Syntha Lab RU

Collection and range

The collection plan in money, contribution of styles and the size curve.

For a brand Development and production

Open interactively →

About the topic

A collection in money means we look at the range not as an attractive display but as a set of investments, each with its own return. The topic helps decide three things: which models to keep in the collection, which to cut or drop, and what depth (how many pieces) to order for each. For a retailer a similar task is solved in the topic “Assortment and sizes”; here we look from the side of the brand that makes the collection and answers for batch size.

The reasoning has three steps. Step one: convert the plan from pieces to money. A grid of 30 models at 200 pieces looks even, but if one model costs RUB 1,500 and another RUB 6,000, their contribution to revenue differs fourfold. Step two: find which models give the main part of revenue and margin. This is ABC analysis: class A makes up roughly 70-80% of the result, B the next 15-20%, C the rest. Say that of 30 models six (20%) give RUB 63 million of 90, that is 70%, while the twelve weakest give only 9 million (10%). The tail need not be dropped, but each such model takes space at the factory, in cutting, in the warehouse and in your head.

Step three: check what exactly is counted. A class A model by turnover can be weak on margin if it sells at a discount or is hard to produce. So classification is done twice: by revenue and by margin. For stability one also looks at demand spread (ABC-XYZ): a stable model allows ordering deeper, an unstable one does not.

The size curve solves the second half: how to spread the batch across sizes. If a model of 120 pieces splits as S 15%, M 35%, L 35%, XL 15%, that is 18, 42, 42, 18 pieces; but the factory minimum may be 300, and then the model must be strengthened, colors merged or the model dropped. The result: a collection in money links the plan to the batch from the topic “Order volume and break-even”. Typical traps: counting only pieces, keeping a long tail “for variety”, taking a curve from one period when stock-outs of sizes distorted sales. Start with the last collection and see where the core and the tail are.

Key ideas

Money instead of pieces

Money changes even the order of models. Model A: 400 pcs at RUB 2,000 give RUB 800,000. Model B: 150 pcs at RUB 7,000 give RUB 1,050,000. In pieces A leads by almost three times (400 vs 150), in money B leads by RUB 250,000. So models are ranked by revenue and margin, not by pieces.

ABC: core and tail

ABC is built on the cumulative share: models are ranked by revenue from largest to smallest and added up. Revenue of five models, RUB million: 40, 25, 15, 12, 8, cumulatively 40, 65, 80, 92, 100%. The first three already give 80%, that is class A, and the last two account for 20%. The class boundary is set by convention, but the order and the cumulative share are calculated unambiguously.

Revenue and margin: double classification

A class A model by revenue can be weak on margin: sold at a discount or expensive to sew. So ABC by revenue and by margin is computed separately and compared. A model with RUB 4 million revenue and 18% margin brings less than one with RUB 2.5 million and 40% (RUB 720,000 against RUB 1,000,000).

Size curve

The curve gives shares, but a batch consists of whole pieces. For 130 pieces at 15/35/35/15% you get 19.5/45.5/45.5/19.5. Rounding to 20/45/45/20 keeps the total at 130, but the end size is now 15.4% and the middle one 34.6%. On small batches check the shares after rounding and add a piece where demand is confirmed.

Depth versus minimum batch

The choice is easy to compare through the expected loss on surplus stock. Plan 120 pcs, minimum 300: 180 extra. With a cost of RUB 1,000 and a leftover exit price of RUB 700, a loss of RUB 300 per piece gives 180 × 300 = RUB 54,000. The model margin on 120 pieces at RUB 900 is RUB 108,000. If the loss eats half the margin, merging colors or dropping the model pays better.

A trend is kept within a budget ceiling, not a count of SKUs

Of 48 articles 10 are trend ones: 20.8% by count, but 21.0% of the budget, RUB 3,024,000 out of RUB 14,400,000. The policy ceiling is 20%, the excess is 1.0 pp: the share by count looks normal, by money it does not. The trend life window is 26 weeks, delivery 10 weeks, 16 weeks are left for sales after arrival. So trend SKUs are taken as a small test batch, not “at full depth”, and their share is counted by money.

Formula: Sales window = trend life − delivery time; trend share = trend budget ÷ budget

What to do: Set the trend share ceiling in money and check it before the order, not after.

Think first: Trend articles 20.8%, ceiling 20%. Is the ceiling broken?

