Imagine a buyer in Raleigh lifting a freshly washed event shirt off a drying rack. The print has curled at one corner, and the hem sits higher than it did at the artwork review. The purchase order covers 600 shirts, with screen printing due to start Friday. The vendor asks whether to reserve the press. Before the buyer answers, she needs to know whether the finished shirt changed size beyond the agreed limit, whether the print failed its own appearance check, and whether the sample was washed the way recipients will be told to wash it.
There is no single shrinkage percentage a buyer can use to accept every branded shirt. A usable result comes from the exact finished garment, a recorded wash and dry procedure, and limits agreed before anyone sees the results. Set separate criteria for body length, width, and appearance. If the first cycle was run incorrectly, treat it as invalid for the agreed acceptance decision. If the measurements pass but the logo lifts, hold the decoration approval while you investigate. That distinction lets you release only what the evidence supports.
Specify the sample before discussing a percentage
Ask the supplier to identify the base garment, size, color, fabric lot if available, and decoration process. Get a physical preproduction sample that represents the product people will actually receive. If a logo will be screen printed, a blank sample or a digital proof cannot show how the cured print behaves after laundering. If the proposed run uses embroidery, test the intended thread, backing, location, and garment together. Keep a matching unwashed sample when possible so you have something to compare with the washed piece. A blank from the same lot can help investigate what changed, but the decorated garment is the one being considered for approval.
Write down how you will select samples before testing. A purchase order may include several sizes, colors, base-garment lots, or decoration setups. Decide which combinations need representation and why. Three tagged samples may be practical for a small screening exercise, but three shirts chosen by convenience cannot establish a defect rate for 600 pieces. If different factories, fabrics, or print processes feed the same order, document which configurations the sample represents. Take additional samples where a difference could affect the outcome, or record the scope still waiting for evidence.
Agree on the disposition rule at the same time. For instance, you might specify a maximum loss of four percent at an agreed length measurement and three percent at an agreed width measurement, with both limits applied to each tested shirt. Those numbers are hypothetical order terms, and a buyer could select different limits based on fit, supplier capability, and the intended use. If your rule uses averages, specify whether an individual limit also applies. Decide what happens when one result fails, when an observation is close to a limit, and when the sample cannot be traced to production. The same recorded rule should govern both parties.
Ask the vendor who has authority to approve the wash result and who can release production. In some programs, artwork and sourcing are approved by different people. Keep approval of a digital proof separate from approval of a physical shirt, and keep both separate from permission to decorate the bulk lot. That way, a quick acknowledgment of the logo position does not become an accidental approval of the garment’s wash performance. Put the approval names and the exact event deadline next to the decision rule so a late exception reaches the right person.
Choose the wash and measurement method
Start with the intended care instructions. In the United States, the FTC Care Labeling Rule requires a reasonable basis for care information for covered manufacturers and importers. The rule is useful context for checking what you tell recipients. It does not set a universal purchasing tolerance for a branded shirt. If a label is provisional or inconsistent with the supplier’s proposed wash procedure, resolve that difference before the test. The people wearing the shirts should be given care instructions supported by the actual product evidence.
For an agreed lab route, AATCC TM150 addresses length and width changes in garments after home laundering and notes a limitation for certain stretch fabrics. ISO 5077 covers determining dimensional change after specified washing and drying procedures. ISO 3759 covers preparing, marking, and measuring textiles and garments for dimensional-change tests. Choose the method and edition with your supplier or lab and confirm which procedure and conditions are in scope. Public summaries describe what these methods cover. They do not supply every instruction a lab needs, or establish a pass criterion for your order.
An internal screening test can still help a buyer catch a problem early. Call it an internal screen, and record the equipment, settings and limits that were actually used. If the order requires a named test method, ask the lab to follow its full procedure and record any deviations. Do not describe a home machine test as formal compliance with a method just because you used a similar temperature. Specify the method’s revision, cycle count and reporting needs before sending samples out, especially if two labs or vendors will compare results.
Choose the wash and dry settings as a pair. Record the machine, selected cycle, temperature or setting, load or ballast, detergent, and drying method, following the agreed procedure. Write down when each cycle started and ended and the number completed. A flat dry and a high heat tumble dry are different exposures. If the purchase decision is based on one intended care route, a result from another route can provide investigative information, but it does not replace the required test. If the program expects industrial laundry, establish a relevant procedure with the parties rather than assuming a domestic wash represents it.
