A bespoke bag programme should be measured on five families of KPI, not on sales alone: quality, speed, cost, consistency and relationship health. Sales is a lagging indicator that reflects last season's design and marketing as much as the programme itself, so a brand that judges its production partner by sell-through alone is flying blind. The metrics below — first-pass inspection yield, sampling turnaround, unit-cost trend, specification drift and reorder behaviour — are leading indicators, and they tell a brand what its next run will deliver before the customer ever sees the range.
Why Sales Is the Weakest Signal of Programme Health
Sales is necessary but insufficient. It is influenced by pricing, seasonality, channel and marketing, none of which the production programme controls, and it arrives too late to change anything about the run it describes. A programme can be technically excellent and still record weak sales, and it can be technically fragile and record strong ones. The purpose of a wider KPI set is to separate the quality of the programme from the quality of the season, so that decisions about the next run are made on evidence rather than on a single number that mixes both.
Programme parameters at a glance
| Parameter | Commitment |
|---|---|
| Minimum order quantity | 500 pieces per style / colourway |
| Sampling lead time | 6–10 working days (12–15 days for complex constructions) |
| Bulk production lead time | 35–50 days after sample approval |
| Inspection standard | AQL 2.5 (ISO 2859-1), single sampling, general level II |
| Incoterms | FOB Xiamen — EXW, CIF and DDP available on request |
| Payment terms | T/T 30% deposit / 70% before shipment; L/C at sight for container loads |
| Verified production floor space | 4,950 m² (SGS-verified) |
| Team | 137 staff across 7 production lines |
| Equipment | 149 machines including computerised saddle-stitch and edge-painting units |
| Monthly capacity | 200,000 pieces |
| Entity | Founded 2014; principal has worked in bag development since 2004 |
The most useful reframing is to treat the programme as a small operating system with inputs, throughput and outputs. Inputs are materials and briefs. Throughput is sampling, production and inspection. Outputs are delivered units and their condition. Sales sits downstream of that system and is affected by many things outside it. A brand that instruments the inputs, throughput and outputs can diagnose a disappointing season honestly — was the design wrong, or was the execution inconsistent? — rather than blaming the wrong party.
| KPI family | What it answers | Leading or lagging |
|---|---|---|
| Quality | Did the run meet the standard? | Leading — predicts returns and reputation |
| Speed | Did the programme keep its dates? | Leading — predicts launch reliability |
| Cost | Is the unit cost moving in the right direction? | Leading — predicts margin |
| Consistency | Did every tranche match the sample? | Leading — predicts drift and rework |
| Relationship | Is the partnership getting easier or harder? | Leading — predicts resilience |
| Sales | Did the market respond? | Lagging — reflects many factors |
Building a Measurement Framework
A workable framework is small. Choose two or three metrics per family, define each precisely, and record them at the same points in every programme cycle so that comparison is meaningful. The trap is ambition: brands that try to track twenty metrics end up tracking none carefully. The right number for most teams is between eight and twelve, each with an owner, a data source and a review cadence tied to the programme calendar rather than to the financial year.
- Choose two or three metrics in each of the five families, no more.
- Define each metric in one sentence so it cannot be interpreted two ways.
- Name an owner and a data source for every metric before the run begins.
- Record at fixed points — after sampling, after inspection, after delivery.
- Review at a fixed cadence, ideally per tranche rather than per season.
- Act on the review: change one variable in the next run and watch it move.
Quality Metrics: Yield, Defects and Inspection Outcomes
Quality is the family most often reduced to a single pass or fail, which wastes most of the information available. The starting point is the inspection standard, typically AQL 2.5 under ISO 2859-1 using single sampling at general inspection level II, which defines how many units are checked and how many defects are acceptable. Around that standard a brand should record first-pass yield, defect category, rework volume and the time to close out any corrective action. Together those four numbers describe not just whether a run passed but how close it came to failing.
| Quality metric | Definition | What a good trend looks like |
|---|---|---|
| First-pass yield | Share of units passing inspection without rework | Rising across tranches |
| Critical defects | Units with a fault affecting function or safety | Zero tolerance, flat at zero |
| Major defect rate | Units with a visible fault against the sample | Falling, inside the AQL limit |
| Rework volume | Units corrected before shipment | Falling as operators gain familiarity |
| Corrective-action closure | Days from finding to verified fix | Shortening cycle over time |
| Sample adherence | Conformity of production units to the retained sample | High and stable |
First-pass yield deserves emphasis because it is the metric that most directly predicts the true cost of a programme. A run with a high first-pass yield consumes little rework labour and arrives with fewer surprises. A run with a low first-pass yield can still pass inspection — defects may be corrected before shipment — but it has spent hidden hours and introduced risk. Tracking yield rather than only the final pass result is what allows a brand to see a drifting programme early, while there is still time to correct it.