Answer: Look at the money: the trend budget is 21.0%, which is above the ceiling by 1.0 pp

A trend catch-up is checked against the ceiling, not only against test sales

The test sold 70 units at a threshold of 60, the signal grows by 4.3% a week, the last weeks are above the previous ones by 17.6%. The catch-up window is 9 weeks: the catch-up order is 108 units per SKU and 1,076 units in total for RUB 2,044,649. By sales the decision is “catch up”, but the trend share in the budget after the catch-up rises to 30.8% against a ceiling of 20%. A catch-up is decided twice: by sales and by what it does to the trend share and to the stock at the end of the window.

Formula: Share after the catch-up = (trend budget + catch-up) ÷ (budget + catch-up)

What to do: Catch up a trend only together with the decision on which other purchases to cut to hold the ceiling.

Think first: The test sold more than the threshold. So is the catch-up safe?

Answer: No: the trend share after the catch-up is 30.8% against the ceiling 20%; an answer is needed on what pays for it.

An assortment is a funnel of ideas, not a list

Not everything that is thought up gets into the collection: of 20 ideas 12 reach a tech pack, 6 a sample, 3.6 a pilot batch, 1.26 become hits. This is not a loss but selection: it costs RUB 1.64 million instead of RUB 6.83 million without gates. The assortment matrix wins when models that passed the criteria enter it, and the number of simultaneous pilots is limited by a limit per season.

What to do: Limit the number of pilots per season and put into the matrix only models that passed the gates.

Think first: The funnel’s spending is RUB 1,637,000 against RUB 6,830,000 without gates. How much do the gates save?

Answer: 6,830,000 − 1,637,000 = 5,193,000 ₽, about 76% of the “all ideas to the end” spending.

The price of developing a style falls with the replacement share and the number of styles

A digital sample per style costs RUB 2,800, a physical round RUB 7,500 at 3 rounds. It is cheaper if the replacement is not below 2,800 ÷ (3 × 7,500) = 12.4%. The license of RUB 240,000 a year at 2 seasons is recouped starting from 19.4 styles per season: for a small collection the digital will not pay back.

Formula: Replacement threshold d* = digital sample ÷ (rounds × the price of a physical sample)

What to do: For a collection smaller than the threshold in styles do not buy the digital, or take it as a service for the needed styles.

How to work through it

  1. Assemble the collection grid by category: models, colors, price, planned pieces.
  2. Convert the plan into money: revenue and margin per model.
  3. Run ABC by revenue and by margin and mark the differences.
  4. Review the structure: basics, fashion, image and their shares in money.
  5. Estimate development cost: samples, tech pack, revisions for each new model.
  6. Spread the batch across sizes with the curve and the factory minimum in mind.
  7. Check results against the last season: which assumptions did not hold.
  8. Decide what to cut, what to deepen, what to carry over from the last collection.

Where to start

Start with the collection plan in money, then ABC analysis shows the core and the tail, the size curve sets depth by size and the review of last season tests the assumptions. Calculations of the topic in order: “Collection plan in value”, “Cost of a sample and tech pack: development cost of a style”, “Collection mix: core, fashion, image”, “The trend share in the matrix and its life window: the share ceiling, the test batch, the chase order and the “rising / falling” signal”, “Payback of the digital sample (3D): savings on physical samples, weeks and the return period”, “Collection gates (stage-gate): a funnel of hypotheses, the price of a hit, the saving against “everything into a trial” and the stop rule of the pilot batch”. See also (their primary topic is another one): “ABC analysis of the range”, “Size curve: how to split a buy across sizes”, “Post-season review by category”, “Size curve by store cluster and pack composition”, “SKU classification: ABC and sell-through, bestsellers, slow sellers, no-sale”, “Categories: contribution to growth and margin, this year versus last”, “New and carryover: share, return per SKU, sell-through and margin”, “ABC-XYZ matrix: a SKU’s contribution to sales and demand stability”, “Choosing items under the buying budget and the shelf: a greedy set and an exact scan”, “The six-month plan and model stock: stock = sales × coverage + safety, purchases by month”.

Calculations in this topic

  • Collection plan in value

    From the collection grid (styles, depth, prices and cost) computes revenue, margin and leftover at a given sell-through and markdown, and checks that depth per style is not too small.

    What revenue, profit and leftover the grid yields at a given sell-through and markdown.

  • Cost of a sample and tech pack: development cost of a style

    Adds up the development cost of one style by stage (tech pack, patterns, prototype, fit sample, pre-production sample), calculates the development cost per launched and per successful style and shows what cutting the number of styles in development would give. Enter stage prices from contracts and invoices.

    What developing one style costs and what it costs per successful style if weak styles are dropped late.

  • Collection mix: core, fashion, image

    For collection roles (core, fashion, image or your own) calculates the share of the buy, planned revenue and margin, turnover and the money at risk in leftovers, and compares the shares with your target ranges and the margin with your target. Thresholds are editable: there is no single norm for the mix.