Define the measurement conditions as precisely as the wash. Use the same flat surface and agreed tool before and after. Lay the shirt in a consistent position without stretching it or pressing it into its original shape. Agree on when measurements can be taken after the garment has dried and how it will rest, or follow the conditioning prescribed by the chosen method. Keep the before and after measurements under comparable conditions. If you want immediate post-dry observations for troubleshooting, label them separately from the acceptance measurement.
For an internal protocol, draw the points of measure on a copy of the supplier’s diagram. Body length might run from the high point of the shoulder to the bottom hem along a defined line. Width might be the flat distance between two points at an agreed height below the armhole. These are examples of points to define, rather than instructions for a particular formal method. Note how to handle a ribbed hem, an uneven neckline, or a curved edge. Two operators should be able to find the same points from your drawing and obtain comparable readings without discussing which number they expect.
| Worksheet field | Record before washing | Record after washing | Use in the decision |
|---|---|---|---|
| Sample identity | Sample ID, size, color, base style, lot, decoration method and proof version | Confirm the same tag and any loss of traceability | Limit the approval to the represented configuration |
| Length and width | Point of measure, units, tool and baseline reading at each point | Same points and units, actual reading and operator | Calculate signed change for each point and compare with its agreed limit |
| Wash and drying | Agreed cycle, temperature, detergent, load, drying route, cycle count and measurement timing | Actual settings, time, conditioning and any deviation | Decide whether the run is valid before scoring it |
| Logo and overall appearance | Flat photos with scale, close-up of decoration and agreed appearance criteria | Repeat photos and record cracking, lift, curling or other observed change | Accept or hold decoration independently of dimensional change |
| Seam alignment | Reference position and the agreed way to assess it | Direction and extent of visible movement under the same setup | Assess twist separately and request an agreed method when quantification matters |
Do the arithmetic on each garment and each point
Use signed percentage change. Subtract the before reading from the after reading, divide by the before reading, and multiply by 100. Both readings must use the same unit. A negative result means the measured point got smaller. A positive result means it got larger. If a supplier reports positive shrinkage instead, identify that convention on the worksheet so a minus sign does not turn a failing result into a passing one. Keep the raw readings as well as the calculated percentage. A missing or zero baseline cannot produce a valid percentage for that point.
A body length of 60 centimeters before washing and 57 centimeters afterward gives a change of minus 3 centimeters. Dividing by 60 and multiplying by 100 gives minus 5 percent. With an agreed maximum length loss of 4 percent for this example, the sample fails on length. If a separate width measurement goes from 50 to 49 centimeters, that gives minus 2 percent. It meets the example width limit of 3 percent. Passing width does not change the length decision. The workbook or inspection record should show both readings, both limits and the separate dispositions.
Avoid rounding toward a pass. If the agreed method requires a particular precision or rounding rule, apply that consistently. When a reading sits near the boundary, follow the agreed remeasurement rule and preserve both observations. For example, a second operator may confirm the same points and tool without seeing the first value. Decide in advance whether a sample can be remeasured after a timing or alignment issue and when a new sample is required. Do not rerun a valid failure repeatedly until a garment happens to pass.
Use the same comparison logic across the sample set. Suppose three decorated shirts with valid measurements show length changes of minus 2, minus 3 and minus 6 percent. Their mean loss is about 3.67 percent, yet the last one fails a 4 percent individual limit. Keep the results on individual rows and decide what the failure means for the lot or configuration the shirt represents. If the purchase agreement only states an average, clarify what a single outlier triggers before testing. A comparison of vendors is only meaningful if their results came from comparable procedures and pre-agreed rules.
Worked decision one: The garment meets dimensions and the print still fails
Return to the Raleigh order. Assume three production-representative printed shirts were washed following the agreed label route, and each remains within the hypothetical length and width limits. One has a curled edge at the logo after the agreed rest period. A matching unwashed reference is flat. The purchase agreement requires the logo edge to stay flat under the agreed visual check, so the appearance result fails even though the measurements pass. The buyer keeps the dimensional results, holds the release for that decoration, and asks for a controlled correction.
Compare the failed print with the unwashed decorated reference and, if one is available, a washed blank from the same garment lot. The comparison may help narrow the inquiry. It cannot prove that ink, curing or fabric caused the problem by itself. Ask the supplier to identify the likely process difference and confirm the next production-representative proof. Keep the prior failed sample tagged and photograph both versions using the same setup. If the proposed correction changes the print system or base fabric, reassess which dimensional and appearance evidence needs to be repeated.
