Define the problem correctly before you write the PO
For retail buyers, teddy fleece shade control is a presentation-control problem first and a dyehouse problem second. Flat fleece programmes can often run on swatch approval plus routine final inspection. 320gsm teddy fleece does not behave that way because the lofty pile changes reflectance with viewing angle, nap direction, brushing intensity, fold set and compression during packing. A blanket can sit inside instrumental tolerance and still look wrong beside another blanket on shelf.
That distinction matters because factory controls and buyer controls are different. The mill controls greige allocation, knitting consistency, disperse dyeing discipline, raising, shearing, heat-setting, lot coding and pack orientation. The buyer controls approval hierarchy, acceptable lot count per colour per PO, segregation through the DC, store allocation logic and concession rules. If those responsibilities stay vague, the programme usually ends in subjective claims after goods land.
Treat shade approval as a two-layer system. Layer one is instrumental lot-to-standard control under an agreed method. Layer two is visual appearance approval on the finished blanket, because pile direction and surface geometry can override the instrument reading in real presentation. Keep the article self-sufficient in the PO: do not rely on a separate generic colour policy that the supplier has not signed against.
Distinguish shade variation, appearance variation and physical defects
Buyers should separate three failure types because the commercial remedy is different. Shade variation means a genuine colour movement against the approved standard, measured instrumentally and confirmed visually. Appearance variation means the perceived colour changes because of pile direction, shearing level, compression whitening, opposite fold direction or face-out pack orientation, even if the underlying dye lot is close. Physical defects include barre, rope marks, crease marks, centre-to-edge difference from processing, lean, patchy shearing or finishing streaks.
This distinction matters in chargeback language. If a store complaint is written only as 'shade issue', the supplier may argue instrumental pass and deny liability. If the defect is actually appearance variation caused by mixed pack orientation or vacuum compression, the evidence and corrective action are different. If the defect is barre or rope marking, that is not a normal lot-to-lot tolerance issue at all; it is a process defect and should be treated as hold or reject depending on severity and prevalence.
For teddy fleece, buyers should define in the PO that visible barre, rope marks, severe side-to-centre appearance shift, patchy shearing, obvious nap reversal within the same retail unit and compression whitening that does not recover after a reasonable relaxation period are not acceptable as normal shade variation and are outside AQL treatment. Routine AQL such as 2.5 may cover minor sewing defects, but these appearance-critical issues should be listed as critical shipment holds rather than left to inspector judgment. AQL guidance for broader blanket inspection is covered in [blanket-quality-control-inspection](/blog/blanket-quality-control-inspection.html).
What actually creates mixed-looking teddy fleece blankets
The usual causes are a combination of greige variation, dye uptake spread, pile geometry and finishing variation. On polyester teddy fleece, disperse dyeing is commonly run at high temperature, often around 125-135C, but the workable window depends on machine type, liquor ratio, construction, depth of shade and recipe discipline. Jet and overflow systems can behave differently on circulation marks, bulk and levelness. Buyers should not write a spec around one fixed dye temperature; they should write controls around lot discipline and release criteria.
GSM variation changes perceived colour mainly through surface geometry, but it can also influence dye result indirectly. If one lot runs bulkier because of knitting variation, raising intensity or heat history, the eye may read it lighter, milkier or duller. At the same time, those construction shifts can alter dye uptake and migration behaviour slightly. The accurate buyer wording is that construction variation can change both perceived shade and, to a lesser degree, the dyed result itself.
Dark navy, charcoal, bottle green, taupe and warm greige usually expose the risk fastest. Slight red-blue or yellow-green movement becomes obvious when folded units are stacked side by side. On teddy fleece, additional brushing, uneven shearing, rope marking, barre, centre-to-edge variation, panel lean and inconsistent heat-setting can all create what stores report as a shade complaint.
Teddy-fleece-specific packing and display failures are often missed in factory release. Opposite fold direction can reverse nap reflection so two blankets from the same lot look different face-out. Tight belly bands or vacuum packing can create compression whitening on the exposed panel, especially on lighter shades and high-loft constructions, and recovery may take 24-72 hours depending on pile resilience. Mixed face-out orientation in retail packs can also make acceptable lots look mismatched because one group presents with nap running up and the other down. These are appearance failures, not necessarily dye failures, but they still create shelf claims.
