Syntha Lab RU

Fit and sizing: anthropometry, ease, size runs, returns

The document links body measurements, patterns and size runs to money: what part of returns is caused by fit, which sizes customers need and in what proportion to buy them. Decision: size grid, ease and tolerances, size ratio in buying, size chart rules on the web shop.

Brand, customer, organisation Brand and retail Verified: 2026-11-14

Open interactively →

Why this document

The pattern maker, product manager and buyer read this when they set the size grid, ease allowances and the size split of an order. It matters because returns for size and fit are a direct loss: the item comes back, loses its look and often goes to a markdown. If the size split does not match your customers' bodies, some sizes sell out in the first week and others sit until season end. Body measurements and patterns let you fix this before purchase, not after the returns report.

At a glance

What is this document and why is it needed?
The document links body measurements, patterns and size runs to money: what part of returns is caused by fit, which sizes customers need and in what proportion to buy them. Decision: size grid, ease and tolerances, size ratio in buying, size chart rules on the web shop.
Who owns the document and how often is it rebuilt?
Maintained by the technologist or pattern maker with the product manager and returns analyst; approved by the head of product; reviewed on a last or pattern change and every season based on returns.
What mistakes are the most common?
A flat size ratio -> tails unsold, center short -> ratio by demand share and sales data. A brand-level size chart instead of product measurements -> the customer errs and returns -> a measurement chart per style. Tolerance wider than the grade step -> adjacent sizes indistinguishable -> tolerance no more than half the step.

For whom

Brand
For a brand, fit is built into construction: base patterns, grading, ease and tolerances in the measurement chart, fit sessions on a form and a live model. The brand owns a single sizing policy and a correct size chart for all channels.
Retail
For a retailer, fit shows up in returns and in size proportions: compare suppliers size grids, collect return data under the reason “size” and set size ratios for buying and replenishment.

What it consists of

1. Anthropometry

  • The target audience is described by bust, waist, hip, height and foot measurements; for mass assortment use a distribution from a real measurement base, not only standard tables.
  • The distribution gives demand share by size (see the example); different audiences (region, age, men, women, plus sizes) shift center and spread.

2. Ease and tolerances

  • Ease: the difference between garment and body measurement; depends on cut, fabric (stretch) and category; recorded in the measurement chart for each measuring point.
  • Tolerance per measuring point (for example ±1 cm for girth) checked at batch acceptance; a tolerance wider than the grade step makes sizes indistinguishable.

3. Fit sessions

  • Brand: The sample is tried on a form and a model in the base size, comments are recorded by point (length, shoulder, armhole, waist fit), the sample version and decision are logged in the tech pack.
  • Checking neighboring sizes (smaller and larger than base) for grading; criterion: no more than two comments on critical points before approval.

4. Size runs and vanity sizing

  • Size runs for clothing (bust, waist, hip; height), footwear (foot length, width; EU, RU, UK and conversion) and children (height and age, transitions between sizes).
  • Vanity sizing (understating the label as real measurements grow) makes tables of different brands incomparable: the only reliable guide is product measurements in the chart.

5. Link to returns

  • The return reason “size or fit” as a separate code; share by style, size and channel; a style above the norm goes to pattern or chart review.
  • Separate “too small” from “too big”: a one-sided shift points to grading or the chart, scatter both ways to inaccurate measurements.

6. Brand and retailer: who owns fit

  • Brand: Grading and tolerance: the brand answers for size to the wholesale client and the shopper; vanity sizing is a positioning decision, not an error, but it must be reflected in the garment measurement chart.
  • Retailer: Measurement charts of other brands differ: the retailer unifies them (garment measures, “how to measure”) and shows them on the card; for doubtful brands it flags “runs small/large” from return statistics.
  • Brand: Batch acceptance by measurements: measuring by points, tolerance, an act; pattern deviations are caught on the sample before cutting, not in returns.
  • Retailer: A buying size curve by store and channel: proportions come from demand (the example above), not flat across sizes; extreme sizes are concentrated in the DC and online.
  • Brand: The brand sees returns through wholesale claims and its own D2C data; without data from the retailer the “size” reason remains a guess.
  • Retailer: The retailer has direct data on returns by brand, size and reason: send them to the partner brand with “small” and “large” codes, which is grounds to revisit patterns or buying terms.
  • Brand: The brand size hint is built on its own measurement chart and D2C history; accuracy is checked against returns.
  • Retailer: The retailer size hint uses the customer’s purchase and return history across brands; data processing consents are needed and the result is checked against the “size” return share.

