Two suppliers quote the same 10 mm satin at the same price per metre. One is cheaper. You will not find out from the quotation, because the quotation prices a metre, and a metre is not what you buy. What you buy is a finished unit — a bow, a wrapped bottle, a gift box with a handle, a garment with a trim — and the metres that go into it are not the metres that end up on it.
Every ribbon programme carries three kinds of loss: material consumed in the setup, material lost in process, and material stranded on the roll when it ends. Any one of them can turn a good price per metre into a bad price per unit. Here is a method for measuring all three, converting them into a single effective yield figure, and using that figure to compare quotations, suppliers and designs on the same basis.
A finished trim is rarely one piece of ribbon. Before you can measure consumption you need the complete bill of materials for the unit, because the ribbon is often the smallest part of it. A typical gift-bow programme might contain:
| Component | Typical share of material cost | Metre relevance |
|---|---|---|
| Ribbon (1–3 pieces) | 35–55% | Consumed by length; the main yield variable |
| Wire, elastic or braid core | 5–15% | Fixed length, rarely wasted |
| Foil, printed film or transfer | 5–20% | Consumed by area; high waste on small motifs |
| Beads, charms, tags, labels | 5–25% | Unit count, not metres |
| Carton, polybag, tissue, labels | 5–20% | Per-unit packaging |
Only the ribbon, the foil and the printed film scale with length. Everything else is a unit cost that you can negotiate, bundle or substitute independently. Keeping the two categories separate is what makes the yield analysis useful — a 12% ribbon yield loss on a unit where ribbon is 40% of material cost moves total material cost by under 5%, and that is rarely the biggest number in the programme.
Nominal consumption is the length of ribbon in the finished product, with no allowance for anything. For a simple flat bow it is straightforward: two loops plus two tails, minus the centimetres that disappear inside the knot and under the centre band. For a more complex bow — a double-layer bow, a wired-edge bow, a bow with a gathered or ruffled loop — the nominal length grows because the gather consumes more ribbon than the flat dimensions suggest.
Write the nominal figure down per component, not per unit. For a 10 mm double bow: 2 × 0.28 m loops, 2 × 0.22 m tails, 0.12 m gathered overlay, less roughly 0.05 m lost in the knot and under the centre band. That is 0.85 m nominal against a finished bow that a customer sees as a 12 cm object.
The temptation at this point is to stop, because nominal consumption is the number that appears on the factory's quotation and it looks like the finished cost. It is not the finished cost. It is the input before any loss.
Setup loss is ribbon consumed while the machine is being set up and dialled in: thread, test bows, colour and tension trials, the first few units after a changeover. It is concentrated at the start of a run and it scales with the number of setups, not with the order quantity. A 5,000 m order split into five 1,000 m runs pays setup five times.
In-process loss is ribbon rejected during production — a knot that fails the pull test, a bow that misses the dimensional tolerance, a colour defect, a print registration failure, a re-work. This is the bucket buyers underestimate most, because rejected units are often re-made silently rather than logged. If your supplier does not record rejects against the batch, ask them to start. It is the only way to know whether a reject rate of 3% is normal or exceptional for the construction.
This is the material stranded on the roll when a run ends, plus the remnants that are too short to run but too valuable to bin. A 500 m roll that stops with 40 m remaining means 8% of that roll becomes remnant. With a 10 m minimum usable length, part of that 40 m is recoverable and part is scrap. Remnant management is a real commercial lever here — the difference between a remnant library and a skip is the difference between an asset and a loss.
Collect these three buckets separately for every SKU in the programme. The combined number is your yield loss, but the mix tells you what to fix. High setup loss points to order sizing and run planning; high in-process loss points at construction and quality; high remnant loss points at roll length and order quantity.
Effective yield is the ratio of ribbon that reaches the finished unit to the ribbon that was purchased. The formula is short, and it is the number that should sit next to every price per metre in your sourcing file.
Effective yield = Nominal consumption ÷ (Nominal consumption + setup loss + in-process loss + remnant loss)
Worked example for one 10 mm satin bow SKU at a 5,000 m order:
| Line | Metres per 1,000 finished units | Share |
|---|---|---|
| Nominal consumption | 850 | — |
| Setup loss (3 setups) | 45 | 4.4% |
| In-process loss (2.5% rejects) | 23 | 2.2% |
| End-of-roll remnant (1 roll ends at 60 m) | 60 | 5.8% |
| Total ribbon consumed | 978 | — |
| Effective yield | 86.9% | 13.1% lost |
Now compare it to the alternative. A second supplier quotes the same nominal consumption at 0.85 m but runs 5,000 m in two setups instead of three, has a 1.5% reject rate, and ends both rolls within 20 m of a clean cut. Total consumption is 908 m per 1,000 units, and the effective yield is 93.6%. On a 5,000 m programme the difference is 350 m of ribbon — before anyone talks about price per metre.
That gap, between two suppliers quoting the same price per metre, is the single most under-measured variable in most ribbon programmes. It is frequently larger than the price negotiation itself, and it costs nothing to measure.
Cost per finished unit is the number finance actually cares about, and it is the number to put in front of a brand buyer who thinks in retail margin rather than in metres.
Cost per unit = (ribbon price per metre × total metres consumed) ÷ units produced, plus every other material line
Using the example above at a price of $0.085 per metre for 10 mm satin: 978 m × $0.085 = $83.13 of ribbon per 1,000 units, or $0.083 per unit, against a nominal cost of $0.072 per unit. The 13.1% yield loss adds $0.011 per bow. On a programme of 200,000 units a year that is $2,200 — and it is the kind of number that is invisible on a purchase order because the purchase order buys metres, not bows.