Use a short pre-production trial to lock realistic tolerances
Do not lock shade tolerance from lab dip only. For piece dyed teddy fleece, create tolerance bands from a pilot bulk trial in the real construction and finish. A sensible framework is one pilot dye lot per key colour family, finished through the intended raising, shearing and heat-setting route, then assessed both instrumentally and visually on finished blankets rather than on small hengers alone.
Before bulk, lock at least four items in writing: approved construction range including target GSM and acceptable shipment range; finishing recipe window including raising and shearing settings that materially affect surface appearance; approved standard blanket sealed by buyer QA; and visual examples of acceptable light-side and dark-side limits, ideally as a shade-band set. If the programme is reversible or has two display faces, lock face A and face B appearance separately.
Tolerances should be shade-family specific. Pale neutrals, warm greiges and cool greys may need tighter visual control than the same numerical Delta E band suggests. Deep navy and charcoal often require a lower maximum single-point deviation than medium shades because a local panel shift is more visible after folding. The pilot should therefore produce data by colour family, not one blanket tolerance for every shade in the range.
A practical pre-production release pack should include the sealed standard blanket, corresponding bulk hanger, instrument report, agreed pile orientation for measurement, approved retail fold photos and one-page decision table naming who can release, hold or concede at each stage. That avoids re-arguing method after bulk is already dyed.
Set an approval hierarchy with clear decision rights
Do not rely on one lab dip plus a Delta E line item. Teddy fleece needs a clear approval ladder. A practical hierarchy is: mill lab approves recipe match to the signed lab dip and pilot bulk standard; factory QA has authority to hold any lot that fails instrumental method, shows visual mismatch, or shows appearance defects such as barre, rope marks or severe compression whitening; buyer QA or buyer merchandising, depending on the account structure, has final visual release authority on finished blankets; written concessions can only be authorised by the buyer's named approver in the PO or quality appendix.
The sealed finished blanket should have strong authority, but not unlimited authority. A defensible release rule is: instrumental pass is necessary but not sufficient; visual pass on finished blanket under agreed viewing conditions is also necessary. If visual pass conflicts with instrumental fail, hold the lot and run joint review against the sealed finished blanket, retained bulk hanger and instrument data. If the lot is visually acceptable in side-by-side retail presentation and the instrumental deviation is marginal and repeatable, the buyer may issue a written concession. If visual fail conflicts with instrumental pass, visual fail governs for shipment release because the retail risk is real.
Do not use one universal Delta E limit across all shades. A workable starting point is CMC(2:1) or Delta E 2000, but the band should be shade-family dependent and confirmed in pre-production trials. Pale shades may hold around CMC(2:1) 0.8-1.0, medium shades around 1.0-1.2, while dark navy, charcoal or red-black families often need a combination of instrumental limits and visual shade-band review because a technically passing value can still read different on pile.
Retain not just one sealed blanket but a sealed shade-band set where feasible: target, acceptable light-side limit, acceptable dark-side limit and known reject example. Final decision rights should be explicit: mill lab for recipe sign-off, factory QA for internal lot hold, buyer QA or merchandising for visual release, and only the buyer's designated signatory for written concession on shipment. Without named authority, goods tend to move on email opinions rather than on a controlled release rule.
Specify instrumental measurement for pile fabrics as a usable SOP
Measurement language for teddy fleece needs more detail than 'three representative areas, average reported'. For pile fabrics, specify conditioning, backing, pile preparation, reading geometry and how many points are allowed to fail. Pile fabrics may be measured with SCI or SCE depending on the buyer's standard, but the selected mode must be fixed for the programme and stated on the approval record. Do not imply that one mode is universally correct. Likewise, aperture size depends on the spectrophotometer model; medium and large apertures are common, but 20-30mm is not available on every instrument, so write 'largest suitable aperture available on the approved instrument' rather than assuming one exact size across all labs.
A buyer-usable SOP can read as follows: condition samples for at least 4 hours, preferably 8 hours, at approximately 20 +/- 2C and 65 +/- 4% RH; present each blanket face up on a neutral grey backing card or board that masks show-through and keeps the surface flat; for high-loft or translucent pile zones, use two layers of the same blanket body or one blanket over neutral grey board if agreed in PPS, then keep the same method for all readings; lightly align the pile in one defined head-to-foot direction before every read; do not read over crushed fold lines, seams, edge zones within 5cm of the hem or visibly streaky areas unless the purpose is defect diagnosis.