How to read the result and which decisions follow

  • Flat buying gives a sell-through (sold ÷ bought) of 82.8%: the two edge sizes get 200 units against demand of 117. Decision: buying by the demand curve moves stock from the edges to the core without raising total volume.
  • Brand: size returns of 7,971 a year with a 15% cut save RUB 0.80 m, most from “Women”. Decision: invest in garment measurements, grading and hints in the card where the size share of returns is high.
  • Retailer: for you the size curve is the order by stores and online warehouse; keep the edges in central stores, the core across the network. Decision: the size forecast is built from your own customers sales, not the brand table.
  • Brand: the fit tolerance of no more than half a step is a working assumption; test it on 3-5 figures and record it in the factory brief.
  • Retailer: an online size hint and in-store fitting reduce size returns; decide on introduction by the saving on returns in the table, not by the number of clicks.

Control checks

  • Check: the sum of size shares equals 100% (100.0%), buying by the curve equals 800 units (code).
  • Brand: take anthropometry and the size run for children from current data; the 6 cm height run is aligned with the kids document, do not invent group boundaries by the regulation.
  • Retailer: estimate the curve μ and σ from your own sales and update once a season; the brand curve shifts the core.
  • Brand: the cost of a return (RUB 650 and below in the example) includes logistics and handling; without unsellable losses it is understated.
  • Retailer: determine size return shares from the reasons chosen by the client and check them with a sample of returns.

Where the data comes from

Data sourceFields
Measurement chart and tech packBase size, grade step, ease and tolerances per point
Sales by sizeShare of sales by size, shortages and leftovers by size
Returns dataReason, size, style, channel, return cost
Measurement and survey dataAudience girths, height, foot; fit reviews

How to build it step by step

  1. Step 1. Describe the audience by measurements and find the center and spread of the size distribution.
  2. Step 2. Choose the base size and grade step; the step must not be narrower than tolerance (a working assumption: tolerance at most half a step, not a norm). For kids the 6 cm height run is the same across the whole range.
  3. Step 3. Fill the measurement chart: body measurement, ease, garment measurement, tolerance per point.
  4. Brand: Step 4. Run a fit session on the base and neighboring sizes and record versions.
  5. Step 5. Compute demand share by size and set the buying ratio.
  6. Retailer: Step 6. Match the web size chart to product measurements and add a “how to measure” hint.
  7. Step 7. Add the return reason “size” and separate “too small” from “too big”.
  8. Step 8. Each season find styles with size-return share above the norm and review patterns or the chart.
  9. Step 9. Estimate savings from fewer fit returns in money and compare with correction cost.
  10. Step 10. For footwear and children keep separate size systems and conversion rules.
  11. Brand: Step 11. Brand: fix the pattern version and the measurement chart in PLM, add batch acceptance by measurements and a comparison with the web chart.
  12. Retailer: Step 12. Retailer: unify brand measurement charts into one format, send brands size-return data and keep a list of brands with a persistent bias.

Formulas

MeasureFormulasNote
Demand share by sizeShare = Phi((interval top - mu) / sigma) - Phi((interval bottom - mu) / sigma); end sizes absorb the tailsmu and sigma are estimated from the audience measurement base; normality is checked.
Retailer: Buying size ratioBuy per size = total buy x size demand shareA flat ratio leaves the tails unsold and center sizes short.
Brand: EaseGarment measurement = body measurement + easeEase depends on cut and stretch; verified in a fit session.
Saving from fit returnsSaving = orders x return rate x “size” share x reduction x cost per returnThe reduction is an assumption until verified on a control group of styles.