    How the collection is divided among core, fashion and image, how much margin each role brings and how much money it holds in leftovers.

  • The trend share in the matrix and its life window: the share ceiling, the test batch, the chase order and the “rising / falling” signal

    From the number of SKUs and the budget of the matrix, the trend SKUs and their budget it counts the shares and compares them with your ceiling; from the expected trend life, the lead time to the shelf, the test weeks and the reorder lead time it counts the selling window and the chase window; from the result of the test batch (sold, %) and the fading of the rate it determines the chase order in units and money; from a series of a signal index (queries, mentions, test sales) it determines the growth rate per week and shows whether the trend rises, falls or stands. It gives the decision: chase or not. The share ceiling, the life window and the test threshold are set by your policy, the calculation has no universal values; the default values are illustrative.

    What share of the matrix and the budget the trend takes against your ceiling, whether a chase order has time to sell and what the signal series shows: rising, falling or flat.

  • Payback of the digital sample (3D): savings on physical samples, weeks and the return period

    From the number of styles, the rounds of the physical sample, the price of a sample and the share of rounds that 3D replaces counts the savings on physical samples per season, the costs of the digital sample and the license, the weeks saved on the critical path and the payback of the one-off investment. Shows the threshold: at what replacement share the digital sample is cheaper than the physical one.

    How many physical sample rounds 3D will replace, what it costs and saves, how many weeks the critical path gains and after how many seasons the one-off investment pays back.

  • Collection gates (stage-gate): a funnel of hypotheses, the price of a hit, the saving against “everything into a trial” and the stop rule of the pilot batch

    From the number of ideas, the cost of the stages per model and the share of models passing each gate (idea, tech pack, sample, pilot batch) it counts how many models reach a hit, how much is spent in all, what one hit costs, what part of the spend went to rejected models and how much the gates save against “all ideas go to the end”. Separately it counts the stop rule of the pilot batch: the sales threshold at the control week, the loss on stopping and the loss avoided by not scaling blindly. Take the pass shares from past seasons; here illustrative.

    How many models will reach a hit, what one hit costs, how much the gates save and at what pilot result the batch is stopped.

  • ABC analysis of the range

    Splits styles or groups into A, B and C by contribution to revenue or margin: what feeds the business, the middle, and what can be cut.

    Which items give most revenue or margin and which can be cut.

  • Size curve: how to split a buy across sizes

    Uses sales by size to compute each size’s share and splits the quantity to buy across sizes. Compares with a flat split and shows how many units would end up in the wrong sizes.

    How to split the quantity to buy across sizes using historical sales.

  • Post-season review by category

    After the season shows what sold at full price, how much went on markdown, the margin that was left and how much money sits in leftovers. A basis for the next buy.

    What sold at full price, how much went on markdown, the margin and how much money sits in leftovers.

  • 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.

    How many pieces of which sizes to send to each store cluster and which pack to choose.

  • SKU classification: ABC and sell-through, bestsellers, slow sellers, no-sale

    Ranks SKUs by net sales, assigns class A, B or C by the cumulative share and compares it with sell-through. A bestseller is class A with a high sell-through, a slow seller is one with a low sell-through, “no-sale” is a SKU with no sales. You set the thresholds; the calculation shows the quartiles of your data as a hint.

    Which SKUs are bestsellers, which are slow sellers and which do not sell at all, and how much stock lies in them.

  • Categories: contribution to growth and margin, this year versus last

    Shows which categories raised or dropped sales and how many points of growth each gave, and splits the change in the total margin into a mix effect (category shares shifted) and a rate effect (margin within categories changed).

    Which categories gave growth and which lost margin and what changed the total margin.

  • New and carryover: share, return per SKU, sell-through and margin

    Compares groups of SKUs of a season, for example new items, carryover (carried over from past seasons) and basics: the share in the range and in sales, sales per SKU, sell-through and margin. Shows which group gives a return above its share in the range.

    Which group of SKUs (new, carryover, basics) returns more than its share in the range.

  • ABC-XYZ matrix: a SKU’s contribution to sales and demand stability

    From a table of SKU sales over several periods splits them into classes A, B, C by contribution to sales and X, Y, Z by demand stability (the coefficient of variation), shows the matrix of sales shares and the number of SKUs and suggests a stock policy for each SKU.

    Which SKUs bring sales, which of them are stable, and what stock policy to assign to them.

  • Choosing items under the buying budget and the shelf: a greedy set and an exact scan

    From a list of items (each taken whole) chooses the set with the highest expected margin under the buying budget and shelf limits: a greedy set by margin per unit of scarcity and the exact set by scan, the gap between them, the use of the budget and the shelf and the value of an extra 10% of budget.