The buyer can ask whether the press time can be reserved without authorizing decoration. If another already-approved design is available, it can be considered separately with its own sample and order evidence. The approval message for the corrected shirt should name the artwork and garment identifiers, the proof revision, the wash procedure, and the passed criteria. Until that evidence exists, a deadline or a passed dimension should not stand in for the unresolved decoration decision.
Worked decision two: The wash procedure was wrong
A second supplier sends a test result showing that length went from 60 to 57 centimeters. The vendor used a high heat dryer, while the agreed acceptance plan specified a different drying procedure. The minus 5 percent calculation is accurate for what happened to that shirt. It is still an invalid run for the contract decision. Record the deviation and keep the result for investigation. Take an unused matching sample with a valid baseline and rerun the agreed procedure. Rewashing the already altered shirt would not reproduce the initial test conditions.
Assume the valid rerun on a fresh sample measures 60.0 to 58.2 centimeters, a 3 percent loss. That meets the example length limit of 4 percent. The buyer must still apply the agreed rule to the other required measurements, samples and appearance criteria. If another valid sample fails, keep that failure visible and follow the disposition agreed in advance. If the approved procedure cannot be reproduced before the deadline, keep the item on hold or record a specifically approved scope change. A new schedule is a business decision that should be made openly.
If both supplier reports arrive without a machine program, dry route or cycle count, ask for the original logs before comparing the vendors. If the logs cannot be recovered, use a common test procedure on new traceable samples. Do not infer that the report with the smaller percentage describes a more stable garment. Even the same shirt can show a different result under different care conditions. The purpose of the repeat is to get comparable evidence, rather than to recreate a number already in a spreadsheet.
Keep twist, surface changes and fit in their own rows
A body may measure within its length and width limits and still have a seam that shifts toward the front. Record side seam movement or visible twist separately, using the agreed reference position, direction and method. If the buyer requires a quantified method, review scope and reporting with a lab. AATCC TM179 addresses changes in skew after home laundering. Ask whether that specific procedure fits the article and decision. A photograph of a twisted shirt can justify an appearance hold under the agreed criteria, but should not be labeled a numerical result from a method you did not follow.
Appearance can include print cracking, edges lifting, puckering, color change or a hem that no longer hangs as intended. Define the acceptance view before the test. Take a full-garment image, a close-up with a scale and, when helpful, a consistent angled view. Retain the same lighting and garment orientation. Ask whether the defect is visible under ordinary handling and whether that is how your approval standard defines it. Keep the original images and label the sample ID rather than comparing two images with ambiguous filenames.
A fit review can also be useful, especially where a tight fit or stretch fabric makes a small dimensional change material. Use that review to inform the original purchase criteria or a formally agreed revision. Do not silently replace the dimensional test because one person finds a shirt comfortable. If stretch recovery, repeated laundering or an unusual fabric is central to the decision, ask your lab which procedures are suitable, and have the supplier produce a sample that matches the final construction. A single internal cycle is a limited observation.
Put the evidence in the purchase order and the release note
Use an attachment or order specification to describe the sample selection, wash and dry protocol, measurement diagram, criteria by point, appearance checks, and handling of invalid runs. Attach the signed sample approval to the exact design and production scope. State who can approve a changed fabric, process, or factory. If a substitute lot arrives, compare it with the approved sample and decide which checks must be repeated. Keep records with their sample IDs, not just the latest version of a photo in a message thread.
The release note can be short. Say what lot and decoration were tested, which valid cycles were completed, whether each acceptance criterion passed, and which exceptions remain open. Attach the raw worksheet and photos. If the team releases part of the order, identify the specific configurations released and held. Include the next inspection point and what evidence would allow the held work to proceed. This protects the timeline better than a general instruction to proceed while someone is still chasing down a discrepancy.
If you are sourcing the shirt through Giftpack, its custom merchandise guidance describes sample or proof approval where appropriate and quality checks against agreed requirements. Use that discussion to request a production-representative decorated sample, confirm who will run the wash test, and record the criteria that must be met before bulk work. In Raleigh, that means the buyer can reserve time if the supplier allows it while keeping decoration approval tied to the corrected, tested shirt. The decision travels with the sample and evidence all the way to the order.