For lot approval, sample at least 2 finished blankets per lot for small lots and 3 per lot once the lot exceeds roughly 3,000 units. Take 5 body readings per blanket: centre, upper left body, upper right body, lower left body and lower right body, all on the display face. If the blanket is reversible or either side may face outward at retail, repeat the same reading plan on the reverse side and report face and back separately. Read all points with the nap oriented the same way relative to the instrument. If the instrument allows rotational averaging, that may be used only if agreed at PPS and applied consistently.
Use D65 illuminant and 10 degree observer unless the buyer's global standard specifies otherwise. The pass/fail rule should cover both average and maximum deviation: each blanket average must be within the approved shade-family tolerance, the lot average must be within tolerance, and no single point may exceed the approved single-point ceiling. As a practical example, a dark shade might be approved at CMC(2:1) average <=1.0 with single-point max <=1.4, while a pale neutral may require average <=0.8 and single-point max <=1.0. Values should be confirmed in pilot bulk, not copied blindly across programmes.
Instrumental data should never be used alone to clear pile anomalies. If one blanket passes average but shows one side panel drifting near the max limit and the panel is also visually evident, treat it as a hold for visual review. Averaging can hide the exact issue that appears on a folded retail face.
Define visual approval conditions so release is defensible
Visual assessment needs fixed conditions or it becomes an argument. Use a controlled light booth with D65 as primary and TL84 or the buyer's defined retail LED as secondary. If a booth is not available for a field review, specify a viewing area in the range of about 1000-1500 lux with neutral surroundings, but booth review should remain the release standard for chain-store programmes. Observation should be at roughly 45 degrees and 90 degrees because teddy fleece can shift appearance between face-on and angled view.
Define blanket presentation state. Review each lot in two forms: opened blanket with pile smoothed in the approved direction, and retail-folded pack exactly as shipped. Assess from approximately 1 metre for opened blankets and 1.5-2 metres for shelf-style folded facings, because store customers and merchandisers do not judge colour from 10cm. Compare the candidate lot side by side against the sealed standard blanket and, where available, the approved shade-band set. All compared units must be oriented with the same nap direction and same face-out fold orientation.
Acceptance criteria should be written plainly: no immediately visible mismatch against the sealed standard under D65 on opened blanket form; no commercially obvious mismatch between lots when retail-folded units are lined up in the same orientation under D65 and TL84 or store LED; no appearance reversal caused by opposite nap or face-out orientation; no compression whitening remaining after the agreed relaxation period. If there is disagreement, buyer QA or merchandising has the final visual release decision, not the inspector alone.
A short shelf-risk scenario helps buyers defend this rule internally: two navy lots can both pass instrument and still look different if one is folded with nap facing upward and the other downward. Once the DC alternates facings on a store shelf, the shelf reads as mixed shade even though each lot independently passed. That is why lot-segregated replenishment and fixed pack orientation matter as much as the dye result.
Write the PO so the factory and DC do not improvise
A chain-store PO should tell the factory exactly what is acceptable, how goods must be folded, how lots must be segregated, and what creates an automatic hold. Generic language such as 'match approved colour' is too weak for teddy fleece. Use a measurable block and attach the colour protocol as a controlled appendix.
A workable PO block reads like this: `100% polyester piece dyed teddy fleece blanket; finished weight 320gsm +/-4% shipment average, no individual blanket below -5% or above +5%; finished size 130x170cm +/-2cm after finishing; one-face consistent pile lay head-to-foot; brushing and shearing to sealed handfeel and appearance standard; colour to approved bulk hanger and sealed finished blanket; instrumental method and shade-family tolerance per buyer colour protocol; no mixed lot IDs in one carton; one lot per inner bundle; maximum two finished lots per colour per PO unless pre-approved in writing; all retail-facing folds to show the same pile direction and same face-out orientation.`
Add specific segregation wording. Example clause: `Factory shall pack one production lot only per carton and one lot only per inner bundle. Cartons shall not contain mixed lots, mixed nap orientation, or mixed face-out presentation. DC and store operators are not permitted to remix lots for replenishment. Supplier shall identify lot boundaries clearly on carton labels, pallet cards, packing list and ASN.`
Define label format so the data survives handoff: `Carton label to show PO, style, colour, lot ID, carton number, quantity, country of origin, and face-out orientation code.` A practical lot code format is `PO-COLOR-YYMMDD-M##-L##`. If there is a belly band or insert card showing one face panel, require the same orientation in every retail unit. This is a common failure point on teddy fleece.
