Every other variable in a ribbon program — colour, print, finish, packaging — is downstream of one number that most buyers never write down: the woven construction. Pick density and warp count determine the ribbon's weight per metre, its hand-feel, its drape, its price, and its abrasion life. Change them and you have effectively launched a different product. This module is about controlling them like the fixed parameters they are, rather than discovering them in a landed cost that came in 12% over plan.
A typical ribbon tech pack from a brand buyer reads like this:
"1–5/8" satin ribbon, 100% polyester, Pantone 294 C, single-side print, 500 m per carton, target FOB Xiamen."
Every one of those attributes is measurable and checkable. None of them tell you how the ribbon feels in the hand, how it drapes over a gift box corner, whether the selvedge will fray during the customer's converting, or why the same quote came back 12% higher last quarter. Those outcomes are determined by the woven construction — and construction is described, in mill documents, by densities, ends and picks.
Ends per inch (EPI) is the warp count. More ends means a denser, more stable, heavier ribbon with better selvedge strength and lower abrasion. Picks per inch (PPI) is the weft insertion rate, which controls how tightly the structure is packed and therefore how firm the hand-feel is. Total picks per centimetre is what the mill actually programs into the loom, and it is the number that governs how far the loom can run without a pick miss.
These are not interchangeable. A satin ribbon's sheen comes from a floating warp structure — long warp floats over many picks — which is why satin picks-per-inch is far lower than its ends-per-inch would suggest. A grosgrain is the opposite: a firm ribbed weft structure with high ppi. Same fibres, same width, completely different hand-feel, weight and cost.
Rule of thumb: a 10% increase in pick density typically adds 6–9% to material cost and 4–7% to loom cycle time. It is the cheapest performance upgrade available — and the easiest thing for a mill to quietly drop when a program gets tight.
As with any multi-mill program, the first requirement is to know who may change what. Construction parameters are assigned across six layers, each with a different owner and a different threshold.
| Layer | Parameter | Owner | Change threshold |
|---|---|---|---|
| L1 | Fibre type and denier (e.g. 150D/48F PET) | Brand, frozen at FAI | None. Substitution requires re-qualification. |
| L2 | Ends per inch (warp count) | Mill weaving, brand co-sign | ±2 ends/inch without notice |
| L3 | Picks per inch (weft count) | Mill weaving, brand co-sign | ±3 picks/inch without notice |
| L4 | Weave pattern / float length (satin vs. grosgrain vs. organza) | Brand, frozen at FAI | None without new sample round |
| L5 | Loom class, reed, harness, tension window | Mill, internal only | Mill notifies brand on change of loom class |
| L6 | Width and picks-per-cm programme (the actual loom setting) | Mill, brand informed | Deviation beyond agreed pick tolerance |
The critical discipline is at L2 and L3. These are the two numbers a mill can adjust most easily on a running programme — a change of reed or a change in weft insertion rate is a loom adjustment, not a machine purchase — and they are also the two that most directly move weight, hand-feel and cost. A buyer who has not co-signature authority at L2/L3 has, in practice, outsourced their most important specification.
Fibre substitution is the substitution buyers hear about most — "the mill offered us a recycled RPET at the same price". It is also the substitution that most often fails downstream. Recycled and virgin polyester differ in dye uptake, surface hand, abrasion behaviour and shrinkage. A fibre swap made to hold a price will show up as a hand-feel complaint, a colour-drift complaint, or a converting problem six weeks later. Freeze L1 and force any change through a full re-qualification.
Construction is approved before colour, and this ordering matters. A buyer who locks Pantone 294 C first and accepts whatever construction the mill proposes will spend the next three rounds chasing a colour that was never achievable on the proposed structure — because sheen, pick density and fibre together determine which colours are reproducible.
Specify fibre, ends/inch, picks/inch, weave pattern, finished width and target weight per metre. No colour reference at all. The mill returns a construction sample plus a weight statement. This stage exists to settle what the fabric is before anyone argues about what colour it is.