Speed Metrics: Sampling Turnaround and Lead-Time Adherence
Speed governs whether a brand can launch when it said it would. The two essential measures are sampling turnaround, which typically runs six to ten working days for standard constructions and twelve to fifteen for complex ones, and bulk lead-time adherence, measured against the thirty-five to fifty days that follow sample approval. A third measure, reorder velocity, describes how quickly a repeat tranche can be produced once the specification and sample already exist, and it is the number that rewards the discipline of freezing a specification.
| Speed metric | Baseline to expect | Signal to watch |
|---|---|---|
| Sampling turnaround | Six to ten working days standard | Slippage beyond the quoted range |
| Complex sampling | Twelve to fifteen working days | Repeat rounds beyond two |
| Bulk lead time | Thirty-five to fifty days after approval | Adherence to the committed date |
| Reorder velocity | Shorter than a first run | Whether it is genuinely shortening |
| Inspection turnaround | One to two days | Compression that skips steps |
| Internal approval time | Brand-controlled | The hidden delay in most programmes |
It is worth separating the two sources of delay explicitly. Some slippage is production-side; more is often brand-side, because internal review of samples and artwork takes longer than the calendar allowed. A programme that measures only the partner's lead time will systematically miss its own contribution to lateness. Recording both, side by side, removes the argument and usually reveals that the fix lies in the brand's own approval process as much as in the production schedule.
Cost Metrics: Unit Trend, Amortisation and Cost of Poor Quality
Cost metrics should answer a directional question rather than an absolute one: is the cost of a given specification moving in the right direction as the programme matures? Two mechanisms drive that movement. One-off development costs — sampling and tooling — amortise as later tranches reuse them. Recurring unit costs fall modestly as yield improves and operators gain familiarity with a style. A programme that is not showing either trend is either re-developing every tranche or has a consistency problem that is quietly consuming the savings.
| Cost metric | What it reveals | Direction of travel |
|---|---|---|
| Development amortisation | How much of the one-off cost each tranche carries | Falling across tranches |
| Unit cost trend | Movement in recurring cost at fixed specification | Flat to slightly falling |
| Cost of poor quality | Value lost to rework, scrap and returns | Falling |
| Yield efficiency | Usable material per hide | Improving with better nesting |
| Landed cost per unit | Including freight and duty | Tracked, not assumed |
| Change cost | Cost of altering specification mid-programme | Avoided by freezing the spec |
Cost of poor quality is the metric most often left unmeasured and most often underestimated. It includes rework labour, scrapped material, expedited freight and the returns that follow a defect reaching a customer. On a programme where first-pass yield is already high it is a minor line; where yield has drifted it can silently exceed the entire development budget. Measuring it forces a brand to see quality and cost as the same conversation rather than as two departments with different interests.
Consistency Metrics: Detecting Specification Drift
Consistency is the quiet KPI. A programme can pass inspection on every tranche and still drift, unit by unit, until a bag from the fourth tranche looks subtly different from one in the first. This is why the retained approved sample matters so much: it is the physical anchor against which drift is measured. Specification drift shows up as small variations in colour, hardware tone, stitching density or branding placement, none of which fails a standard on its own but all of which, cumulatively, make a range look inconsistent on the shelf.
- Colour variation between hide batches, measured against a fixed swatch.
- Hardware tone and finish variation between component shipments.
- Stitch density and edge-paint consistency against the retained sample.
- Branding placement tolerance, especially on debossed and engraved marks.
- Lining and interior specification holding to the approved build.
- Reusable components — straps, hardware, closures — staying interchangeable.
The discipline that controls drift is procedural rather than technical. Keep one approved sample held by each party, inspect every tranche against it rather than against a drawing, and record deviations as data rather than as complaints. When a deviation is identified, decide explicitly whether it is within tolerance or a corrective action, and if it is the latter, update the sample so that the reference itself reflects the corrected state. A programme that never updates its reference gradually redefines its own standard downward.
