Beauty Appliance Reorder Points and Safety Stock: Replenishing Without the Guesswork

2026-11-02 · By the Aolemon export team

Most importers reorder when the shelf looks thin and then spend the next seven weeks apologising to customers. That is not a decision failure at the moment of reordering; it is a failure to write down, in advance, the stock level at which a purchase order has to leave your desk. The number has a name, the reorder point, and it is arithmetic rather than instinct: expected demand during the replenishment pipeline plus a buffer for the fact that neither demand nor the pipeline behaves as planned. The common trap is measuring the pipeline as a single figure, the lead time a supplier quotes, which is a commercial promise rather than an observed distribution. Before the arithmetic can work, the pipeline has to be split into stages and timed from your own receipt dates.

Ask for the milestone dates behind your last six shipments and you will normally find that the quoted lead time covers production only. A realistic pipeline runs from purchase order to shelf: order confirmation and artwork approval, motor and component availability, production, pre-shipment inspection, booking wait, inland haulage, ocean transit, destination port dwell, and the warehouse appointment. Booking wait is the stage that quietly doubles in the weeks before a peak season, and destination dwell is the stage importers never measure because the forwarder reports arrival rather than availability. Convert every stage into calendar days from your own records, then compute a mean and a spread per stage, weighted by the lanes and seasons you actually ship on.

The standard safety-stock formula is worth using in full, because it puts the trade-off on the table: safety stock equals the service-level factor multiplied by the square root of lead time times demand variance plus demand squared times lead-time variance. In plain terms, uncertainty in demand and uncertainty in when the container lands both consume buffer, and for ocean imports the second term usually dominates. Take a SKU selling 300 units a week with a weekly demand deviation of 90 units, an eight-week pipeline, a pipeline deviation of 1.5 weeks and a 95 percent target, where the factor is 1.65. The buffer lands near 850 units, about three weeks of cover, and the reorder point near 3,250 units. Raise the target to 98 percent and the buffer grows by roughly a quarter, which is what two points of availability actually cost.

The arithmetic then collides with how factories and containers actually work. Order quantity is not a free variable: it is a multiple of a minimum order quantity and it is usually pushed upward to fill a container. A reorder quantity the formula likes may sit below the MOQ, and rounding up quietly adds weeks of cover and ties up cash. Rather than optimise an economic order quantity that the MOQ step invalidates, compare total landed cost per unit at the three or four quantities you can actually buy, carrying cost included, which for consumer electronics typically runs 20 to 25 percent of inventory value a year. Mixed containers across SKUs and colour versions are often the better answer than a larger single-SKU buy, and a lower MOQ at a slightly higher unit price is frequently cheaper than the inventory it avoids.

None of this survives without a forecast the factory can plan against, and a forecast is only useful when it separates commitment from intention. Send a rolling twelve-month view every month with two windows: a frozen window inside the pipeline where quantities are firm, and a flexible window beyond it where they are indicative. Per line, the factory needs SKU, colour and plug version, quantity by month, and the share of the forecast for which you accept liability on long-lead components they buy ahead. Keep the same file format each month so that bias is visible, because a forecast that is consistently 30 percent high is worse than a flat one: the factory stops believing both. Air freight is a rescue tool for launches and hero SKUs rather than a planning method, so price the air premium per unit against the margin lost per week of stockout before you use it.

Close the loop with five numbers reviewed monthly: weeks of cover by SKU, fill rate against the orders you accepted, forecast bias, stock aged beyond 180 days, and inventory turns. Then put the planning assumptions into the purchase order and the supply agreement, because a buffer is only as good as the promises behind it. The clauses that matter are a lead-time commitment measured to departure rather than to booking, a monthly capacity reservation with a stated consequence for missing it, a rolling-forecast liability cap, the right to pull orders forward or reduce them inside the flexible window, and an explicit sentence that factory-caused delays do not silently consume your safety stock. Collect the five numbers, agree the two windows, and the reorder point stops being a guess.

← Back to all articles