but creates a window in which multiple customers can request the last unit. The right rule depends on order volume, payment mix, and verification speed. Whatever rule is selected must be applied consistently.
When an order is packed or dispatched, the system should convert its reservation into a completed stock-out according to the business’s accounting rule. A cancellation before dispatch releases the reservation. A failed delivery does not automatically mean the item is sellable again: the parcel must return, be inspected, and be received back into the correct condition. This event-based approach creates an audit trail instead of silently overwriting a quantity.
Cycle counts check selected SKUs regularly without closing the whole warehouse. Count high-value and fast-moving items more frequently than stable, low-risk products. Investigate differences before posting adjustments. Common causes include an incorrect SKU, unrecorded damage, duplicate dispatch, picking from the wrong location, a return put directly on the shelf, or unit-of-measure confusion.
A product is the customer-facing concept; a variant is the exact sellable choice. “Cotton Polo Shirt” may be one product, but navy-medium and navy-large are different inventory items. If one shared quantity is used for all combinations, the store can continue selling a size that is already exhausted.
Good SKU practice includes a documented naming rule, barcode support where practical, no reuse of retired identifiers, separate costs when variants have different costs, and a mapping between supplier codes and internal SKUs. Bundles need an additional rule: selling one bundle should reduce every component required for that bundle. A gift or promotional item also consumes inventory and should not disappear from reporting simply because the customer price is zero.
When importing products, validate duplicate SKUs before creating records. When renaming a product, keep the SKU unchanged unless there is a controlled migration. When discontinuing a variant, deactivate it for future sales while preserving historical orders and stock movements.
Each inventory movement should have a type and reference. Stock-in may come from a purchase receipt, customer return, warehouse transfer, production, or approved correction. Stock-out may come from a sale, supplier return, transfer, sample, internal consumption, expiry, loss, or approved correction. A bare “+5” or “−3” without a reason is not a useful audit record.
Use approval controls for manual adjustments, especially for high-value products. Record the quantity before and after, reason, user, timestamp, and supporting reference. Review adjustment totals by location and employee. Repeated adjustments may reveal a process problem even if the closing number eventually matches.
Available to sell = Sellable stock on hand − Reserved stock
If 120 units are physically sellable and 28 are reserved for confirmed orders, available-to-sell quantity is 92. Damaged or quarantined units should already be excluded from sellable stock.
Average daily demand = Units sold during the period ÷ Number of selling days
If 900 units were sold over 30 days, average daily demand is 30. Use a representative period and consider seasonality. Eid demand should not be forecast solely from a quiet month.
A simple operating estimate is: Safety stock = (Maximum daily demand × Maximum lead time) − (Average daily demand × Average lead time).
Suppose maximum daily demand is 45 units, maximum supplier lead time is 8 days, average daily demand is 30, and average lead time is 5 days. Safety stock is (45 × 8) − (30 × 5) = 210 units. This is a planning model, not a universal answer. Improve it with reliable demand and lead-time history.
Reorder point = Average daily demand × Average lead time + Safety stock
Using the figures above, reorder point is (30 × 5) + 210 = 360 units. When the inventory position reaches 360, procurement should review or place the replenishment order. Inventory position can include sellable and appropriate incoming stock minus commitments, depending on the chosen policy.
Inventory turnover = Cost of goods sold ÷ Average inventory at cost
Average inventory is commonly calculated as (opening inventory + closing inventory) ÷ 2. If annual cost of goods sold is BDT 6,000,000 and average inventory at cost is BDT 1,000,000, turnover is 6 times. Compare products with their own history and category characteristics; a high-turnover grocery item and a premium appliance should not share the same target blindly.
Sell-through rate = Units sold ÷ Units received × 100
If 400 units were received and 260 sold during the review period, sell-through is 65%. State the period and treatment of opening stock so the metric remains comparable.
Stock accuracy = Correctly counted SKU-location records ÷ Total records counted × 100
If 190 of 200 counted SKU-location records match the system, record accuracy is 95%. Quantity-weighted or value-weighted measures can supplement this, because a one-unit difference on an expensive item may matter more than a small variance on packaging material.
A stockout occurs when demand exists but sellable inventory is unavailable. Its causes include delayed purchasing, inaccurate records, unexpected demand, supplier delays, unprocessed returns, and stock held by abandoned orders. Use reorder alerts, lead-time tracking, safety stock for critical items, faster order verification, and alternative suppliers where justified. Review lost-sales and cancellation reasons rather than relying only on completed sales, because completed sales cannot show demand that the business was unable to serve.
Overstock is inventory beyond a reasonable demand and risk horizon. It often results from optimistic forecasts, quantity discounts, duplicated purchasing, weak variant analysis, or continuing to reorder a declining product. Review inventory age, weeks of cover, open purchase orders, and variant-level sell-through before buying. A product may sell well overall while one colour or size remains overstocked.
Dead stock has had no meaningful movement for a defined period and has little near-term demand. Define age bands appropriate to the category, such as 0–30, 31–60, 61–90, and more than 90 days. Then choose a documented action: stop reordering, improve merchandising, bundle with a relevant seller, transfer to a better location, negotiate a supplier return, offer a controlled markdown, or write off unusable goods. Do not hide dead stock by repeatedly changing its purchase date.
A returned parcel should enter a receiving workflow, not go directly from the courier bag to a sellable shelf. Match it to the original order and SKU, inspect seals and condition, record missing components, and classify the outcome. Common conditions include unopened and sellable, opened but sellable, needs repacking, damaged, expired, quarantine, and supplier-claim eligible.