Formulas in notation

Size share of demand
p_s = Φ((u_s − μ) / σ) − Φ((l_s − μ) / σ)
l_s and u_s are the bounds of the size interval, μ and σ the mean and spread of girth; the edge sizes take the tails.
Buying by size
Q_s = Q × p_s
Q is total buying; flat buying Q / n errs at the edges.
Size returns
R_s = Orders × r × φ
r is the return rate, φ the share of size returns.
Saving
Saving = R_s × δ × c_ret
δ is the cut in returns (15%), c_ret the cost of a return.

Worked example

Exercise: size returns and savings from a 15% cut by category of the web shop

CategoryOrders a yearReturnsOf them for sizeSize returnsReductionReturns preventedCost per return, RUBSavings, RUB m
Women22,00034%58%4,33815%6516500.42
Men8,00024%50%96015%1446500.09
Kids6,00026%55%85815%1295000.06
Footwear9,00030%65%1,75515%2638000.21
Accessories5,00012%10%6015%94000.00
Total50,00028.5%55.9%7,97115%1,1960.80

Notional: 50,000 orders a year; returns average 28.5%, of which 55.9% are for size. A 15% cut in size returns (a better size chart, patterns, a size hint) prevents 1,196 returns and gives RUB 0.80m a year. The largest saving is in Women: many orders and a high share of fit returns; in accessories size barely matters, so the effect is small and the return reason is sought in expectations and quality. Cost per return differs: footwear is heavier and costlier to process. All inputs are exercise assumptions.

Figures

Demand by size and flat buying
Bar chart: 84 85, 88 97, 92 140, 96 158, 100 140, 104 970501001502008484Demand by curve: 8585Flat buying: 1001008888Demand by curve: 9797Flat buying: 1001009292Demand by curve: 140140Flat buying: 1001009696Demand by curve: 158158Flat buying: 100100100100Demand by curve: 140140Flat buying: 100100104104Demand by curve: 9797Flat buying: 100100108108Demand by curve: 5252Flat buying: 100100112112Demand by curve: 3232Flat buying: 100100pcs
  • Demand by curve
  • Flat buying
Data as a table
MeasureDemand by curveFlat buying
8485100
8897100
92140100
96158100
100140100
10497100
10852100
11232100

Demand by size and flat buying. Buying equally over eight sizes gives an excess at the edges and a shortage in the core; sizes by chest girth, cm.

Saving from cutting size returns
Bar chart: Women 0.42, Men 0.09, Kids 0.06, Footwear 0.21, Accessories 000.10.20.30.40.5WomenWomenSaving: 0.420.42MenMenSaving: 0.090.09KidsKidsSaving: 0.060.06FootwearFootwearSaving: 0.210.21AccessoriesAccessoriesSaving: 00RUB m
Data as a table
MeasureSaving
Women0.42
Men0.09
Kids0.06
Footwear0.21
Accessories0

Saving from cutting size returns. The saving in RUB m when size returns fall 15%: footwear and women give almost all of it.

Returns and the size share in them by category
Scatter plot: 5 points10203040506070101520253035Women: 34; 58WomenMen: 24; 50MenKids: 26; 55KidsFootwear: 30; 65FootwearAccessories: 12; 10AccessoriesReturn rate, %↑ Share of size returns, %
Data as a table
Return rate, %Share of size returns, %
3458Women
2450Men
2655Kids
3065Footwear
1210Accessories

Returns and the size share in them by category. The higher the return rate and the size share, the greater the fitting potential: women and footwear are upper right, accessories lower left.

How to take measurements: woman. Measure over light underwear; keep the tape level and snug, not tight.

How to take measurements: man. Take girths at the fullest points; the inside leg runs from crotch to floor.

How to take measurements: child. Measure height barefoot against a wall; girths follow the same rules as for adults.

Dress: measurement points. Dress spec: bust, waist, hip, length, shoulders and hem sweep.

