Most small and mid-sized retailers manage purchasing through email threads, spreadsheets, and supplier phone calls. It works, in the sense that stock generally arrives and operations continue. What it does not do is give you visibility into what was ordered, what was received, and what the variance between the two was. That gap shows up as stockouts, over-ordering, and receiving errors that accumulate into margin erosion.
The Cost of Informal Purchase Order Management
Most small and mid-sized retailers manage purchasing through a combination of email threads, spreadsheets, and supplier phone calls. It works, in the sense that stock generally arrives and operations continue. What it does not do is give you visibility into what was ordered, what was received, and what the variance between the two was.
The cost of informal PO management shows up in three places. First, stockouts: without a structured system tracking reorder points and lead times, replenishment decisions are reactive rather than predictive. You run out of a product, notice the gap, and place an emergency order at whatever cost is available. Second, over-ordering: without visibility into what is already on order, buyers sometimes place duplicate orders or order quantities that exceed demand, tying up cash in excess inventory. Third, receiving errors: without a purchase order to check deliveries against, discrepancies between what was ordered and what was delivered are often not caught until a cycle count reveals the variance, weeks after the delivery.
A structured purchase order management system addresses all three by creating a documented, traceable record of every order from the moment a replenishment need is identified to the moment goods are confirmed received and matched against the invoice.
The Five Stages of a Purchase Order Lifecycle
A well-designed PO management system tracks every order through a defined lifecycle with clear state transitions. Each state transition creates a record, assigns accountability, and triggers the next action.
Stage 1 is Draft. A buyer identifies a replenishment need, selects a supplier, and creates a draft PO with the items, quantities, and expected unit costs. The draft stage allows review and adjustment before the order is committed.
Stage 2 is Submitted. The PO is sent to the supplier, either electronically or by email. The submission records the date, the agreed quantities, and the expected delivery date. From this point, the PO is a contractual commitment and should not be modified without a formal amendment process.
Stage 3 is Confirmed. The supplier acknowledges the order and confirms availability, quantity, and delivery schedule. Not all suppliers provide order confirmations, but capturing this step when available creates a cleaner audit trail and makes delivery discrepancy resolution easier.
Stage 4 is Received. Goods arrive at the warehouse or store, and a receiving team counts and inspects the delivery against the PO. Quantities received are recorded against the PO line items. Discrepancies, short shipments, or damaged goods are flagged for follow-up. This is where blind receiving protocols (counting before seeing the PO quantity) catch supplier fraud and errors.
Stage 5 is Completed. The received quantities are matched against the PO and the supplier invoice. Any discrepancies are resolved. The PO is closed, inventory is updated, and the invoice is approved for payment. The completed PO record serves as the audit trail for the inventory receipt and the accounting entry.
- Draft: buyer creates and reviews the order before commitment
- Submitted: order sent to supplier, quantities and delivery date locked
- Confirmed: supplier acknowledges availability and schedule
- Received: goods counted against PO, discrepancies flagged
- Completed: invoice matched to PO, inventory updated, payment approved
Setting Reorder Points and Safety Stock
The most impactful improvement a retailer can make to purchase order management is shifting from reactive replenishment (ordering when stock runs out) to predictive replenishment (ordering before stock reaches the minimum level needed to cover the lead time).
The reorder point (ROP) is the inventory level at which a new order should be placed so that stock does not run out before the order arrives. The formula is: ROP = (Average Daily Sales x Lead Time in Days) + Safety Stock.
Safety stock is a buffer against demand variability and lead time variability. A product that sells an average of 10 units per day but varies between 5 and 20 units per day needs more safety stock than a product that consistently sells 10 units per day. The safety stock calculation accounts for the standard deviation of both demand and lead time.
For most retailers, calculating and maintaining ROPs manually across hundreds or thousands of SKUs is not feasible. The practical solution is a system that tracks daily sales velocity by SKU, records supplier lead times, and automatically calculates and updates reorder points. When any SKU reaches its ROP, the system generates a replenishment recommendation that a buyer can review and convert to a PO with one action.
How ML-Powered PO Pre-Fill Works
Advanced inventory management systems go beyond reorder points to generate purchase order quantity recommendations using machine learning models trained on historical sales data.
A basic reorder point calculation tells you when to order. An ML-powered replenishment model tells you when to order and how much to order, based on projected demand over the order cycle. The model accounts for sales velocity trends, day-of-week and seasonality patterns, upcoming promotional events, and lead time variability by supplier.
The output is a pre-filled purchase order draft that a buyer reviews rather than creates. Instead of calculating quantities for 200 SKUs across 8 suppliers, the buyer reviews and approves pre-calculated quantities, adjusting where their judgment departs from the model (knowledge of an upcoming promotion, awareness of a supply constraint, a planned product discontinuation). This shifts the buyer's role from calculation to review, compressing the time to order placement significantly.
The quality of ML-powered recommendations improves over time as the model accumulates more transaction history and learns the demand patterns specific to each SKU and location. A retailer with 24 months of clean transaction data typically sees better recommendations than one with 6 months, because seasonal patterns, year-over-year trends, and supplier lead time variance are better represented in the longer dataset.
Landed Cost Allocation and True Inventory Costs
The cost recorded in your inventory system when goods are received should be the true cost per unit, including all costs incurred to get that unit to your warehouse. This is the landed cost, and it differs from the supplier invoice price whenever shipping, customs, duties, or handling fees are involved.
Landed cost allocation distributes these additional costs across the units in a shipment proportionally. If a $500 shipment of 200 units arrives with a $50 freight charge, the landed cost per unit is $2.75 instead of $2.50. That difference flows into COGS when the units are sold, affecting gross margin calculations. If you record only the supplier invoice price and expense freight separately, your per-product margin reports will consistently overstate profitability for imported or shipped goods.
For retailers with international suppliers, freight, customs duties, and insurance can represent a substantial portion of true landed cost. Accurate landed cost allocation requires the PO management system to support the addition of freight and duty costs to a shipment and the distribution of those costs across SKUs, typically in proportion to the value or weight of each line item.
Getting landed costs right matters most for pricing decisions. If your cost basis is understated because landed costs are not allocated to inventory, your margin calculations will indicate acceptable profitability even when the true margin on international goods is insufficient. This leads to chronic underpricing in categories sourced from suppliers with significant freight costs.
Purchase order management is the connective tissue between your inventory records and your supply chain. When it works well, stockouts are rare, over-ordering is controlled, and every inventory receipt is verified against a documented order. When it is informal, the gaps accumulate as shrinkage, margin erosion, and cash tied up in inventory that should not have been ordered. The infrastructure to do it right is not complex: a five-stage PO lifecycle, reorder points calculated from live sales velocity, and landed cost allocation at the line-item level. The key is that all of it needs to be connected to your inventory system and your general ledger so that every PO completion automatically updates inventory counts and creates the corresponding accounting entry.
POs That Write Themselves, Receiving That Catches Every Discrepancy
Momentum's purchase order management includes ML-powered PO pre-fill from live sales velocity, a five-stage order lifecycle with full audit trail, blind receiving support, and landed cost allocation. Every completed PO creates the correct journal entry automatically. Book a demo to see the full workflow.