Relationship and Commercial Metrics
The final family measures the health of the partnership, because a bespoke programme is a relationship before it is a transaction. The useful metrics here describe responsiveness, problem-solving and repeat behaviour: how quickly questions are answered, how often issues are resolved without escalation, what proportion of styles are reordered, and whether the terms remain stable. These are softer than defect rates, but they are leading indicators of whether the programme will still be viable when a difficult season arrives.
| Relationship metric | Definition | Why it predicts resilience |
|---|---|---|
| Response time | Time to answer a specification or quality query | Fast answers prevent compounding issues |
| Issue resolution | Share of problems closed without escalation | Shows ownership of outcomes |
| Style reorder rate | Share of styles brought back for a second run | Reflects fit and quality |
| Specification stability | Number of changes after freeze | Low drift means a clear brief |
| Terms stability | Consistency of payment and shipping terms | Underpins planning and cash flow |
| Sample retention | Both parties holding the approved reference | Enables objective quality discussion |
Fill rate and on-time delivery belong here as well, because they are the commercial consequences of the quality and speed families. A style that is reordered is the strongest possible signal that design and execution aligned. A style that is never reordered, despite adequate sales, usually points to a quality or consistency issue that did not fail inspection but reduced the customer's willingness to carry the bag — a signal no sales report will show clearly, and one that a reorder-rate KPI surfaces immediately.
Turning KPIs Into Programme Decisions
Metrics are worthless until they change a decision. The purpose of the framework above is to answer three recurring questions: which styles should be reordered and which should be retired, which parts of the specification should be adjusted, and whether the partnership is improving or deteriorating. A programme review that ends without a decision about at least one of those three has merely produced a report. The value comes from changing one variable in the next run and observing the effect, tranche by tranche.
It helps to fix the cadence. Reviewing the metrics per tranche, rather than once a season, keeps the feedback loop short enough to influence the next run. Reviewing them only at the financial year end means that every insight arrives after the opportunity to use it has passed. A programme that reviews per tranche and acts on one variable each cycle will typically show measurable improvement in first-pass yield and lead-time adherence within a small number of runs, which is the clearest evidence that the measurement itself is doing its job.
Instrument your programme
Ask the programme team to set up a KPI sheet for your next tranche — quality, speed, cost, consistency and relationship metrics recorded at fixed points. Email service@junyuanbags.com or message us on WhatsApp at +86 177 5002 0688.
Frequently Asked Questions
Why are sales not enough to measure a bespoke bag programme?
Because sales is a lagging indicator shaped by pricing, seasonality, channel and marketing, none of which the production programme controls. A technically strong programme can record weak sales, and a fragile one can record strong sales, so the number alone cannot tell a brand whether the programme is healthy. Leading metrics such as first-pass yield, lead-time adherence and unit-cost trend describe what the next run will deliver, which is what a brand can actually act on.
How many KPIs should a bespoke bag programme track?
Between eight and twelve is the practical range, distributed as two or three metrics in each of five families: quality, speed, cost, consistency and relationship health. The temptation is to track twenty or more, which usually results in none being recorded carefully. Each metric should be defined in a single sentence, assigned an owner and a data source, and recorded at the same points in every cycle so that comparisons across tranches remain meaningful.
What is first-pass yield and why does it matter?
First-pass yield is the share of units that pass inspection without any rework. It matters because it predicts the true cost and risk of a run more directly than the final pass result does. A run can pass inspection while consuming large amounts of hidden rework labour, and first-pass yield is the metric that exposes this. A rising first-pass yield across tranches is one of the clearest signs that a programme is maturing.
Which quality standard should programme inspection use?
AQL 2.5 under ISO 2859-1, applied as single sampling at general inspection level II, is the widely used reference in this trade. It defines how many units are inspected and how many defects are acceptable before a consignment is rejected. Recording first-pass yield, defect categories, rework volume and corrective-action closure time alongside that standard gives a far fuller picture of quality than the pass or fail outcome alone.
How do I measure sampling speed fairly?