Specifics by category

Apparel
Size is set by chest, waist, hip girth and height; for a mass grid the distribution is taken as normal: at 96 ± 8 cm and a 4 cm step the central size takes about 20% of demand and the extreme ones 3-4% each. Ease depends on cut and fabric: a T-shirt 4-8 cm, a coat 12-20 cm (an assumption). A ±1 cm girth tolerance is a working assumption, not a norm; the grade step is not narrower than the tolerance, otherwise adjacent sizes are indistinguishable.
Footwear
Size is set by foot length: Mondopoint (mm, ISO 9407), EU (2/3 cm step, Paris point), UK and US with a different scale origin; they are matched through foot length in mm. The last is longer than the foot by a toe allowance (10-15 mm, an assumption). Width is a separate axis and a frequent reason for a “too tight” return; a half size splits demand between two SKUs, so it is introduced by data. Tolerance is set on last length in mm, not on the number; vanity sizing is weaker in footwear but brand scales differ.
Accessories
Most items have no size, so fit is adjustment and universality: a belt by waist girth (a 5 cm step, an assumption, adjusted through holes), a hat by head girth (a 1 cm step), gloves by palm girth, a ring by diameter. A bag has dimensions and capacity instead of size (for example A4 format or a laptop). “Does not fit” returns are few but exist for belts, rings and hats, so a measurement chart and “how to measure” are mandatory on the card.
Kids
A 6 cm height run (56 ... 176) is the same for toddlers and teens; age in the table is only a hint. Growth ease is half a size or one size, and tolerance depends on the item. For children under 1 year TR CU 007/2011 sets a height range up to 74 cm and chest girth up to 48 cm; teens (14-18) are closer to adults in proportions, and the run joins the adult grid by girths. A fit session uses a real model of the right group, not a scaled-down adult mannequin.
Men
The distribution is wider in height and waist; the grid is built on chest and height and often on sleeve length (for example heights 176 and 182). Fit (slim, regular, relaxed) diverges between brands more than the number; vanity sizing is weaker than in women. Size returns are lower, but when present they usually concern sleeve and trouser length, so return codes are split: “short in sleeve”, “long in length”. Ease: a shirt 8-12 cm on chest (an assumption), a suit cut is fixed.
Women
The grid is more complex: chest, waist and hip do not grow proportionally, so figure types and a three-point grade are needed. Vanity sizing shifts the number by 1-2 sizes at some brands (an assumption), so the table must rest on garment measurements. Fabric stretch changes ease: jersey needs less than suiting. Fit returns are higher than in men; a tolerance of at most half a step is a working assumption with no normative source.

Common mistakes

  • A flat size ratio -> tails unsold, center short -> ratio by demand share and sales data.
  • A brand-level size chart instead of product measurements -> the customer errs and returns -> a measurement chart per style.
  • Tolerance wider than the grade step -> adjacent sizes indistinguishable -> tolerance no more than half the step.
  • “Size” returns not split into “too small” and “too big” -> cause invisible -> two reason codes.
  • Footwear sizing mixed with apparel -> conversion errors -> separate systems and conversion tables.

Russian specifics

  • Garment size and labeling must meet the technical regulation for light industry products (TR CU 017/2011): size, composition and care instructions are stated. 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).
  • Children products have a separate regulation (TR CU 007/2011) with safety and age labeling requirements; children size runs are subject to it. 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).
  • Consumer rights on returns of properly made goods (size, style) are governed by the consumer protection law; check terms and exceptions against the current wording and sale rules. Rule: Law 2300-1 of 07.02.1992: Law of 07.02.1992 No. 2300-1 on consumer rights protection.
  • Brand: size and composition labeling on the item must match actual measures and composition; a mismatch is grounds for claims and inspection (light-industry labeling content requirements are unverified, check against the regulation).
  • Retailer: size information on the product card is part of the product information for the buyer; a wrong chart raises the risk of claims and returns, and responsibility for the description is shared with the supplier by contract (details unverified).

Open interactively →