Assortment matrix and assortment plan
The matrix shows which styles, colors and sizes enter the collection, in what quantity and at what price, and the assortment plan links product roles to the budget. It decides what to buy or make, what to drop and which items get depth across sizes.
Why this document
The matrix stops the assortment from growing beyond the budget and demand. Each item gets a role: traffic, margin, image or filler; then it is clear when a collection is all novelties or all expensive pieces. The buyer and merchandiser check the number of styles against size depth: too many styles with limited money mean small lots, missing sizes and a sale at season end. That is why extras are removed before purchase, while that costs nothing.
At a glance
- What is this document and why is it needed?
- The matrix shows which styles, colors and sizes enter the collection, in what quantity and at what price, and the assortment plan links product roles to the budget. It decides what to buy or make, what to drop and which items get depth across sizes.
- Who owns the document and how often is it rebuilt?
- Brand: product director and merchandiser, before the season development starts, agreed with the commercial director. Retailer: category manager and buyer, before season buying, agreed with the head of buying.
- What mistakes are the most common?
- Width without depth → many styles at 1-2 units per cell → sales do not build, turnover falls → check the cell minimum. Identical styles on different ladder tiers → cannibalization → each style gets its own function and tier. The core role is filled with fashion → unsold stock remains after the season → tie the core share to repeat-sale history.
For whom
- Brand
- A brand builds the matrix for its own collection: number of styles, colors, sizes and the price ladder; it must fit fabric and factory MOQ. Roles: core (repeats from season to season), fashion (novelty), image (showcase and recognition).
- Retail
- A retailer assembles the matrix from other brands and the private label: it splits space and budget between brands, price tiers and category roles. The key is not to duplicate model functions or leave empty “price × role” cells.
What it consists of
The “Product role” axis
- Core, fashion, image: shares in money and in styles; an editable hint is 60-70 / 20-30 / 5-10% (the source is not verified: take it from your own seasons).
- Category role: traffic driver, profit core, image, seasonal (the “Category roles” calculation).
- Brand: price steps and MOQ in numbers: a fabric MOQ of 300 m and consumption of 1.6 m per item give a minimum of 187 items per color; with the minimum per SKU of 37.5 units this limits the number of colors per style; steps of 3,900, 4,500 and RUB 5,400 excluding VAT give steps of 15% and 20% (an editable hint, not checked).
The “Price” axis
- Price ladder good-better-best: lower, middle and upper tiers with a gap visible to the shopper (the “Price tiers” calculation).
- Price points: anchor on rounded levels, limit the number of points.
- Retailer: a “brand × price segment” matrix: rows are brands and the private label, columns are price steps (up to 3,000, 3 to 6 thousand, above RUB 6 thousand); each cell holds the purchase budget, the SKU count and the role (core, fashion, image); an empty cell is a hole in the assortment, two cells with the same role and price are duplication. It differs from the brand example in that the role is tied to the brand and the cell minimum is counted per store, not per factory SKU.
The “Width and depth” axis
- Width: number of styles and colors; depth: units per style-color-size.
- Size curve and core sizes; for footwear: the size run and the last.
- Capsules and drops by wave: dates and volume of each.
Brand and retailer: what is computed differently
- Brand: a brand builds the matrix for its own collection: the number of styles, colors and sizes, the price ladder and roles (core, fashion, image), all within the fabric and factory MOQ.
- Retailer: a retailer assembles the matrix from third-party brands and the private label: it allocates space and budget across brands, price steps and category roles, and keeps no empty “price × role” cells or duplicates.
Brand: how the document differs and what to add
- Brand: a brand assortment is a collection by role: core, fashion, image; each role has its own budget share, depth and risk.
- Brand: factory constraints (fabric MOQ, MOQ per color and size) set the minimum depth and number of colors per model; without them the matrix is unrealistic.
- Brand: brand price tiers are tied to the price chain (cost, wholesale, retail): each model must fit a tier with the required margin.
- Brand: the brand size curve depends on the channel: wholesale orders by the partner curve, site and marketplace give their own; the common curve is refined from last season sales.
- Brand: the share of novelties and carry-over models is set by policy: repeat basics reduce risk and development cost.
- Brand: for footwear and accessories the brand adds last width, color batches and minimum runs per pattern; for kids, height ranges.
Retailer: how the document differs and what to add
- Retailer: a retailer assortment is a matrix of brand × price segment × category: empty cells and tier imbalances are visible at once.
- Retailer: width and depth are set by store format: mall, street and small format get different SKU counts and size depth.
- Retailer: the share of private label, exclusives and third-party brands determines assortment margin: the balance is written into policy.