    Which items to take whole so that expected margin is highest under the budget and shelf limits and how far the simple greedy layout falls short of the best.

  • The six-month plan and model stock: stock = sales × coverage + safety, purchases by month

    From sales by month, coverage (in months of sales, in steps by season), the display minimum (width × depth × the number of points) and a share of sales for the spread it counts the model stock at the start of each month. It takes the end stock equal to the model stock of the next month and counts the purchase of the month in units and at cost. It subtracts what is already ordered and shows what is left to place, over-ordering, the average stock, weeks of cover and turnover.

    How many units to buy and in which months so that there is a model stock before each month, how much can still be placed and whether there is over-ordering.

Self-check questions

  • What share of revenue and what share of margin falls on the core?
  • Are there class A models by turnover that are weak on margin?
  • What does the tail cost to keep: samples, cutting, warehouse?
  • What share do basics, fashion and image take and does it match the strategy?
  • What is the depth of each model and is it not below the minimum batch?
  • Where does the size curve come from and are stocked-out sizes accounted for?
  • What share of the collection is new and what share carryover and how do they sell?
  • What did the last season show: which assumptions proved wrong?

Knowledge check

Multiple-choice questions. The correct answer and the explanation follow each question.

  1. Trend budget RUB 3,024,000 out of RUB 14,400,000, ceiling 20%. By how much is the ceiling exceeded?

    • A. 0.0 pp
    • B. 1.0 pp
    • C. 20.8 pp
    • D. 20.0 pp

    Correct answer: B. 1.0 pp. 21.0% − 20% = 1.0 pp The share by count (20.8%) and by money differ, the ceiling is counted by money. Covered in: A trend is kept within a budget ceiling, not a count of SKUs.

  2. The trend lives 26 weeks, delivery 10, test 3, catch-up delivery 4. What is the catch-up window?

    • A. 16 wk
    • B. 13 wk
    • C. 9 wk
    • D. 26 wk

    Correct answer: C. 9 wk. 26 − 10 − 3 − 4 = 9: delivery, the test and the catch-up delivery are subtracted from the window. Covered in: A trend catch-up is checked against the ceiling, not only against test sales.

  3. Catch-up RUB 2,044,649, trend budget RUB 3,024,000, total RUB 14,400,000. What is the trend share after the catch-up?

    • A. 21.0%
    • B. 35.2%
    • C. 14.2%
    • D. 30.8%

    Correct answer: D. 30.8%. The catch-up increases both the numerator and the total budget: (RUB 3,024,000 + RUB 2,044,649) ÷ (RUB 14,400,000 + RUB 2,044,649) = 30.8%. Dividing by the old budget (35.2%) is wrong. Covered in: A trend catch-up is checked against the ceiling, not only against test sales.

  4. The funnel’s spending is RUB 1,637,000, the spending “all ideas to the end” RUB 6,830,000. How much do the gates save?

    • A. 5,193,000 ₽
    • B. 1,637,000 ₽
    • C. 6,830,000 ₽
    • D. 8,467,000 ₽

    Correct answer: A. 5,193,000 ₽. 6,830,000 − 1,637,000 = 5,193,000 ₽. 8,467,000 ₽ comes if the spending is added instead of subtracted. Covered in: An assortment is a funnel of ideas, not a list.

  5. What share of the funnel’s spending goes to models that did not become hits?

    • A. 26%
    • B. 50%
    • C. 74%
    • D. 98%

    Correct answer: C. 74%. A hit passes all gates and costs 4,000 + 25,000 + 55,000 + 257,500 = RUB 341,500; 1.26 hits give RUB 430,290 of RUB 1,637,000, the rest RUB 1,206,710, that is 74%. This is the normal price of selection. Covered in: An assortment is a funnel of ideas, not a list.

  6. A digital sample of RUB 2,800 per style, 3 physical sample rounds at RUB 7,500. At what replacement share is the digital cheaper (without the license)?

    • A. from 37.3%
    • B. from 12.4%
    • C. from 8.3%
    • D. from 40%

    Correct answer: B. from 12.4%. d* = 2,800 ÷ (3 × 7,500) = 12.4%. 37.3% results if one round is counted instead of three. Covered in: The price of developing a style falls with the replacement share and the number of styles.

Common mistakes

  • Mistake: Depth below the minimum batch

    What it costs: The factory refuses the order or the unit price rises

    How to avoid: Plan with the minimum in mind and merge colors.

