Syntha Lab RU

Size curve by store cluster and pack composition

Splits the buy across store clusters and sizes using each cluster’s past sales, picks the composition of a ready pack (prepack) for the cluster’s curve, rounds the buy to whole packs and shows how many pieces end up in the wrong sizes. Compares a pack per cluster with one pack for the whole chain, and picking to order with a ready pack.

Assortment and sizes For whom: Chain buyers and merchandisers who distribute a style to stores and order ready packs.

Calculate on your data →

At a glance

What question does it answer?
How many pieces of which sizes to send to each store cluster and which pack to choose.
What data is needed?
The buy for the style and the pieces per pack.
How to read the result?
Default example: a buy of 1,250 pcs with a pack of 6; the center gets 1-1-2-1-1 (83 packs, 498 pcs, rounding rest 2 pcs), residential areas 0-1-2-2-1 (75 packs), regions 0-1-1-2-2 (50 packs). In total 208 packs, 1,248 pcs shipped. In the wrong sizes: 126 pcs (10.1%) with a pack per cluster against 155.7 pcs (12.5%) with one pack for the chain: a pack per cluster is better by 29.7 pcs. Picking to order costs 35,000 more (a difference of 28 per piece). Example on the default values: Packs in total (pack per cluster) — 208; Pieces shipped (pack per cluster) — 1,248 pcs; Rounding to packs: shortfall (−) or excess (+) — −2 pcs; Pieces in the wrong sizes: pack per cluster — 126 · 10.1%.

Why and how to reason

Size allocation across stores is the same only on paper: city-center stores buy different sizes from residential districts and regions. So the purchase is split by store cluster and size using each cluster’s past sales, then a pre-pack composition is chosen to follow the cluster curve. Boxes are whole: size 6 pieces, and the purchase is rounded to a whole number of boxes. In the example 1,250 pieces, clusters center 40%, residential 36%, regions 24%; curves differ (regions have more L and XL). A purchase of 1,250 gives 208 boxes, 1,248 pieces (−2 from rounding). Pieces in the wrong sizes with a box per cluster are 126, with one box for the whole network 156: a difference of almost 30 pieces. Made-to-order assembly is dearer than the ready box by 35 thousand ₽ (40 against RUB 12 a piece).

When to use it

Reach for it when buying size ranges for a heterogeneous network: center, residential districts, regions. You need sales by size by cluster for at least last season. Assumption: the structure of demand by size will remain. It does not fit sizeless goods, a small network (one or two stores) or a collection with a fundamentally new fit where history does not work. A box is convenient when the factory or warehouse has pre-pack packaging.

How to read the result

Look at pieces in the wrong sizes: the fewer, the better the box matches demand. A box per cluster gains 30 pieces over a single box: each such piece is a probable markdown. Rounding to boxes gives a shortfall of 2 pieces: acceptable. Compare assembly cost: if made-to-order is 35 thousand dearer while the gain of 30 pieces in sizes at RUB 3,000 margin is 90 thousand, a ready box is better. If the gain is small, take a single box and save on complexity. It is worth testing boxes on a trial batch in one or two stores and check size remainders after a month: the calculation assumes size demand is stable and this needs confirming.

Common mistakes

Using sales only by sizes in stock (missing sizes did not sell, and the curve is biased toward availability); not normalizing cluster sales; setting a box bigger than a store accepts. A box ignores the specific needs of a store: check against sales. Do not make a box larger than a store can unpack and display in a day.

What data is needed?

  • The buy for the style and the pieces per pack.
  • Per cluster: the share of the buy and last season’s sales by size.
  • Optional: the rates for picking to order and for a ready pack.

Inputs

FieldUnitValue in the exampleMeaning
Buy for the stylepcs1,250Total pieces for the style across all clusters.
Pieces per packpcs6The size of a ready pack under your supplier or warehouse terms. 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.
Picking to order, per piece₽40At your warehouse or fulfillment rate, excl. VAT. Empty: the cost comparison is not calculated.
Shipping a ready pack, per piece₽12At the same rate. Empty: the cost comparison is not calculated.

How it is calculated

StepFormula
Size share in a clustersize sales ÷ cluster sales
Pack compositionshare × pieces per pack, integer part; missing pieces by the largest remainders
Packscluster buy ÷ pieces per pack, rounded to a whole number
Wrong sizesΣ over sizes max(0; shipped − cluster buy × size share)

Worked numbers

Default values from the app; in the interactive version you enter your own.

Packs in total (pack per cluster)
208
Pieces shipped (pack per cluster)
1,248 pcs
Rounding to packs: shortfall (−) or excess (+)
−2 pcs
Pieces in the wrong sizes: pack per cluster
126 · 10.1 %
Pieces in the wrong sizes: one pack for the chain
155.7 · 12.5 %
What a pack per cluster gives, pieces
29.7
Picking to order costs more than a ready pack
35,000 ₽(Order-picking rate − pack rate) × the buy.

Pack XS-S-M-L-XL by cluster

ClusterBuy, pcsPack per clusterPacksRounding restWrong sizesWrong sizes, common pack
City center, malls5001-1-2-1-183213.3 %16.5 %
Residential areas4500-1-2-2-17506.7 %6.7 %
Regions3000-1-1-2-250010 %14.4 %

The curve comes from the cluster’s past sales and does not know about stock availability: if a size was out of stock, its share is understated. For a new item use analogues.

Nuances and limits

  • Taking a curve from sales where a size was out of stock: its share is understated.
  • Using a pack smaller than the number of sizes: some sizes will not reach the store.
  • Not rounding the buy to whole packs: the order will contain a fractional number of packs.

Terms used

Calculate on your data →