For failed delivery, keep the unit unavailable while it remains with the courier. When the parcel physically returns, record the return date and condition. This prevents the same unit from being promised to another customer before it is back in the warehouse. Track return reasons separately: customer refusal, wrong address, courier damage, product defect, incorrect item, or failed delivery are operationally different problems.
Damaged inventory should retain its cost and reason for financial review even though it cannot be sold. Periodically reconcile the damaged-stock location, approve disposal or supplier return, and preserve evidence where required by company policy.
A seller may receive orders through a website, Facebook, a physical outlet, telephone agents, and wholesale partners. If each channel maintains a separate uncontrolled number, overselling becomes likely. Use one product master and one inventory source of truth. Channel orders should enter the central workflow quickly enough to reserve stock under the same rules.
For multiple warehouses, record stock at SKU-location level. The business needs both a consolidated view and the ability to see what each location can fulfil. Order allocation should consider availability, delivery destination, courier coverage, handling capacity, and transfer cost. Do not treat a unit in Chattogram as immediately available to a Dhaka warehouse without accounting for transfer time.
Transfers require two sides: stock-out from the source, in-transit quantity, and stock-in at the destination after receiving. A transfer should not increase total company stock. Differences between dispatched and received quantities must be investigated.
| Approach | Works well when | Main limitation | Control needed |
|---|---|---|---|
| Notebook/manual register | Very small, single-person operation | Slow updates, weak search and audit trail | Daily reconciliation |
| Excel or Google Sheets | Limited SKUs, one location, disciplined users | Concurrent edits and order synchronization become fragile | Protected fields, version history, one owner |
| Inventory software | Multiple users, channels, locations, or growing order volume | Requires setup, training, and clean master data | Roles, approvals, backups, routine audits |
Excel is not automatically wrong. It can be a sensible starting tool when transaction volume is low and one person controls the file. Use separate movement records instead of overwriting the closing balance, validate SKUs, protect formulas, and back up the workbook. The warning sign is not a particular order count; it is when updates arrive from several people or channels faster than they can be reconciled accurately.
Software should reduce duplicate entry, connect stock changes to business documents, enforce user permissions, show location-level balances, and provide traceable reports. Buying software without cleaning product data or defining status rules simply digitizes confusion.
Daily, review negative or unexpectedly zero stock, failed imports or orders, low-stock alerts, unprocessed returns, and large manual adjustments. Weekly, count selected fast-moving or valuable SKUs, review open purchase orders, inspect inventory ageing, and resolve location mismatches. Monthly, close the movement period, reconcile warehouse totals with finance, review turnover and dead stock, revise reorder parameters, and document improvement actions.
Useful automation includes reserving stock from confirmed orders, releasing reservations after cancellation, creating low-stock alerts, calculating reorder suggestions, synchronizing channel orders, recording purchase receipts, tracking transfers, and producing ageing or movement reports. Automation should have failure handling: an API error, duplicate event, or missing SKU must create a visible exception rather than silently losing a stock movement.
Start with clean data and the events that cause the largest errors. Measure stock accuracy, stockout frequency, adjustment value, fulfilment delay, and inventory age before and after the change. Automation is valuable when it improves control and decision quality, not merely when it replaces clicks.
DevzCart publicly presents connected product management, inventory management, order-dashboard, stock-alert, and reporting and analytics capabilities. This connected structure can help a business maintain product and variant records, monitor inventory, process orders, and review results without treating each function as an isolated spreadsheet.
For businesses using more than one selling route, DevzCart also describes multi-channel sales. Courier operations can connect through its verified courier integrations. Exact modules and package availability should be confirmed against the current plan and implementation before migration. A sensible rollout begins with SKU cleanup, opening-stock verification, roles, locations, and documented order-status rules.
This guide complements DevzCart’s earlier analysis of the hidden inventory crisis. That article explains the business risk; this guide provides the operating controls and calculations. Inventory also connects directly to return-rate reduction and supplier management.
Stock on hand is the physical or recorded quantity at a location. Available stock excludes units already reserved and units that are damaged, quarantined, or otherwise not sellable.
Choose a consistent event that matches the operation, such as confirmation, allocation, packing, or dispatch. Many businesses reserve earlier and finalize stock-out at packing or dispatch. The system must also release reservations correctly after cancellation.
Use average daily demand multiplied by average supplier lead time, then add an evidence-based safety-stock quantity. Review the inputs regularly because demand and lead time change.
It can be enough for a small, controlled operation with few SKUs, one location, and limited concurrent updates. It becomes risky when multiple channels, warehouses, users, reservations, or frequent returns require real-time coordination.
Keep it unavailable while it is with the courier. Receive and inspect the parcel when it returns, then classify it as sellable, needs repacking, damaged, or another controlled condition.
No single KPI is sufficient. Stock accuracy is foundational, but it should be reviewed alongside stockouts, turnover, ageing, sell-through, and adjustment value.
Use risk-based cycle counting. Count valuable, fast-moving, or historically inaccurate items frequently and complete a broader reconciliation at an interval appropriate to the business.
No. It improves visibility and alerts, but supplier delays, unexpected demand, inaccurate receiving, and poor parameters can still cause stockouts. Good software must be paired with disciplined processes.
Effective e-commerce inventory management is a chain of controlled events: create clean SKU data, receive accurately, store by location, reserve against real orders, record every movement, inspect returns, count routinely, and replenish from evidence. Bangladesh sellers do not need to automate everything at once, but they do need one trustworthy inventory record and clear ownership of every change.
If disconnected product, order, stock, and courier records are making daily reconciliation difficult, review DevzCart’s verified inventory and operational modules and request a workflow demonstration using your real SKU, warehouse, order-status, and return scenarios.