Add concession wording because instrumental pass can still leave visual risk. Example: `Where multiple lots are instrumentally within tolerance but visually distinct in retail presentation, supplier shall hold shipment pending buyer written instruction. No substitution, remixing, relabelling or release is permitted without buyer concession referencing PO, colour, lot IDs, quantity and approved channel/store allocation.`
Control shipment and store allocation, not just factory approval
For chain-store programmes, the safest rule is one approved lot per colour per PO where capacity allows. If volume makes that unrealistic, keep the approved lot count low. A practical ceiling is no more than two approved lots per colour per PO, and preferably one lot per store allocation. Once the count rises beyond two, the burden on DC segregation and replenishment discipline increases sharply.
Assign stores or regional DC allocations by lot, not by mixed-carton availability. Lot A should feed one defined store list or one DC allocation block; Lot B should feed another. Do not alternate lots within the same store unless the buyer has reviewed side-by-side folded presentation and accepted the risk in writing. If e-commerce and store channels share the same PO, direct the visually tighter lot to store retail and the less shelf-sensitive lot to online fulfilment only if buyer merchandising agrees.
Required shipment documents should be listed in the PO pack: lot map by carton range, carton count by lot, pallet count by lot, retained standard references, final inspection photos of opened and folded presentation, instrument report by lot, and signed release sheet stating approved lot IDs. If the programme runs under FOB Ningbo, FCA or another Incoterm, these quality documents still need to be complete before ex-factory release; freight term does not reduce the supplier's documentation obligation.
For buyers comparing constructions, tight lot segregation is usually more critical on lofty teddy fleece than on flatter microfleece. Similar logic appears in [travel-airline-blanket-weight-packing](/blog/travel-airline-blanket-weight-packing.html) and [piece-dyed-280gsm-polyester-fleece-throws-batch-to-batch-shade-toleran.html](/blog/piece-dyed-280gsm-polyester-fleece-throws-batch-to-batch-shade-toleran.html), but teddy fleece needs stricter control of orientation and compression as well as colour.
Root-cause map with buyer-side detection, evidence and action
Buyers need a defect map that tells inspectors what to isolate, photograph and decide. For shade variation between lots, inspectors should pull at least 3 cartons per suspect lot, open 2 blankets per carton, photograph opened blankets and retail-folded units side by side with lot cards and the sealed standard, then compare under D65 and TL84. Outcome: hold shipment if lots are visually distinct; if difference is only between lots and cartons are already segregated, shipment may proceed only with buyer-approved lot allocation and no-remix instruction.
For opposite nap or face-out pack orientation, inspectors should photograph the same style from two cartons showing fold direction arrows or visible pile reversal on the face panel. Isolate at least 32 units or one inner per carton across the suspect range to confirm whether the issue is local or systematic. Outcome: 100% sort and repack at source if orientation differs within the same lot or same store allocation. This is not a dye concession issue; it is a packing control failure.
For compression whitening from vacuum or tight banding, inspectors should photograph packs immediately after opening and again after the agreed relaxation period, often 24 hours at ambient conditions. Sample at least 8 units per suspect lot from top, middle and bottom of pallets because compression can vary by stack position. Outcome: hold if whitening remains commercially obvious after relaxation; repack or change pack method if recovery is acceptable but current presentation is not. Chronic non-recovery on shelf-facing panels is a reject risk.
For barre, rope marks, centre-to-edge variation or patchy shearing, inspectors should photograph defects under both straight-on and angled viewing, include a ruler or marked board for scale, and isolate at least 10 additional units from the same machine lot or finishing batch. Outcome: automatic hold pending factory root-cause review. If defects are recurrent or visible at normal retail viewing distance, treat as reject or 100% sort depending on spread. These are physical or process defects, not routine lot shading.
For mixed lots inside one carton or pallet, inspectors should photograph carton label, inner labels and the mixed contents together. Check 100% of cartons in the affected pallet zone if one mixed carton is found. Outcome: carton reject and repack at source; if shipment timing forces exception review, buyer must issue written concession with exact carton ranges and destination control. Mixed cartons are the fastest way to turn an acceptable bulk into a chargeback.
Finish the approval and release protocol with a complete decision rule
The release protocol should end with one clear rule. Step 1: mill lab confirms recipe match to approved standard. Step 2: factory QA reviews lot instrument data and blocks any lot failing average or max tolerance, or showing visible process defects. Step 3: buyer QA or merchandising performs visual release on finished blankets under agreed lighting in opened and retail-folded states. Step 4: shipment control verifies lot-segregated packing, label format, orientation consistency and lot map completeness. Only after all four steps pass should the lot move to final packing release.
