In-House Test Reports Get Rejected: What to Commission Instead
When a hotel brand rejects your factory’s in-house report, commission the tests the brand actually names, from a laboratory accredited to ISO/IEC 17025 for...
read more…
A cutting ticket for striped upholstery must state seven things before the blade drops: the measured repeat of the fabric in both directions, one named control stripe, one direction rule for the whole suite, the seams where matching is mandatory, which panels are mirror pairs, the extra yardage that matching consumes, and the rule that stripes meet at the seam line rather than at the cut edge. If the ticket only says striped fabric, cut to pattern, every cutter invents their own starting point, and the delivery you received, with stripes drifting at every seam, is exactly what that ticket predicted. The fix is not a sharper eye at the sewing machine. It is a document chain that runs from the purchase order to the cutting table to a signed first piece, so the match you approved is the match that ships.
Stripes are the least forgiving pattern in upholstery because the eye finds the error even when the seam is neat. A solid fabric hides a cutter’s choices; a stripe publishes them. That is why a ticket written for solids fails on stripes: it contains no repeat, no direction, and no named reference line, so nothing in it can be violated. A cutter who lays the fabric for yield, one panel at a time, produces a suite in which every cushion is internally correct and no two agree with each other.
The failure you saw at every seam also points at the ticket rather than at one careless hand. When one cushion is off, look for a sewing error. When every seam is off by a similar amount, the reference itself was missing or misplaced, most often because match points were keyed to the cut edge instead of the seam line, so each sewn seam landed one seam allowance out of register. Uniform misalignment is a document defect, and documents are cheaper to fix than re-cuts.
Treat stripe matching as a five-stage chain in which each stage produces something you can look at and sign. If a stage produces nothing, the chain is broken there.
Before any ticket is written, measure the repeat on the roll that will actually be cut, not on the sample book swatch. The vertical repeat runs along the roll length and the horizontal repeat runs across the width, and both matter: the vertical repeat sets how much extra length each panel needs, and the horizontal repeat decides whether stripes can align where panels meet side by side, as they do where a seat deck meets an inside back. Record the two numbers on the ticket with the fabric code and dye lot.
Then name one control stripe. The control stripe is the single stripe the whole suite is set out from, for example the dominant stripe centered on the cushion top, the seat front edge, and the headboard face. Without a named line, every cutter centers something different: one centers the widest stripe, another centers the space between stripes, and a third keys everything to the panel edge. The control stripe is also what you will audit against at delivery, which is why it must be named, not implied. Buyers who approve fabric under controlled light rather than in a showroom glow already know this discipline, and the same approval logic extends to where the stripe sits on the panel.
Write the ticket as measured fields, each one short enough to be checked by someone who was not in the room when it was drafted. These are the fields that decide whether a striped suite can come back matched.
A ticket can be perfect and still die at the cutting table, so the cutting room owes you two pieces of evidence before the batch proceeds. Neither is elaborate, and both take minutes.
The marker is the full lay of fabric with all pattern pieces positioned. Photograph the marker with the ticket number visible, taken from above, so the stripe positions at every panel edge can be read later. The photograph proves that the lay followed the ticket: same direction on every piece, mirror pairs side by side at the same stripe position, and the control stripe landing where the ticket says it lands. When a delivery dispute arrives months later, this image is the difference between re-cutting one bad lay and re-cutting an argument. Ask that the marker photo be attached to the batch paperwork the same way the packing photos are attached to the shipment file.
Before the batch is sewn, the first complete pair, for example one seat cushion mated to its seat deck, is assembled and checked at eye level under the light the workshop actually sews in. The person who checks signs the ticket line, and the signed first piece stays at the sewing line for the batch to be compared against. This is the batch anchor: when a new cutter joins mid-run, or a fabric roll changes, the comparison is against a physical, signed object rather than against memory. Keep the first piece with the order file at handover, because it remains the reference for reorders years later.