- Retailer: assortment is checked by store cluster: one set for the whole network leaves some shelves empty and overloads others.
- Retailer: delisting and listing of SKUs is planned by turnover and ABC-XYZ: class C SKUs with no role in the matrix are removed first.
- Retailer: novelties and carry-over SKUs are matched by the retailer with the brand’s sales history in the network: a new brand without history gets a limited share and is tested with a small batch.
How to read the result and which decisions follow
- The planned number of SKUs is above the role maximum: cut models or colors or raise the budget; launching without a fix gives one-unit cells and no size choice.
- The share of the “fashion” role grows at the expense of core without a risk review: check fashion sell-through of past seasons; if it is below 60% (a guide) the share should fall.
- More than 40% of SKUs in one price tier (a guide): the matrix is skewed; add a tier above or below based on sales results.
Control checks
- Role budgets 3.7 + 1.7 + 0.6 = RUB 6 m = purchase budget 6; SKUs 48 + 36 + 4 = 88, cells (SKU × sizes) 440; size curve 8 + 22 + 30 + 25 + 15 = 100%.
- The sum of role budgets equals the purchase budget; the number of cells equals SKUs × sizes.
- The size curve adds up to 100%; the per-size minimum is checked for the rarest size, not the average.
- Typical traps: average depth hides zeros at edge sizes, factory MOQ is ignored, colors make one model appear in several roles.
Where the data comes from
| Data source | Fields |
|---|---|
| Past-season sales | Sales by style, color, size, price; sell-through; markdown share. |
| Market and competitors | Price monitoring, width of competitors matrices, empty price tiers. |
| Platform reference books | Categories, attributes, size tables (lab-refs), seasons. |
| The “Product” template | SKU, style, color, size, barcode (the product template) to check the matrix. |
How to build it step by step
- Step 1. Review the past season: bestsellers, slow movers, leftovers, the share of new versus carryover.
- Step 2. Set the category budget from OTB and the category role.
- Step 3. Split the budget by role (core / fashion / image) in money.
- Step 4. Define the number of styles per role and colors per style: the width.
- Step 5. Compute depth: money ÷ unit cost ÷ (SKU × sizes) = units per cell; check the minimum cell.
- Step 6. Place styles on the price ladder; fill empty tiers or drop them deliberately.
- Step 7. Spread across the category size curve; for footwear choose lasts and the core of the size run.
- Brand: Step 8. Check fabric and factory MOQ: if a style falls short, merge colors or drop it.
- Step 9. Reconcile with the sales plan and OTB: the matrix sum equals the budget; record the version.
Formulas
| Measure | Formulas | Note |
|---|---|---|
| SKU and cell count | SKU = Styles × Colors; Cells = SKU × Sizes | Each cell gets at least the minimum (cellmin), otherwise it does not work as a product. |
| Depth | Units per cell = Role budget ÷ Unit cost ÷ Cells | The “Width and depth” calculation. |
| Price tier step | Step = (Upper price − Lower price) ÷ Lower price | A step too small (under 15% as a starting hint, not verified) is not seen by the shopper; the “Price tiers” calculation. |
| Share of new styles | Share of new styles = New styles ÷ All styles (in money and units) | The “New vs carryover” calculation. |
| Minimum per SKU | Minimum per SKU = cellmin ÷ Share of the smallest size in the curve; SKU_max = Role units ÷ Minimum per SKU | With the curve 8/22/30/25/15% and a cell minimum of 3 an SKU needs 37.5 units; for fashion with 1,050 units this is at most 28 SKUs. |
Formulas in notation
- Role units
U_r = B · s_r / c_r
B is the purchase budget, s_r is the role share, c_r is the unit cost of the role.- Number of SKUs
N_sku = M · C
M is the number of models, C is colors per model; matrix cells are N_sku · Z where Z is the number of sizes.- Maximum number of SKUs
N_max = U_r / Min
How many SKUs the role supports at the per-size minimum; the planned number is compared with N_max.
