How to Reduce Clothing Returns When Fit Changes After Washing
Fit returns on a core hoodie have climbed since the autumn drop, the size chart has not changed, and the Monday trading meeting wants to know whether the problem sits in the pattern, the fabric or the customer. It is a fair question, and the answer is often a mix. To reduce clothing returns of this kind you first have to find out how much of the fit problem is really a dimensional-stability problem that appears only after the garment has been laundered at home.
The scope here is knit casualwear, hoodies above all. The approved fit sample is measured before it has ever been washed, while the customer wears it after several wash cycles. Any gap between those two states is invisible to a buyer who checks only as-sewn measurements, yet it shows up in the returns file as "too small", "too short" or "not as expected". The deadline is bulk cut: after that you can only choose between a rework, a size chart change, a care label change or an accepted return rate.
All figures above are illustrative, built for this guide, and not benchmarks; replace them with your own returns data and wash tests. If your product is an oversized block, read it alongside our guide to oversized hoodie fit without increasing returns.
Step 1: Analyse Garment Returns by Size and Reason
Aggregate return rates hide the pattern that tells you what to fix. Cut returns by style, colour, size and reason code before you form a view, and keep the denominator consistent: this guide uses fit return rate = units returned with a fit reason code divided by units sold (net of cancellations) over the same period. Retailers define the denominator differently, so state yours in the report.
Reason codes are self-reported, so treat them as signals, not measurements. "Too small" can mean the garment shrank, the size chart misled the customer, or the wrong size was ordered. Separate these by reading the code against size, time owned and measurements of returned units.
Illustrative returns by size and reason (one hoodie style, 1,000 units sold)
| Size | Units sold | Returned: too small or short | Returned: too large | Returned: other reasons | Fit return rate |
|---|---|---|---|---|---|
| S | 150 | 6 | 9 | 5 | 10.0% |
| M | 350 | 28 | 14 | 8 | 12.0% |
| L | 300 | 33 | 10 | 7 | 14.3% |
| XL | 200 | 22 | 6 | 8 | 14.0% |
| Total | 1,000 | 89 | 39 | 28 | 12.8% |
Illustrative data, hypothetical and not a benchmark. Fit return rate = (too small or short + too large) divided by units sold. Total returns in the example are 156 units, of which 128 are fit returns.
Read it in three passes. First, "too small or short" accounts for 89 of the 128 fit returns, about 70 per cent, which points towards a dimensional or sizing cause rather than random customer choice. Second, the rate rises with size, which fits a garment that loses length and width in the wash. Third, the "too large" returns cluster in the smallest size, a different signal that belongs with the size curve, not with shrinkage.
Add a returned-garment audit to the routine: have the returns team measure a sample of fit-coded units against the approved specification, with the sample size agreed with your quality team. Measurements of real returned garments beat free-text comments.
Step 2: Test Wash Shrinkage and Spirality
A wash shrinkage test measures how much a garment changes dimensionally after laundering under stated conditions. It is a product test: it compares a garment against a method, and the method comes from a standards body such as ISO, AATCC or the BS EN series. Test methods are revised, so confirm the latest version and the method your programme requires with your lab. Do not confuse the test with a certification or with a claim. A "pre-shrunk" statement on a hangtag is a marketing claim until a test supports it.
How to run a wash shrinkage and spirality test
- 1Fix the conditionsMatch the care label: temperature, cycle, drying method and number of cycles.
- 2Mark and measure beforeMeasure every point on the specification and mark reference points.
- 3Wash and dryLaunder and dry as stated, repeating for the agreed number of cycles.
- 4Relax and measure afterRest the garment as the method states, then measure the same points.
- 5Report change and skewCalculate per cent change for every point, plus side seam displacement.
Confirm wash cycles, ballast and drying method with your lab and the care label.
Two calculations matter. Dimensional change per cent = (measurement after wash minus measurement before wash) divided by measurement before wash, multiplied by 100. A negative result is shrinkage. Spirality per cent = side seam displacement at the hem, relative to the matching point on the body, divided by the measured garment length, multiplied by 100. Some labs report skew with a different convention, so ask for the definition in the test report.
