Size-run completeness and core sizes
Finds the core sizes of a style, shows whether the run is intact, how many weeks of sales each size covers and which sizes to top up or redistribute.
At a glance
- What question does it answer?
- Whether a style’s size run is intact, which sizes to top up and which to redistribute.
- What data is needed?
- Stock and sales by size for one style over a period when all sizes were in stock.
- How to read the result?
- The core is the sizes giving the set share of sales. If any core size has stock below the display minimum the run is breaking: the customer cannot find their size. The core shortfall shows how many pieces are missing to the season end; surplus outside the core shows what to redistribute or mark down. Example on the default values: Sizes in the core — 3 of 6 · 77% of sales; Run completeness (core in stock) — 66.7% · run is breaking; Core shortfall to the season end — 107 pcs; Surplus outside the core — 51 pcs.
Why and how to reason
A size run is complete not when all sizes are in the warehouse but when the sizes in demand are. So we first find the core: sizes in descending order of sales until their cumulative share reaches a threshold (for example 75%). The rest are edges. Run completeness = key sizes with stock at least the minimum / number of key sizes.
Cover is for how many weeks a size’s stock lasts at the current pace: stock / (sold / weeks of data). Core deficit to the end of the season is how much must be added to last until the end; excess outside the core is what lies beyond demand.
Example. Four-week sales: S 32, M 46, L 38, which is 77% of all 150: a core of three sizes. In the warehouse S 0, M 10, L 90. Completeness 2 of 3 = 66.7%. To the end of the season (6 weeks) S 48 and M 69 are needed, in stock 0 and 10: a deficit of 107 units. Outside the core 51 excess. Margin at risk 107 x 2,100 = RUB 224.7 k.
When to use it
Use it in season while the style is still selling, to redistribute or top up sizes in time. Suitable for styles with a long size run. Assumption: past weeks reflect future demand. Count sales data only when the size was in stock, otherwise zero sales will mislead. The difference between regions' curves is useful when you distribute between them.
How to read the result
Completeness below the threshold (60%) means a “broken” run: the buyer will not find their size and the style’s sales fall. In the example 66.7%: above the threshold, but size S is missing. The core deficit is converted into money of margin at risk. Excess outside the core is stock better redistributed or included in markdown. Decision: top up or transfer the deficit sizes now and show the surplus edge sizes on clearance. If the run is broken, first redistribute stock between stores and only then reorder. Check whether the needed sizes lie elsewhere in the network: that is cheaper than reordering. For the edges of the run decide: clear, keep or drop.
Common mistakes
Computing the core from buying rather than sales; not noticing that a size sold out and has no sales; taking weeks of data and weeks to the end in different units; setting the display minimum above what is feasible. Also: ignoring differences between regions. Also: not recomputing the core as the season progresses.
What data is needed?
- Stock and sales by size for one style over a period when all sizes were in stock.
- Weeks of data and weeks to the season end.
- Optional: price, cost and sales of another region.
Inputs
| Field | Unit | Value in the example | Meaning |
|---|---|---|---|
| Weeks in the sales data | weeks | 4 | Take weeks when all key sizes were in stock and there were no discounts. |
| Weeks to the end of the season | weeks | 6 | How many weeks remain to the end of the selling season: the core shortfall and the surplus outside the core are calculated for them. |
| Core: share of sales | % | 75 | A rule of thumb is 70–80%, set your own. The core is the smallest set of sizes giving this share. The defaults are set for apparel. For footwear count the shares by foot length (half sizes), accessories have no size curve, for children’s use height: see the size reference. Take your own data. |
| The run counts as broken if completeness is below | % | 60 | The run completeness threshold in percent: a run with fewer than this share of sizes in stock counts as broken. The choice is yours; the example value is illustrative. |
| Display minimum per size | pcs | 2 | A size with less stock counts as missing. Empty: one piece is enough. |
| Price excl. VAT (optional) | ₽ | 4,100 | Unit price excl. VAT. Together with cost gives the margin at risk in the core; empty: it is not calculated. |
| Unit cost (optional) | ₽ | 2,000 | Unit cost excl. VAT. Empty: the margin at risk is not calculated. |
| Threshold: difference between two regional curves | pp | — | If the difference is larger, the regions need their own curve. Empty: no threshold, the color does not change. There is no single standard in Russia: enter your own. |
How it is calculated
| Step | Formula |
|---|---|
| Core | sizes by falling sales until the cumulative share reaches the threshold |
| Run completeness | key sizes with stock at least the minimum ÷ key sizes |
| Cover | stock ÷ (sold ÷ weeks of data) |
| Regional difference | Σ |share of size A − share of size B| |
Worked numbers
Default values from the app; in the interactive version you enter your own.
- Sizes in the core
- 3 of 6 · 77 % of sales
- Run completeness (core in stock)
- 66.7 % · run is breaking
- Core shortfall to the season end
- 107 pcsHow many core pieces are missing to cover demand to the end of the season at the current pace.
- Surplus outside the core
- 51 pcsPieces above demand to the season end in outer sizes: candidates for redistribution or markdown.
- Margin at risk in the core
- 224,700 ₽Missing core pieces × unit margin. An upper estimate: some buyers will take another size.
- Difference between two regional curves
- 10.7 p.p.Sum of absolute differences in size shares, in percentage points. You set the threshold in the field above; there is no single standard, set your own.
Sizes: share, cover and shortfall
| Size | Share of sales | Core | Stock | Cover, weeks | Needed to season end | Stock − needed |
|---|---|---|---|---|---|---|
| XS | 4 % | 12 | 8 | 9 | 3 | |
| S | 21.3 % | yes | 0 | 0 | 48 | −48 |
| M | 30.7 % | yes | 10 | 0.9 | 69 | −59 |
| L | 25.3 % | yes | 90 | 9.5 | 57 | 33 |
| XL | 13.3 % | 60 | 12 | 30 | 30 | |
| XXL | 5.3 % | 30 | 15 | 12 | 18 |
Key sizes missing: S. The customer cannot find their size and leaves; outer sizes pile up and end in markdown. Redistribute surplus from stores and top up the core while the season runs.
Build the curve only from weeks when all key sizes were in stock and without discounts: otherwise the demand for a missing size is invisible in sales. Do not mix online and offline: try-on returns distort the online curve.
Nuances and limits
- Do not build the curve from weeks with discounts or with sizes missing.
- Do not mix online and offline: try-on returns distort the curve.
- Regional differences are a hypothesis, check them on your data.