Measure from the moment the brief is complete and materials are confirmed, not from the moment the idea was discussed, and compare against realistic baselines — six to ten working days for standard constructions and twelve to fifteen for complex ones. The fair measure is turnaround per full sampling round, including the brand's own review time, because internal approval is the hidden delay in most programmes. Measuring only the production side systematically flatters the programme.
What lead time should bulk production be measured against?
Thirty-five to fifty days after sample approval is the standard baseline for bulk production. The important detail is the starting point: the clock begins at sample approval, not at order placement, because a specification that is still being revised cannot be produced at a predictable rate. Measuring adherence to the committed dispatch date across tranches shows whether the programme keeps its promises and whether a reorder can be planned on a shorter, more confident cycle.
What is specification drift and how is it detected?
Specification drift is the gradual divergence of production units from the approved sample across successive tranches, showing up as small variations in colour, hardware tone, stitching density or branding placement. It is detected by holding one approved sample with each party and inspecting every tranche against that physical reference rather than against a drawing or a photograph. Individually the deviations may sit within tolerance; cumulatively they make a range look inconsistent.
How should a brand measure the cost of poor quality?
Add up everything consumed by defects: rework labour, scrapped material, expedited freight and the returns that follow a defect reaching a customer. On a programme with high first-pass yield this is a minor line, but where yield has drifted it can quietly exceed the entire development budget. Recording it forces quality and cost into a single conversation instead of treating them as separate concerns owned by different people.
What does the unit cost trend tell a brand?
It shows whether the cost of a fixed specification is moving in the right direction as the programme matures. Two mechanisms drive it: one-off development costs amortising across later tranches, and recurring unit costs falling modestly as yield improves and operators gain familiarity with a style. A programme showing neither trend is either re-developing every tranche or losing the savings to a consistency problem that is consuming them invisibly.
Which relationship metrics predict a resilient programme?
Response time to specification and quality queries, the share of issues resolved without escalation, the proportion of styles reordered, specification stability after freeze, and consistency of payment and shipping terms. These are softer than defect rates but they indicate whether the partnership can absorb a difficult season. A style that is reordered is the strongest signal that design and execution aligned, and a style that is never reordered often hides a quality issue that never failed inspection.
How often should programme KPIs be reviewed?
Per tranche rather than once a season. A tranche-level cadence keeps the feedback loop short enough that an insight can influence the very next run, whereas an annual review delivers every lesson after the opportunity to apply it has passed. Programme reviews should ideally end with a decision — reorder, retire, adjust the specification or address a process problem — because a review that produces only a report has not yet done its work.
What is the difference between a leading and a lagging KPI?
A leading indicator predicts future performance and can still be acted upon, such as first-pass yield or lead-time adherence. A lagging indicator reports what has already happened, such as sales or returns, and cannot be changed for the run it describes. A programme should be judged mainly on leading indicators, using lagging indicators as confirmation rather than as the primary basis for decisions about the next production cycle.
How can a brand tell whether a poorly performing style should be retired?
Look at the reorder rate alongside the quality metrics. A style with acceptable sales but a low reorder rate and rising defect or rework figures usually has a fit or quality problem that the customer noticed even though the units passed inspection. A style with weak sales but clean quality and a normal reorder rate points instead to a design or positioning issue. The two diagnoses lead to different decisions, which is why quality and reorder KPIs should be read together.
Does a brand need a retained sample to measure consistency?
Yes. Without a physical approved sample there is no objective reference against which drift can be measured, and every quality discussion becomes a matter of opinion. Both the brand and the production partner should hold one. When a deviation is identified and corrected, the sample should be updated so the reference reflects the corrected state; otherwise the programme quietly redefines its own standard downward with each tranche.
What is reorder velocity and why does it matter?
Reorder velocity measures how quickly a repeat tranche can be produced once the specification and approved sample already exist. It is typically shorter than a first run because sampling and tooling are already complete, and it is the metric that rewards the discipline of freezing a specification. A programme whose reorder velocity is genuinely shortening can respond to demand faster and carry less inventory, which improves working capital as well as launch reliability.
How should KPIs change decisions about the next production run?
They should resolve at least one of three recurring questions: which styles to reorder and which to retire, which part of the specification to adjust, and whether the partnership is improving or deteriorating. The practical method is to change one variable in the next run and observe the effect across tranches. Programmes that act on a single variable each cycle usually show measurable improvement in yield and lead-time adherence within a few runs.