Test the points that drive fit complaints, and test them by size. Length measurements often move more than width in a knit, so include body and sleeve length. A single medium-size test cannot describe the whole size run, so test at least the smallest, middle and largest sizes in the first round. Our guide to stopping hoodie shrinkage, twisting and fit changes covers causes in the knit.
Fit Returns, Shrinkage and Size Exchanges: The KPIs Defined
Three KPIs carry this topic. Define each the same way in every report. The table gives the formula, the denominator and the way each can mislead.
The KPIs to track when you reduce clothing returns
| KPI | Formula | Denominator and variations | Illustrative reading | How it can mislead |
|---|---|---|---|---|
| Fit returns | Units returned with a fit reason code divided by units sold | Units sold net of cancellations over the same period; some retailers use units shipped or order lines | 128 fit returns on 1,000 units = 12.8% | Reason codes are self-reported; a size chart error and shrinkage look the same |
| Shrinkage | (Measurement before wash minus after wash) divided by before wash, times 100, per point of measure | Reported per point of measure, per size and per lot; state the wash conditions and cycle count | Body length 68.0 cm to 65.5 cm = 3.68% shrinkage | An average across points hides the one that breaks the fit; one lot hides lot variation |
| Size exchanges | Units exchanged for a different size divided by units sold | Count exchanges separately from refunds; split by direction (up or down) | 64 exchanges on 1,000 units = 6.4% | Exchanges can look cheap while still costing freight and handling twice |
Formulas are standard definitions; the example readings are illustrative inputs, not benchmarks. Definitions of returns and exchanges vary by retailer, so write down yours.
Read the KPIs as a chain. Shrinkage shows up in fit returns with a delay, because customers need time to wash the garment. Size exchanges move sooner when the size chart or the grade is the problem. Watching the three together tells you which end of the chain to repair.
Step 3: Set Tolerances and Graded Measurements That Survive the Wash
A tolerance is the allowed deviation from a specified measurement, and a graded measurement is the planned change in that measurement from one size to the next. Both are normally written for the as-sewn garment. The gap is that the customer experiences the after-wash garment, so a specification that is silent about laundering lets the fit you approved drift away without any single document being wrong.
The remedy is to carry two reference states in the tech pack for each critical point of measure: the as-sewn target and the after-wash target, with the expected movement between them and the tolerance on each. Our guide to making a clothing tech pack shows where those columns belong, and the grading and size chart guide explains how grade increments are built.
Illustrative size M specification with two states (hoodie, cm)
| Point of measure | As-sewn target | Illustrative tolerance | Illustrative grade step per size |
|---|---|---|---|
| Body length, high point shoulder to hem | 68.0 | ± 1.0 | 2.0 |
| Chest width, half measure | 56.0 | ± 1.0 | 2.5 |
| Sleeve length, centre back | 66.0 | ± 1.0 | 1.5 |
| Rib hem width, relaxed | 40.0 | ± 1.0 | 2.0 |
Illustrative values for one hypothetical block, chosen for the worked example. They are not recommended tolerances or grade increments; set yours for your fit, fabric and price point.
Compare the tolerance with the grade step before the wash result arrives. If a point of measure may legitimately sit one centimetre either side of target, and the grade step is only two centimetres, then adjacent sizes already overlap at the extremes. Add wash movement on top and the overlap grows. A useful rule is that the expected after-wash movement, plus the production tolerance, should stay clearly smaller than the grade step on the measurements that customers notice. If it does not, the options are tighter control, less shrinkage, or larger steps. Our UK size chart guide covers how the chart is presented to customers.
Worked Illustration: A Hoodie Loses Body Length After Three Wash Cycles
The example below is illustrative and hypothetical. It follows one size M hoodie from the specification above through three wash cycles at the care label conditions. The assumptions are: pre-wash measurements are exactly on target; the test uses the care label wash and drying method for three cycles; measurements are taken at the same marked points; and the grade steps are those in the specification table. No part of this is a real test result or a supplier outcome.