Receive the construction sample and measure weight per metre on a calibrated balance across three cuts. Compare against the theoretical weight implied by EPI, PPI and denier. A deviation beyond 5% between theoretical and actual is the signal that the mill is running a different construction from the one you specified — and it is far cheaper to catch here than at a pre-shipment inspection.
Run the construction sample through the actual converting operation — winding onto the customer's bobbin, folding through a bow machine, or feeding a bag-sealing line. Construction that passes a hand-feel check can still fail here: high-construction ribbons can be stiff, can crease at the slitter, and can generate more static and dust. This stage is skipped by most buyers and should not be.
L1, L2, L3 and L4 are locked and signed. Colour development proceeds against a now-fixed structure. From this point the construction numbers go into the quality agreement verbatim, and any movement becomes a non-conformance rather than a conversation.
Buyers discovering a construction drift usually discover it through a price variance. The ladder below turns that variance into a diagnosis.
Before arguing about money, establish the physical facts. Weigh the delivered ribbon. Count ends and picks on a section using a pick glass or a loupe. Measure width with a micrometer. If measured weight is above nominal, the mill has over-constructed; if below, it has under-constructed. In our experience most price disputes at this stage are measurement disagreements, not disagreements about intent.
Rebuild the cost from first principles: yarn consumption per metre (derivable from denier, EPI and PPI), dye and finishing cost, conversion cost, and overhead. Compare to the invoiced figure. A mill running a heavier yarn than specified shows up immediately as a yarn-consumption divergence. This is also where an FX or energy surcharge should be visible as a separate line — if it is buried inside a unit-price increase, the mill has an incentive problem you should address contractually.
Only when Tiers 1 and 2 are settled should the conversation become commercial: is this a one-off adjustment, or does the programme now sit on a different economics basis? Answer honestly. A programme that was priced at 40 EPI is not the same programme as one at 43 EPI, and pretending otherwise for eight months rarely ends well for either side.
Diagnostic shortcut: weight up + price up, with no construction change on paper = over-construction. Weight flat + price up = a cost or surcharge issue, and belongs in Tier 2. Weight down + price flat = the most dangerous case — under-construction, which usually surfaces later as an abrasion or fray complaint.
Construction drift is measurable continuously. A buyer running a multi-season programme should track two derived indicators per lot.
Construction variance index (CVI) compares actual weight per metre against the frozen nominal, normalised so that a reading of 1.000 is on-target. A CVI above 1.05 or below 0.95 triggers a review.
Pick-deviation count tracks observed pick misses and ends-down per 100 metres. A rising trend here is an early signal of loom-setting drift — before weight has moved measurably. Pick misses are cheap to detect and expensive to ignore, because they produce visible striped defects that a brand customer will attribute to printing.
| Indicator | Green | Amber | Red |
|---|---|---|---|
| Construction variance index (CVI) | 0.98–1.02 | 0.95–0.98 / 1.02–1.05 | Below 0.95 / above 1.05 |
| Pick-deviation count per 100 m | 0–1 | 2–4 | 5+ |
| Selvedge fray rate (per 1,000 m) | 0–2 | 3–6 | 7+ |
| Width variance from nominal | ±0.2 mm | ±0.2–0.4 mm | >±0.4 mm |
Two consecutive amber readings, or any single red reading, should trigger a hold-and-investigate before the lot moves into assembly. The cost of holding one lot is a few hours. The cost of a multi-season programme quietly running 6% heavy on a nominally fixed construction is not.
The most common case by a wide margin. Construction was never actually frozen with co-signature authority, so the mill adjusted EPI to manage a yarn cost or capacity issue. Detection: weight up, specification unchanged. Fix: this is a contract problem, not a technical one — which is exactly why L2/L3 co-signature must exist before the first order.