A practical release rule is: `Release only those lot IDs that pass instrumental tolerance, visual approval, and pack/orientation verification. Any lot failing one of the three is HOLD. Lots that instrumentally pass but are visually different in folded retail presentation are not releasable without buyer written concession. Lots that visually pass but instrumentally fail are HOLD pending joint review; release is allowed only if buyer issues written concession referencing test data and commercial channel restriction. Physical defects such as barre, rope marks, patchy shearing or mixed-carton packing are not concessionable unless buyer accepts them explicitly in writing.`
The shipment file accompanying release should include sealed standard references used for comparison, instrument report by lot, lot map, carton count by lot, pallet map, final inspection photos, and signed release sheet listing approved lot IDs and destination allocation. Keep one retained standard blanket and one retained bulk hanger at both mill and buyer side where practical, sealed and labelled with approval date.
If the goods later move under FOB, FCA, CIF or DDP, that commercial term does not change the quality release rule. The quality gate should sit before shipment booking confirmation, not after goods are already handed to the forwarder. For lead-time implications, use the same discipline discussed in [custom-blanket-lead-times-shipping](/blog/custom-blanket-lead-times-shipping.html) when building the critical path.
Escalation path after goods land
Post-landing disputes should follow a fixed sequence so the claim stays evidence-based. First isolate the affected lot or store batch and stop remixing. Second verify the complaint against the retained sealed blanket and bulk hanger, not against memory or a photo. Third review lot map, carton labels, pack orientation and whether the DC mixed lots or reversed facings. Fourth compare instrument data by lot and inspect the complaint units under D65 and store lighting. Only then decide corrective action.
The practical options are usually limited to repack, relabel, store reallocation, selective withdrawal or supplier claim. If the problem is visual mismatch between otherwise acceptable segregated lots, reallocate stores by lot and stop mixed replenishment. If the issue is wrong face-out orientation or mixed cartons, relabel or repack may solve it. If the issue is real process defect such as barre, persistent compression whitening or patchy shearing, the claim should be against the supplier with quantity isolation by lot and supporting photos.
Build the escalation clause into the PO: `After landing, buyer may isolate affected lot IDs pending joint review against retained standards and shipment records. Supplier shall respond with lot data, packing records and release photos within 2 working days. Corrective action shall be agreed as repack, relabel, store reallocation, replacement, debit note or other written settlement by identified lot and quantity.` This keeps disputes tied to evidence rather than broad store-level complaints.
Frequently asked
What is a realistic maximum number of dye lots per colour for a chain-store teddy fleece blanket PO? One lot per colour per PO is safest. If capacity or volume requires more, keep it to a maximum of two approved lots per colour per PO unless the buyer has a strong DC segregation system and agrees store allocation by lot. More than two lots raises shelf-risk quickly on lofty pile fabrics.
Should buyer release be based on Delta E only? No. On 320gsm teddy fleece, instrumental control is necessary but not sufficient. Use a shade-family-specific instrumental tolerance such as CMC(2:1) or Delta E 2000 confirmed in pilot bulk, then require visual approval on finished blankets under D65 and a secondary retail light source. Visual fail should block release even if instrument data passes.
How should teddy fleece be measured on a spectrophotometer? State the method in the PPS or colour appendix. Fix SCI or SCE according to the buyer standard, use the largest suitable aperture available on the approved instrument, align nap in one direction, read conditioned finished blankets on neutral backing, and report both average and single-point maximum. For reversible products, measure both faces separately.
What defects should be treated as process defects rather than normal lot shade variation? Visible barre, rope marks, centre-to-edge variation, patchy shearing, severe streaks, persistent compression whitening and mixed nap orientation inside the same retail presentation should not be treated as routine shade tolerance issues. They should trigger hold, sort, repack or reject depending on spread and severity.
What packing rules matter most for teddy fleece shelf consistency? One lot per carton, one lot per inner, one consistent face-out orientation, and one consistent fold direction. Tight vacuum or aggressive belly-band compression should be validated before bulk because it can create whitening or reversed appearance on the exposed panel.
What documentation should accompany shipment release? At minimum: instrument report by lot, visual approval sheet, lot map by carton range, pallet count by lot, carton count by lot, retained standard references, final inspection photos of opened and folded blankets, and a signed release record listing approved lot IDs and destination allocation.
Have a project in mind? Send us your spec — we'll reply within one business day with indicative pricing and a sample plan.