Illustrative pre and post-wash measurements, size M hoodie (cm)
| Point of measure | Before wash | After 3 washes | Change (cm) | Change (%) |
|---|---|---|---|---|
| Body length | 68.0 | 65.5 | -2.5 | -3.68% |
| Chest width, half | 56.0 | 55.0 | -1.0 | -1.79% |
| Sleeve length | 66.0 | 64.5 | -1.5 | -2.27% |
| Rib hem width, relaxed | 40.0 | 39.0 | -1.0 | -2.50% |
Illustrative arithmetic: change per cent = (after minus before) divided by before, times 100. Example: (65.5 minus 68.0) divided by 68.0 = -3.68%. Hypothetical values, not test results.
The chart below shows the percentages. The point to watch is the one where the loss is large relative to the grade step, which is body length.
Illustrative shrinkage by point of measure, size M hoodie after three washes
Shrinkage %
Illustrative values from the worked example, shown as shrinkage per cent (the absolute value of the change). Not a benchmark or typical result.
A 3.68 per cent change in body length looks modest in a report. In centimetres it is 2.5, which is 1.25 times the 2.0 cm grade step. The chest lost 1.0 cm, which is 0.4 of its 2.5 cm grade step. The garment has not shrunk evenly, so its proportions have changed as well as its size. That matters for a hoodie where the body length sets where the hem falls on the hip.
What the Result Means for the Size Curve
Translate the result into what the customer experiences. Someone orders size M from the size chart, expecting a 68.0 cm body length. After three washes the hoodie measures 65.5 cm. The unwashed size S on the same chart is 66.0 cm. The customer now owns a medium that is shorter than the chart small, which is exactly the situation that produces "too short" returns and a request to exchange for a large.
Body length by state against the size curve (illustrative, cm)
Target plus or minus tolerance
Illustrative bands: target plus or minus the example tolerance of 1.0 cm (S 66.0, M 68.0, L 70.0 as sewn; M after wash 65.5). Hypothetical data.
The after-wash band for M sits almost entirely below the as-sewn band for M and overlaps the small. If the published size chart is the only communication, the customer cannot know. There are four reasonable responses, and the right one depends on cost, the customer and the product.
Options for closing the gap, with the evidence each one needs
| Option | What changes | Main risk | Evidence needed before approving |
|---|---|---|---|
| Compensate in the pattern | Cut M longer so that it lands near 68.0 cm after washing. At the measured 3.68% shrinkage, the as-sewn body length would be about 70.6 cm | Only works if shrinkage is repeatable: the same cut gives about 69.5 cm at 2.1% shrinkage | Shrinkage results across several fabric lots, then a repeat wash test on the revised sample |
| Reduce the shrinkage | Change fabric finishing, relaxation or compaction, yarn or knit structure, or garment pre-treatment | May alter hand-feel, weight or cost; needs re-sampling | New wash test and a hand-feel and weight check on the revised fabric |
| Change the size chart or care label | Publish after-wash measurements, or specify a gentler wash and dry method | Customers may not follow care instructions | Customer wash behaviour and legal review of care instruction wording |
| Re-grade the size run | Increase the length grade or add a size so that after-wash positions stay distinct | Affects stock mix, ticketing and presentation | Size-level sell-through and fit return data |
Illustrative figures from the worked example. Pattern compensation arithmetic: 68.0 divided by (1 minus 0.0368) = 70.6 cm. Confirm any change with a revised sample.
Do not compensate in the pattern from a single test. If shrinkage moves between 2.1 and 3.7 per cent across lots, a fixed allowance cannot be right for all of them. Reduce and stabilise the shrinkage first, then compensate for what is left.
Step 4: Validate Bulk Consistency After Finishing
An approved sample proves a result is possible, not that bulk will repeat it. Shrinkage depends on fabric lot, knitting tension, dyeing, relaxation, compaction and any garment wash, and each can drift between lots. The final step is to make bulk prove its consistency before shipment.
A validation sequence from sample to shipment
- Before bulk cutAgree wash conditions, points of measure, sizes tested and buyer limits.
- Fabric lot arrivesTest shrinkage and spirality per fabric lot before cutting that lot.
- Pre-production sampleWash a sample made from bulk fabric and the bulk finishing route; measure all points.