The practical use is comparison. Put two suppliers on this basis and you can finally answer the question a buyer actually cares about: which one delivers the cheaper finished unit.
In most programmes we have measured, the recoverable loss sits in a predictable order:
Two things that usually do not deliver: negotiating a lower price per metre in isolation, and switching to a supposedly cheaper construction without re-measuring yield. Both can easily be given back by a worse yield figure.
Any supplier that cannot answer these four questions with numbers is not managing yield — and you are paying for their uncertainty inside your price.
The method is only worth the effort if it changes decisions. Three moments when it earns its keep:
At quotation stage. Convert every quotation to cost per finished unit before comparing. Where two suppliers are close, ask each for the four numbers above and model the difference — the supplier with the higher nominal price frequently wins once yield is included.
At design review. Take the candidate constructions and compare nominal consumption, not just ribbon appearance and unit price. A 6% difference in nominal consumption across a 200,000-unit programme is a material line item.
At reorder. Compare the actual consumption from the last run against what was planned. The gap is either a plan error or a process problem, and it is always cheaper to find it at reorder than at annual review.
Standalone yield studies do not survive contact with a buying calendar. What survives is a small number of fields in the system of record: nominal consumption per component, measured reject rate, setups per run, standard roll length, remnant recovery policy. If those five fields are captured per SKU per supplier, cost per finished unit becomes a report rather than a project — and the programme can be re-costed in an afternoon whenever a price moves.
At Smith Ribbon, a 15,000 m² factory in Xiamen with more than 200 employees and daily output of 100,000 metres, we measure and report consumption per SKU for every OEM programme, because we are quoting per metre and we need to know what it becomes in finished units as much as you do. Certification includes OEKO-TEX, FSC, BSCI, SEDEX/SMETA and ISO 9001; we export to more than 50 countries and support OEM and ODM customisation from 1,000 metres, with small-batch runs from 500 metres. If you send us one of your SKUs with the current quoted consumption and the measured rejects, we will run the yield model and tell you where the missing 10 to 15% sits.
Email: xmmsd@126.com | Mobile / WeChat: +86 137 7995 1780 | Office: +86-592-5095373
© 2026 Xiamen Smith Ribbon & Bow Co., Ltd. | Professional Ribbon Manufacturer Since 2004
These are the questions brand buyers, packaging developers and procurement managers most often ask us about ribbon cost and yield. If your question is not here, ask us directly.
Because a metre is what you buy, not what you use. Setup loss, in-process rejects and end-of-roll remnants all consume ribbon that never reaches the finished unit, and those three buckets together commonly account for 10 to 15% of purchased material. Two suppliers quoting the same price per metre can therefore deliver very different cost per finished unit, and only the second number survives contact with a brand buyer's margin model.
Setup loss is ribbon burned while dialling in a run — thread, test pieces, colour and tension trials — and it scales with the number of setups rather than with order quantity. In-process loss is ribbon rejected during production for quality reasons, including silent re-work. End-of-roll loss is material stranded on the roll when a run ends, part of which may be recoverable as usable remnant. Keeping them separate tells you what to fix: run planning, construction and quality, or roll length and order sizing respectively.
Effective yield equals nominal consumption divided by total consumption, where total consumption is nominal consumption plus setup loss plus in-process loss plus end-of-roll remnant loss. In the worked example, 850 m of nominal consumption for every 1,000 finished units plus 128 m of loss gives 978 m consumed and an effective yield of 86.9%. That single figure should sit next to every price per metre in your sourcing file so that quotations can be compared on a like-for-like basis.
Run consolidation and order sizing, because setup loss is typically the largest single recoverable bucket. Fewer setups, longer runs and fewer colour changes per run reduce it directly. The next largest is usually matching roll length to order quantity, which is an arithmetic problem rather than a negotiation. Construction selection and simplification of gathered or ruffled components come after that. Negotiating price per metre in isolation usually delivers least, and can easily be given back by a worse yield figure.
Yes, and if your supplier does not record them you should ask them to start. Rejected units are frequently re-made silently rather than logged, which makes the in-process loss bucket invisible. Without recorded reject rates you cannot distinguish a normal 3% loss from an exceptional one, and you cannot price the programme correctly — nor can the supplier price it correctly, which is part of why their quotation carries an unexplained risk premium.
A gathered or ruffled component consumes ribbon out of proportion to its visual value, because the gather adds length that the flat dimensions do not show. Comparing constructions on nominal consumption rather than on appearance and unit price alone can reveal differences of several percentage points, which on a 200,000-unit programme is a material line item. Where two constructions deliver a comparable finished appearance from flat components, the flat version usually wins on yield as well as on cost.
Five fields per SKU per supplier: nominal consumption per component, measured reject rate, setups per run, standard roll length, and the remnant recovery policy. With those captured, cost per finished unit becomes a standing report rather than a one-off project, and the programme can be re-costed quickly whenever a price, a construction or a run pattern changes.
Yes. Send one SKU with its current quoted consumption per unit, the measured reject rate if you have it, and the order quantity and roll length you are working to. We will run the yield model, show you where the missing material actually sits — setup, rejects or remnants — and tell you which of the three is worth attacking first. If the numbers are not being tracked yet, we can start from nominal consumption and build the model from your next run's actuals.