  • Mistake: Size curve from a period of stock-outs

    What it costs: Scarce sizes are under-ordered again

    How to avoid: Reconstruct demand, do not take sales as is.

  • Mistake: Classifying by revenue when margin is needed

    What it costs: An A model by turnover turns out unprofitable

    How to avoid: Compute both classifications.

  • Mistake: Planning in pieces without converting to money

    What it costs: Expensive models get lost in the mass

    How to avoid: Always count in money too.

  • Mistake: Keeping a long tail “for variety”

    What it costs: Costs grow and attention scatters

    How to avoid: Limit the number of models and review the tail each season.

  • Mistake: A trend is caught up by test sales

    What it costs: The trend share breaks the ceiling Money in a trend with a short window: 2,044,649 ₽ at a catch-up window of 9 weeks.

    How to avoid: Stock at the end of the window → check the catch-up by the share and the window, not only by the test Before a catch-up count the share after it and the window; pay for the catch-up by cutting other purchases.

    What it looks like (illustrative figures): The test sold 70 units, the catch-up of RUB 2,044,649 raises the trend share from 21.0% to 30.8%.

Russia and the EAEU

For a Russian brand the collection is planned around the calendar: seasons, school, holidays, trade-show pre-orders and shipment deadlines to chains. Each model in Russia needs a conformity confirmation (certificate or declaration, with higher requirements for kids and underwear) and Chestny ZNAK marking: each item has its own GTIN, and codes are ordered for the run. These costs are fixed per model, so a long tail of small batches costs more than it seems.

Samples and the tech pack are part of development cost; they must be spread across the run, not written off “in general”. Buyer prices are quoted with 22% VAT or without it depending on the regime: check against the contract. For children’s items on the list a 10% rate is possible with confirmation; check the item.

When planning depth consider the requirements of chains and marketplaces: they set size runs, packaging, acceptance deadlines and penalties for short delivery, which affects which models make sense to produce. Take platform tariffs from current terms. Basic models that sell all year lower the risk because leftovers pass to the next season without a discount.

The assortment for EAEU countries depends on climate, tradition and sizing habits, not on taxes. Indicative notes, verify with local partners: in Kazakhstan and Kyrgyzstan the share of warm outerwear in winter and of national festive looks (Nauryz, Kurban Bayram) is higher, in Armenia and Belarus seasonality is closer to the European one. Size grids are mostly RU or EU, but in Central Asia and Armenia Turkish and Chinese ones occur; a conversion table is needed. For Kazakhstan light-industry marking starts on 1 December 2026 (indicative), so check depth per model against the cost of codes; for the other countries the dates: indicative as of 14.11.2026, check the text of the act.

Categories and audiences

Apparel
On apparel: the widest grid and the longest tail, so ABC yields the most here. The balance between basics that sell all year and fashion that lives a season determines leftovers. The size curve runs across three to five sizes, and the extremes cost more when scarce.
Footwear
In footwear: the grid is narrower than in apparel, but each model unfolds into 8-12 sizes, and a mistake in the curve is costly because a pair cannot be combined with another size. The minimum batch for tooling is higher. Check the curve by half sizes and rely on the core of middle sizes.
Accessories
For accessories: a size curve is hardly needed, but color and price structure matter. Models are cheap, so counting revenue of a single item in ABC makes little sense; look at category contribution and the image role (bags and scarves draw people to the brand). One-size and small depth per color work well.
Kids
If it is children’s wear: the size grid runs by height in 6 cm steps, and a child quickly moves to the next size, so the curve shifts through the season: autumn has more large sizes for school. The grid covers boys and girls, and ABC is computed separately. Certification and marking fix a cost per model.
Men
In the men’s segment the size curve sits tightly in the core (for example M–XL), so one proven curve can be applied to several models of a category and depth can be set per category rather than for each model separately. Cut the tail harder: the buyer looks for a familiar item, not variety, and an extra model only takes space at the factory and in the warehouse.
Women
In a women’s line the size curve depends on fit: a fitted model has higher demand at the end sizes than a loose one, so the curve is taken by fit type rather than for the whole category. Set the shares of “basics, novelties, image” in advance, otherwise novelties will push basics out of the plan, and sizes and colors will scatter across minimum batches.

Questions this topic answers

  • What are the revenue and margin of our collection grid?
  • Which styles bring most of the revenue?
  • How to split the run by size?
  • What share of revenue comes from the core and what to do with the tail?

Key measures and terms

Study first

What next

Next step: “Range: width, depth and price tiers”. Related: “Channels: wholesale, retail, platforms”, “Season results: brand, category, SKU”, “Assortment and sizes”.

Documents on this topic

Open interactively →