Two green gardening gloves with striped knitted cuffs on grey fabric.
The first garden glove order was a test: twenty pairs each of small, medium and large, one store, one season. Four weeks in, medium is gone and the other two sizes are still on the peg. The reorder has to be placed now, and the table of sales figures looks like it contains an answer.
It contains a question instead. Medium sold out because it sold out, and after that nobody could buy medium. The twenty pairs of sales in that size measure demand up to the day the peg emptied, and nothing after it.
Read sales alongside the days each size was available
The first correction is to put the sales number back into the time it happened in.
Here is a worked example, and it is an assumption constructed for this article rather than a record from any store. Twenty-eight days observed, no restocking, no returns.
| Size | Opening stock | Days available | Units sold | Closing stock |
|---|---|---|---|---|
| S | 20 pairs | 28 | 5 | 15 |
| M | 20 pairs | 10 | 20 | 0 |
| L | 20 pairs | 28 | 8 | 12 |
Read as a table of sold units, medium is the winner at twenty. Read as a rate, the picture changes completely.
Small sold 5 pairs over 28 days, which is about 0.18 pairs per day. Large sold 8 over 28, about 0.29 per day. Medium sold 20 pairs in 10 days, which is 2.0 pairs per day.
Medium's observed in-stock sales rate is roughly seven times large's and eleven times small's. Those ratios compare different availability windows, not fully observed demand under identical conditions. The table does not show how many medium pairs customers would have bought after day 10, or whether the first ten days were typical. Closing stock alone cannot answer either question.
The second thing to fix is that medium's demand after day 10 is unobserved. The 2.0 pairs per available day is its observed in-stock sales rate, not a proven lower bound on average daily demand over the full 28 days or the next selling period.
That is the honest conclusion from a sold-out size: the available stock limited what could be observed. A measured increase in the next test order may be worth considering, with its size based on the next selling period and the risk of leftover stock.
Ask what the returns are telling you
The second input is the one a sales table cannot contain, and it changes the size recommendation more often than the sales data does.
Returns have reasons, and the reasons separate into three groups that must be handled differently.
Fit reasons: too small, too large, tight across the knuckles, loose at the wrist. These are evidence about fit and can inform the next order's size mix. A pattern of small returns in one size and large returns in another can indicate that the sizing runs differently from what customers expect, which is a specification issue rather than a mix issue.
Quality reasons: a seam failure, a coating worn through, a tear at the thumb. Record these separately from fit complaints and investigate whether they affect one size, a batch or the whole range. Do not interpret them automatically as weak demand for that size, but do consider whether the quality issue must be resolved before it is reordered.
Other reasons: wrong item, changed mind, gift. Record these separately and investigate any exchange or fulfilment pattern before deciding whether they inform the size mix.
Getting this data requires the store to record a reason at the point of return. A retailer that only counts returns, without reasons, has thrown away the most useful half of the information, and a supplier selling into that retailer has an interest in asking for it.
The worked example deliberately has no returns, because a table with assumed return reasons would look like data. In practice, the second analysis is where a size decision usually gets corrected, and it should be run before the mix is set.
Adjust for the next selling period
The third step is the part that separates a reorder from an extrapolation.
Several things change between the observed period and the next one, and each shifts demand in a direction.
Season. Gloves are bought for spring planting, for summer maintenance, for autumn clearance. A size that sold fast in a spring promotion may be a different proposition in a quiet month, and the mix that fits one period does not automatically fit the next.
Promotion. If the sold-out size was on promotion, its rate reflects the promotion as well as the size. If a promotion is planned for the next period on a different size, the previous period's rates are a poor guide.
Restock timing. Medium ran out on day 10. The sales it would have achieved with stock available for all 28 days are unknown; use the stockout timing to plan the next observation period and replenishment test, not as an automatic reason to over-order.
Existing stock. Small has 15 pairs left and large has 12. The next order is not a fresh mix; it is an addition to inventory already on hand. Compare the remaining stock with each size's planned requirement before deciding what to order.
Lead time and pack configuration. The size mix has to be ordered in whatever pack ratios the supplier offers, and a mix of, say, 30% small, 45% medium and 25% large may not be available as a carton configuration. The order is a negotiation between the desired mix and the pack, and the pack constraint should be known before the mix is set.
Write the size mix with its assumptions attached
The output of the analysis is a short written rationale, and its value is in what it admits.
Write the observed rates with the days-available correction. Label the sold-out size's figure as an observed in-stock sales rate, and state what demand remains unobserved. State the return reasons, and state which ones were excluded from the size analysis and why. Note the seasonal and promotional conditions of the observed period, and how the next period differs. State the existing stock position by size. State the pack configuration constraint.
Then write the order: which sizes up, which sizes down, by how much, and the reason for each. A larger medium allocation can be tested if comparable selling conditions are expected, while the small and large quantities should reflect their planned requirements and stock on hand. State the chosen quantities as a purchasing decision under uncertainty, not as numbers proved by this table.
Finally, write what to watch in the next period. If medium sells out again at a higher stock level, compare the quantity, days available, promotion and selling conditions before concluding that its rate increased or ordering more. If small continues at 0.18 pairs per day with 15 pairs in stock, the next conversation is whether to carry that size at all.
No general ratio transfers between stores. A garden centre in a mild climate with an older customer base and a hardware store in a cold one with a landscaping trade will have different size curves from the same glove. The method transfers; the numbers do not.
And a caution that belongs in the file: this analysis covers one store and 28 days. A second store, or a second period, would either confirm the pattern or break it, and the useful discipline is to run the same analysis on the next period rather than to extend this one's conclusion indefinitely.
Written by
ScarecrowGarden
💡About Scarecrow Garden Supplier Co., Ltd.
Scarecrow Garden Supplier Co., Ltd. is a China-based sourcing and wholesale partner specializing in garden tools, landscaping equipment, and outdoor supplies for international wholesalers, distributors, contractors, and brands.
With hands-on experience rooted in real garden use scenarios, we focus on durable materials, functional design, and stable large-volume supply. Our product range covers pruning tools, watering systems, hand tools, outdoor hardware, and customized garden solutions to support both retail and professional landscaping markets.
Beyond products, we help our partners navigate supplier selection, quality control, compliance requirements, and long-term sourcing strategies in China. Through our blog, we share practical insights on product selection, material comparisons, industry trends, and cost-effective purchasing—helping global buyers build stronger, more competitive supply chains.