Worked example
Brand: a collection matrix on a RUB 6 m budget at cost (an independent example, not tied to Models B and R)
| Role | Budget share, % | Money, RUB m | Unit cost, RUB | Styles | Colors | SKUs | Cells (with sizes) | Units | Units per cell (average) | Units per SKU | Minimum per SKU at a cell minimum of 3 and an XS share of 8% | Sizes below the minimum | SKUs that can be held |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Core (basics) | 62 | 3.72 | 1,200 | 16 | 3 | 48 | 240 | 3,100 | 12.9 | 64.6 | 37.5 | 0 | 82 |
| Fashion (seasonal) | 28 | 1.68 | 1,600 | 18 | 2 | 36 | 180 | 1,050 | 5.8 | 29.2 | 37.5 | 1 | 28 |
| Image (showpieces, capsules) | 10 | 0.6 | 3,500 | 4 | 1 | 4 | 20 | 171 | 8.6 | 42.9 | 37.5 | 0 | 4 |
| Total | 100 | 6 | 38 | 88 | 440 | 4,321 | 9.8 | 1 | 114 |
88 SKUs and 4321 units in total. The average depth is 5.8 units, above the minimum (3), but the minimum applies to every size: with the 8/22/30/25/15% curve an SKU needs at least 37.5 units (3 ÷ 8%). Fashion gets 29.2 units per SKU, so size XS gets 2.3 units, below the minimum: either cut fashion to 28 SKUs (instead of 36), or raise the role budget, or narrow the size run. The average depth without the curve hides this problem. Core gets the greatest depth because it repeats and sells more evenly; image keeps a narrow width at a high unit price. For a retailer the same matrix describes a set of brands: the role is attached to the brand and width is counted by styles per brand per unit of space.
Figures
- 2.3 … 19.4 units
Data as a table
| XS | S | M | L | XL | |
|---|---|---|---|---|---|
| Core (basics) | 5.2 | 14.2 | 19.4 | 16.1 | 9.7 |
| Fashion (seasonal) | 2.3 | 6.4 | 8.8 | 7.3 | 4.4 |
| Image (showpieces, capsules) | 3.4 | 9.4 | 12.9 | 10.7 | 6.4 |
Units per SKU by size at the size curve. The cell minimum for size × SKU is three units: at the rarest size (XS) the fashion role gets 2.3 units, below the minimum, so SKUs must be cut.
- Planned
- Supported by budget
Data as a table
| Measure | Planned | Supported by budget |
|---|---|---|
| Core (basics) | 48 | 82 |
| Fashion (seasonal) | 36 | 28 |
| Image (showpieces, capsules) | 4 | 4 |
How many SKUs are planned and how many the budget supports. If more SKUs are planned than the per-size minimum supports, part of the cells will be empty or single-size.
Specifics by category
- Apparel
- A color and size matrix (XS-XL or 40-56); core is held in a wide color range, fashion in 1-2 colors; the size curve is set by category and audience, the core sizes receive most of the depth.
- Footwear
- Width is counted by lasts and the size run: one last makes several styles, but each style multiplies by 8-12 sizes, so SKUs are many and depth per cell is low. The core sizes are critical (for example women 37-39); extreme sizes are bought in small lots or to order.
- Accessories
- There are no sizes (or one universal size), so width is counted by styles and colors and depth is simpler. The price ladder is more pronounced: image items (bags, belts) set the brand face; costume jewelry and hats cover the lower tier.
- Kids
- The matrix is by age and height, not by letters: the size grid goes in height steps, rapid growth shifts demand to larger sizes; a separate school capsule and seasonal waves; safety requirements (TR CU 007/2011) apply to each style.
- Men
- Narrower width, deeper stock: core classes (shirts, trousers, knitwear) repeat from season to season; the color palette is tighter; the size curve leans to M-L (check against your sales).
- Women
- The widest matrix: more colors, styles and price tiers, a higher fashion share; plus a wider size range (including plus size). A price ladder is needed per occasion (office, evening, casual).
Common mistakes
- Width without depth → many styles at 1-2 units per cell → sales do not build, turnover falls → check the cell minimum.
- Identical styles on different ladder tiers → cannibalization → each style gets its own function and tier.
- The core role is filled with fashion → unsold stock remains after the season → tie the core share to repeat-sale history.
- MOQ ignored → the matrix cannot be executed → check MOQ before approval.
- One size curve for all categories → a shortage of core and a surplus of extreme sizes → set the curve by category and audience.
Russian specifics
- Kids styles need conformity confirmation under TR CU 007/2011; without it the 10% VAT rate does not apply. Rule: CU Commission decision 797 of 23.09.2011: TR CU 007/2011 on safety of products for children and adolescents (Customs Union Commission Decision of 23.09.2011 No. 797) (rule check not completed).
- Clothing and shoes fall under TR CU 017/2011; composition and size information on the product is mandatory for the listing and marking. Rule: CU Commission decision 876 of 09.12.2011: TR CU 017/2011 on safety of light-industry products (Customs Union Commission Decision of 09.12.2011 No. 876) (rule check not completed).
- The Chestny ZNAK marking requires a GTIN for each size-color SKU: the matrix must be aligned with master data. Rule: Decree 1956 of 31.12.2019: Government Decree of 31.12.2019 No. 1956 (marking rules for light-industry goods) (rule check not completed).