Three weeks before a launch, the cartons are still on a press in another country, and the marketing spend is already booked. The order was placed on time by the calendar the buyer had in their head, which was built from a quoted lead time rather than from the range of lead times the supplier actually delivered over the previous year. Packaging safety stock is the buffer that closes that gap, and the number behind it is arithmetic rather than judgement.
What the buffer is protecting against. Two different risks get mixed together in many planning conversations. One is demand variance: the forecast was wrong and the promotion sold more than planned. The other is supply variance: the order took longer than expected, or arrived with a defect that required a partial reprint. Packaging faces both, and the supply side is usually the larger of the two because print runs cannot be topped up quickly. Separating the two is what makes a reorder point meaningful.
The basic calculation. A reorder point equals average demand during the replenishment period plus a buffer for uncertainty. The replenishment period is not the quoted lead time; it is the quoted lead time plus transit, customs, inward inspection and any internal approval before the stock is usable. Averaging lead time over twelve months of actual receipts gives a realistic figure, and taking the standard deviation of those receipts gives the size of the uncertainty. The buffer is then a multiple of that deviation, chosen by how much stockout risk the business is willing to carry.
Why packaging needs a bigger buffer than the formula suggests. Three features of print work push the requirement up. Minimum order quantities mean a shortfall cannot be covered by ordering a few hundred extra units. Tooling and plate changes add a fixed setup that a supplier will not run for a small top-up. And a reprint has to go through artwork approval, proofing and scheduling again, which means the recovery time after a stockout is longer than the original lead time. A buffer sized only on statistical variance ignores the recovery cost.
Seasonality and the calendar. Demand for packaging is not flat. A retailer’s promotional calendar, a holiday season and a Chinese New Year shutdown all create periods where a normal reorder point is too low. The practical fix is to hold a seasonal buffer that is added on top of the base figure for a defined window, then released. Factoring the factory closure into the replenishment period for orders placed in the last quarter is the single change that prevents the January shortages.
The cost side of the trade. Carrying packaging stock costs warehouse space, working capital and, for printed cartons, obsolescence risk if the artwork changes. Board also absorbs moisture over time, so a carton held for a year in a humid warehouse can lose stacking strength even if it looks fine. The buffer should therefore be sized to cover a defined risk window, not to cover a year. For fast-moving items a few weeks of cover is typical; for slow-moving items with long print runs, the buffer often sits in the supplier’s finished goods rather than in the buyer’s warehouse.
Using supplier-held stock. A supplier with spare capacity can hold finished cartons and release them against call-offs, which moves the buffer off the buyer’s balance sheet and shortens response time. This works when the print is stable and the storage conditions are controlled, and it needs a written agreement covering who owns the stock, who insures it, how long it can be held and what happens if the artwork changes. Without those clauses, supplier-held stock is a favour rather than a plan, and it tends to disappear when the factory is busy.
Setting the reorder point in a system. Planning systems want a single number per SKU, refreshed periodically. A workable routine is to review actual lead times quarterly, recalculate the deviation, and adjust the buffer for known events such as a holiday shutdown or a capacity crunch. The review matters more than the initial calculation, because lead times drift and a reorder point set two years ago is usually wrong in both directions across a range. Where a system cannot handle a seasonal overlay, a manual calendar note achieves the same result.
The signals that a buffer is too small. Repeat expedites, air freight on cartons, partial shipments split across two sailings and repeated requests for early release are all symptoms. So is a pattern of running the last few days of a run on a substitute pack. Each of these has a cost that is easy to see in isolation and hard to total, which is why the buffer conversation often needs a number: the cost of one air freight shipment against the annual carrying cost of the extra stock usually settles it.
The signals that a buffer is too large. Slow-moving stock ageing past a year, artwork changes stranding printed cartons, and warehouse space becoming the constraint all point the other way. In those cases the fix is usually in the ordering pattern rather than the buffer: smaller and more frequent releases from a supplier-held stock, or a review of the print run length against the actual annual volume. A run length that covers eighteen months of demand is a decision to hold eighteen months of risk.
Tying the plan to the supply base. A buffer is cheaper when the supplier is reliable, because the deviation term in the calculation shrinks. That is the argument for tracking on-time delivery per supplier and reviewing it with them, and for keeping a second source qualified even at low volume. Buyers who do both often find they can cut the buffer without raising risk. The lead time drivers are worth understanding before the number is set, since half of the variance usually comes from two or three specific steps.
If a stockout has cost you a promotion, or the buffer number has not been calculated properly, send us the annual volume, the print run length and the lead times you have seen over the last year. We will tell you how much of the buffer could sit in our finished goods instead of your warehouse.
