E-commerce replenishment: set your reorder points

Every week, you need to decide which products to reorder and how much to buy. Many merchants settle it by guesswork, or with a minimum stock threshold in a spreadsheet that never changes. On one side, the best sellers run out; on the other, the warehouse fills with dead stock. Both are costly: according to analyst firm IHL Group (2025), stockouts and overstocks cost global retail $1.73 trillion a year, 6.5% of sales.

The intelligent replenishment case study follows a multi-brand retailer with seasonal sales through its online store and several marketplaces.

Choose when to order and how much to buy

Before you calculate anything, you have to pick a replenishment approach. The methods differ in two ways: the trigger (a fixed date or a stock threshold) and the quantity (a fixed volume or one recalculated each time).

A fourth method replenishes inventory up to a target level. Push/Pull and DDMRP also distinguish forecast-based orders from orders triggered by actual demand. Ford W. Harris set out EOQ in 1913 in an article later reprinted in Operations Research. The reprint was published in 1990. But EOQ assumes stable demand, rare in e-commerce.

For most e-commerce catalogs, the sensible default is the reorder point coupled with a dynamic Min-Max. Demand is continuous but uneven, supplier lead times are known, and the right reflex is to trigger on a threshold rather than on the calendar. For stores using the WooCommerce WMS integration, imported sales and available inventory provide the inputs for this replenishment work.

Measure the selling pace (12 / 4 / 2-week windows)

Everything starts with sales velocity: how many units a SKU sells per week, on average, over a recent window. Without that figure, no reorder point means anything.

The problem with a spreadsheet is relying on a single sales window. If you always look at the last 30 days, you smooth out sales spikes and miss new launches. Compare several rolling windows:

The short window gives an early warning; the long one keeps the underlying trend in view. In myFulfillment, the system compares recent sales with the same period last year to adjust thresholds for seasonality. In the intelligent replenishment case study, fixed minimums left summer best sellers out of stock while winter items sat idle.

Calculate the days before stockout (available stock, backorders, supplier lead time)

Once you know sales velocity, calculate how many days remain before this SKU runs out of stock. That is the days-before-stockout indicator.

Divide available stock by average daily sales to calculate days of supply.

Available stock is physical stock minus units reserved for open orders. Using physical stock overstates your days of supply and makes you order too late.

Backorders are orders you have accepted but cannot ship because stock is unavailable. Allocate the required units from the next receipt. Your stock needs to cover sales during the supplier lead time. If stock runs out in ten days and replenishment takes fifteen, the delivery will arrive after the stockout.

In myFulfillment, Stock Helper displays days before stockout per warehouse, with supplier lead time built in. When projected days of supply drop below the threshold, the SKU automatically appears in the purchasing recommendations table.

Trigger the order at the right stock level

The reorder point is the stock level that should trigger the order. It has several names: reorder point, replenishment threshold, reorder level, or ROP. The safety stock is the buffer that absorbs the unexpected: a demand spike, a late supplier.

Calculate the reorder point as follows:

Reorder point = (daily velocity x supplier lead time) + safety stock

Place the order when available stock reaches the reorder point. Recalculate the threshold when sales velocity or supplier lead time changes.

The reorder point sets timing; min/max sets quantity.

Match safety stock to the SKU

Safety stock is calculated in three ways, from the simplest to the most statistical.

The average-max formula compares maximum sales and lead time with their averages. The NIST table (2012) gives Z values of 1.282 for 90%, 1.645 for 95%, and 2.326 for 99%. These probabilities set the target service level in the model. A higher target increases safety stock.

For a stable SKU with reliable lead times, start with safety stock expressed in days of sales and review it as demand changes. For a critical SKU, estimate sales variability and check the assumptions before applying the statistical formula. Z sets the service target; standard deviation measures variability.

Let the system propose the min/max per SKU

For every SKU due for replenishment, Min-Max suggests a minimum and a maximum quantity to order. The recommendation also includes an estimated pack quantity and shipping lead time.

The minimum quantity covers demand until the next delivery you can realistically receive. The maximum quantity factors in your capacity and optimal storage duration, so you do not buy too many units of a SKU whose sales are slowing. The operator reads the proposal, adjusts it if they have information from day-to-day operations that the system lacks, and approves it.

The operator checks the suggested quantities before sending the order. For authorized workflows, purchase rules can retain an optional approval step.

Create the purchase order with landed cost and margin calculated

A recommendation must be immediately convertible into a purchase order.

In myFulfillment, purchase orders are pre-built from the replenishment needs and grouped by supplier, ready to review and send in one click. You enter the suggested min or max quantity and add the line to the purchase order. The purchase order progresses from a draft to a sent order, then to awaiting delivery and completion. Partial receipts are supported, and the supplier is notified by email.

The software calculates landed cost, including freight and taxes, and the corresponding margin in the purchase order. You then decide what to buy with the margin already known.

Read inventory turnover alongside stockouts

Inventory turnover measures how many times you sell through and rebuild stock over a period. Calculate average days in stock by dividing the number of days in that period by its turnover rate.

Use the rate to identify problems:

• A rate that rises without creating stockouts signals replenishment well tuned to real demand: you tie up less cash for the same service level.
• A rate that collapses reveals dormant stock, order quantities that are too high relative to actual sales velocity.
• A rate that spikes but comes with stockouts points the other way: reorder point or safety stock set too low.

Track turnover by product category alongside the stockout count. A seasonal product and a year-round staple need different settings.

Reduce purchases when sales slow

If you track only stockouts, you end up filling the warehouse with products that no longer sell. You also need to reduce purchases at the right time.

The adjustment is automatic: dynamic replenishment cuts ordered quantities when velocity declines. Where a fixed minimum keeps triggering purchases of a product that no longer moves, a velocity-based threshold drops by itself. At the retailer we worked with, this brought recurring overstock on slow movers under control.

The Stock Helper adds explicit guardrails: maximum storage capacity, optimal storage duration, warning and ideal stock levels, and discontinued-product flagging. In practice, you set a ceiling per SKU so you do not run out of storage space. You also take products you no longer want to rebuy out of the replenishment calculation. For workflows without storage, the cross-docking guide describes the handoff from receiving to shipping. Examples include the multi-supplier cross-dock of an auto-parts distributor and the supplier-side cross-dock and dropship automation.

Automate purchase orders using your buying rules

Automate orders when manual review no longer keeps up. The retailer in the intelligent replenishment case study managed over 12,000 SKUs and spent 2 days a week on purchasing in Excel.

Automation happens in stages, not all or nothing. Purchase orders can be created manually or automatically from the replenishment needs. Purchase rules run in priority order for each store or warehouse when their conditions are met. You keep an optional approval step: the system generates the PO from your rules, and the operator approves it before sending. For controlled workflows such as dropshipping, the system can send the purchase order automatically by email, file, or API.

In the intelligent replenishment case study, time spent on weekly purchasing fell from 2 working days to a 30-minute review. Stockouts on top sellers fell from 15–20 per month during peak periods to zero within the first quarter.

To see this method tooled end to end, from velocity measurement to purchase order, look at the myFulfillment replenishment and purchasing page. For Magento stores, the Magento ERP integration for inventory and purchasing shows how catalog, orders and sellable stock feed this operating model. Our procurement masterclass walks through it live.