- Inline and final inspectionCheck as-sewn dimensions and pull garments for wash testing. See our AQL inspection guide.
- Release or holdRelease, hold for retest, or escalate to a finishing review.
Illustrative sequence. Timing depends on fabric lead time, finishing route and shipping schedule; each programme must set its own.
The table shows a lot-by-lot read for the same size M specification. The limits are examples set only for this illustration: 3.0 per cent maximum body-length shrinkage and 2.5 per cent maximum spirality. They are not industry standards; agree your own with your supplier and lab.
Illustrative lot-by-lot validation, size M, three washes
| Fabric lot | Body-length shrinkage | Body length after wash (from 68.0 cm) | Spirality | Decision against example limits |
|---|---|---|---|---|
| Lot A | 2.1% | 66.6 cm | 1.2% | Release |
| Lot B | 2.4% | 66.4 cm | 1.5% | Release |
| Lot C | 3.7% | 65.5 cm | 2.9% | Hold: above both example limits; retest and review finishing |
| Lot D | 2.2% | 66.5 cm | 1.4% | Release |
Hypothetical lots and results for illustration. Example limits: 3.0% shrinkage, 2.5% spirality. Mean shrinkage across lots is 2.6%, but the range of 1.6 percentage points (about 1.1 cm of body length) is what matters.
Lot C passes on average and fails on its own: a mean of 2.6 per cent suggests comfort, while one lot in four is near the 3.68 per cent that caused the original problem. Request results per colour and per lot, as the finishing history may differ. For the fabric side of this discussion, see our guide to hoodie fabric: French terry versus fleece.
Common Mistakes When You Try to Reduce Clothing Returns
Most failed attempts to reduce clothing returns share a small set of errors. They are less about technical difficulty than about timing and ownership: tests are run too late, on the wrong sample, or against limits nobody agreed in advance.
Pre-bulk fit and shrinkage checklist
- Return reasons are coded by size, style, colour and lot, with one stated denominator.
- A sample of returned garments has been measured against the approved specification.
- The wash test uses the care label conditions and a stated number of cycles.
- Shrinkage and spirality are reported per point of measure, per size and per fabric lot.
- Acceptance limits were agreed with the supplier before testing, not after.
- Bulk wash testing uses the actual production fabric and finishing route.
A working checklist, not a standard. Add or remove items to suit your product and programme.
Testing only one size is the most common gap, since the largest sizes carry the highest return rates in the example above. Agreeing limits after results arrive turns a test into a negotiation. Relying on supplier results with no independent sample check removes the control that catches a lot like Lot C.
Decision Rules: What to Do and What Would Change the Decision
Use the evidence in sequence and keep decisions small enough to reverse. These rules suit a core hoodie programme.
- If fit returns are high but wash tests show small movement, treat it as a size chart or grade issue and review the curve, not the fabric.
- If wash movement is large on any critical measurement, address the cause before cutting more, then compensate in the pattern only for the stable remainder.
- If shrinkage varies widely between lots, tighten fabric lot control and test every lot before cutting. Do not rely on a pattern allowance.
- If exchanges rise but refunds stay flat, the size communication is the likely cause: check the chart and any fit notes.
Evidence that should change a decision includes a repeat test that contradicts the first, consistent results across several lots, or returned-garment measurements that do not match the lab result. Monitor fit returns, shrinkage per lot and size exchanges weekly during the launch period and monthly afterwards, with the owner for each named.
For a related read on when speed matters, see our guide to test and scale manufacturing, a risk-reduction model that lets you run a wash test on a small first quantity before scaling. One question to finish: if your best-selling hoodie passed its approval measurements today, which three points of measure would you want to see again after three home washes before the next reorder?
How Ruhrose Approaches Fit and Shrinkage
Ruhrose develops hoodies, jersey, denim and woven casualwear for UK buyers, with development and production in Bangladesh. We sample before bulk, and we are happy to build as-sewn and after-wash measurements into the development conversation so that wash conditions, points of measure and your own acceptance limits are agreed up front. We will say early if a target looks hard to hold. The practical starting point is our garment sampling process guide.