The mill is running lighter to protect margin. The ribbon passes a visual check, passes a colour check and passes a hand-feel check at the counter — and then fails in the customer's converting line with selvedge fraying and abrasion failures. Detection: CVI consistently below 0.95 with amber pick-deviation counts. Fix: reject and re-establish; do not accept a price concession in lieu of compliant construction, because the failure mode lands on the retailer, not on you.
You qualify two mills for capacity redundancy. They build subtly different fabric, because each interpreted the construction brief in its own standard. Product shipped from mill B does not quite match mill A, and the inconsistency surfaces in reviews. Detection: side-by-side weight and pick count across mill lots. Fix: specify EPI/PPI/denier absolutely rather than as a range, and require each mill to submit a construction sample through the same Stage 1–4 gate.
The mill proposes a construction that is perfectly acceptable structurally but cannot hold the target Pantone repeatably — typically a low-pick satin that cannot achieve saturation, or a sheen level that shifts the perceived hue. Detection: lab-dip ΔE drifting between dips on the same construction. Fix: reorder the approval sequence — construction first, colour second — as described in Stage 1.
A recycled or lower-denier fibre is proposed to hold the price. It passes all four gates because nothing in the gates checks long-term abrasion life. Detection: abrasion and crocking results deteriorate 3–4 months into retail. Fix: freeze L1, and add an abrasion-life requirement to the construction brief rather than treating it as a finishing-stage property.
Pull the tech pack for your largest active ribbon programme and check whether EPI, PPI and denier are stated at all. Most buyers discover they are not, and that the mill chose them. This is the single highest-value hour in this entire module.
Fill L1–L6 into the quality agreement with named owners and explicit thresholds. Confirm with each mill that L2/L3 co-signature is accepted in writing. Where a mill resists co-signature at L2/L3, treat that as qualifying information about the relationship, not a negotiation point.
Apply the construction-first approval sequence to your next new ribbon. Measure theoretical versus actual weight at Stage 2. Run the convertibility trial at Stage 3. Expect at least one material you currently buy to come back with a different construction than assumed.
Track the construction variance index per lot per mill. Three amber readings in a row on one mill is a supplier-development conversation, not a per-lot non-conformance. Publish the trend to finance: cost variance caused by construction drift is invisible to them until someone connects the CVI chart to the spend curve.
Reconcile actual delivered construction against the frozen spec across the whole programme year. Where a consistent gap exists, decide deliberately: accept it and re-freeze at the actual numbers, or require correction and plan the transition. What you should not do is let a fourth year pass without deciding.
Woven construction is the part of a ribbon that looks like a number on a mill document and behaves like a product decision. Ends per inch, picks per inch and fibre denier together set the weight, the hand-feel, the abrasion life, the conversion behaviour and the price. Every one of those outcomes is a specification you care about.
The architecture is deliberately unglamorous: six layers with named owners, four approval gates with construction ahead of colour, three tiers of reconciliation when the numbers disagree, and two indicators tracked per lot. It costs almost nothing and it converts the most important parameter in your ribbon programme from something the mill decided into something you control.
When you next qualify a ribbon supplier, ask the question that settles it: Show me the EPI, PPI and denier for my programme, and tell me who signs off if they change. A mill that answers with construction numbers and an authority model is a mill that can hold a specification for three years. A mill that answers with price is telling you exactly what will happen in month four.
Smith Ribbon has supplied woven and printed ribbon, bows, and decorative trims to 1,000+ brands and retailers across 50+ countries since 2004. Our 15,000 m² plant runs 200+ employees with 100,000 meters of daily capacity, certified to OEKO-TEX®, FSC®, BSCI, SEDEX, SMETA and ISO 9001.
Talk to a compliance engineer: +86 13779951780 (WeChat) | xmmsd@126.com | www.smithribbon.com
Request the woven-structure spec template and a construction approval walkthrough →
Smith Ribbon OEM since 2004 — woven ribbon, satin, grosgrain, velvet, organza, RPET, printed bows, decorative trims, and private-label accessories for global brand buyers.