At delivery, the handover pack should contain the stripe chain’s outputs, and you should audit against them in one sitting. The table below is the checklist: each deliverable proves something specific, and each audit step has a rule for what happens when it fails.
| Deliverable or check | What it proves | What you do when it fails |
|---|---|---|
| Marker photograph with ticket number | The lay followed the ticket before cutting began | Stop re-cuts until the marker is re-photographed and compared to the ticket |
| Signed first piece kept at the line | The batch had a physical match reference before sewing | Re-anchor the batch: re-check every finished piece against a new signed pair |
| Seam audit against the control stripe | Mandatory seams meet at the seam line as the ticket required | Log seam, amount off, and carton; classify per the triage rules below |
| Fabric lot label on each piece | Rolls and dye lots are known per piece, so repairs use the same lot | No repair proceeds until the lot of the damaged piece is identified |
| Direction check across the suite | All pieces run the stated direction, including headboards against seats | Treat a direction break as a re-cover, because it cannot be sewn away |
Audit in a fixed order so the findings are comparable: headboard centerline first, then seat front edges, then cushion tops against seat decks, then mirror pairs side by side. Photograph each checked seam straight on, from the distance a standing guest sees it, and record the offset in millimeters against the control stripe. A written audit turns the general complaint, stripes out of line, into a list of specific seams with specific amounts, which is what a factory needs in order to answer.
Matching is not free, and a ticket that pretends otherwise forces the workshop to steal fabric from somewhere. Each matched panel pair needs positioning room, so plan roughly one extra repeat per panel beyond the plain-cutting calculation, and let the wide pieces, headboards and seat decks, drive the total. The honest conversation happens before the order: either the budget carries the extra yardage, or the ticket defines where matching is mandatory and where it floats, in writing, with the buyer’s signature. What must not happen is the silent version, where the ticket is written as if the fabric were plain, the yardage arrives short, and the cutting room quietly skips the matching to make the goods stretch.
Time works the same way. A striped order needs a marker planned for the whole batch rather than panel by panel, a first piece sewn and signed before the run, and an audit at the end. On a hotel suite order that is a day of planning and an hour of signing, spread across weeks of production, which is why at EverGoCasa we treat the stripe chain as part of the order’s normal milestones rather than as a special favor. Buyers who are planning a visit can walk the cutting room and the Foshan showroom side by side, and see matched striped sets displayed the way they will be installed.
When the audit finds stripes off the control line, judge each finding against three thresholds agreed before delivery. Small offsets at permitted seams are accepted and recorded, because the ticket said they float. Moderate offsets at a mandatory seam on a visible face go to repair when the trade can re-set the seam without disturbing the panel’s stripes elsewhere. Offsets that cannot be sewn away, or a direction break anywhere in the suite, go to re-cover, and the replacement cover is cut from the same dye lot as the piece it replaces, which is why the lot label on every delivered piece mattered in the audit. The triage decision is made per seam, on the photograph, against the thresholds, not per mood.
Whatever the remedy, the correction loop closes the same way the chain started: the cutting ticket for the replacement is checked against the signed first piece, the marker for the re-cut is photographed, and the new first piece is signed before the panel is made. A factory that runs the chain once will run it cheaply the second time, and the defect class that hit you at every seam becomes the easiest one to prevent.
Stripes drift at every seam when nobody owns the reference. The five-stage chain gives the reference an owner at each stage: the order funds the repeat, the ticket names the control stripe and the direction, the marker photo proves the lay, the signed first piece anchors the batch, and the seam audit closes the loop at delivery. Put the chain into the purchase order as deliverables with names and dates, and a striped order stops being a gamble on workshop habits.
Mills publish the underlying vocabulary for a reason, and Warwick’s upholstery guidance on repeats and railroading is a fair plain-language reference to hand to anyone drafting a first ticket. Your part is to make the vocabulary contractual: measured repeats, a named control stripe, mandatory and permitted seams, mirror pairs, funded yardage, and match points at the seam line. Approve the signed first piece the way you approve the fabric itself, and the suite that arrives for installation will match the way the showroom promised, seam by seam, without a single argument about who saw what.