Custom Bag Sampling Data: Average Rounds and Time to Approval

Sampling is the part of a bag programme that creators most consistently underestimate, and it is underestimated in a specific way: they budget the per-round time correctly and the round count wrongly. A creator who assumes one round at 6 to 10 working days and actually needs three has lost a month, and that month usually lands directly on the campaign launch date or the delivery promise. This article sets out what a round is, how many are typically needed, where the days go, and how to compress the sequence without cutting quality. The wider planning sequence sits in the crowdfunding bag guide.

What Counts as One Sampling Round

The first problem with sampling data is that nobody defines a round the same way. In our usage, one round is the complete cycle from a confirmed specification and released materials through to a physical sample in the creator's hands, reviewed against a written checklist. It includes pattern making, material preparation, cutting, sewing, hardware fitting, internal checking and dispatch. It does not include the creator's review time, which sits outside the clock and is frequently the largest single component of the elapsed calendar.

That definition matters because it separates the two things a creator can control. The cycle inside the workshop is fairly stable at 6 to 10 working days for a standard style. The review time outside it is entirely under the creator's control and is often where a week disappears. Our programme coordinator treats them as separate line items for exactly that reason, and asks creators to commit to a review turnaround before sampling begins.

A revision is not necessarily a new round. If the first sample comes back with a colour that is slightly off and the correction is a material swap with no pattern change, many programmes handle that as a correction within the round rather than a fresh one. If the correction requires a pattern change or new hardware, it is a new round with its own clock. Understanding that distinction lets a creator ask the right question: is this a correction or a re-sample?

There is one more category worth naming: the confirmation sample. This is a unit pulled from the start of mass production and checked against the approved sample, which is technically a first-article inspection rather than a sampling round. It does not add to the round count, but it adds a checkpoint that creators frequently forget to schedule, and it is the single most effective guard against the gap between sample and production. The first article inspection article covers that step.

Typical Round Counts by Project Complexity

Round count is driven by complexity, and complexity in bags is driven by three things: structure, hardware and finish. A flat tote with standard webbing handles is a one-round product. A structured travel pack with a moulded back panel, custom zipper pulls and a printed lining is a three-round product at minimum. The mistake creators make is assuming their product is simpler than it is, usually because they are comparing it to a bag they own rather than to a bag being made for the first time.

Those counts are observational ranges from programmes our production team has supported, not industry-wide statistics. They carry the bias of our own project mix, which skews toward mid-complexity commuter and travel bags. A creator working on something genuinely novel should expect to be at the top of its band or above it, and should budget accordingly rather than treating the table as a promise.

The practical use of the table is to set the launch date. Multiply the expected round count by the per-round time, add the creator's own review time for each round, and add photography time, then check whether that fits before your intended launch. If it does not, the options are to simplify the product, to accept a later launch, or to compress the review turnaround. All three are legitimate; discovering the conflict late is not.

Where the 6 to 10 Working Days Go

The per-round window is often treated as a black box, which makes it hard to see where compression is possible. It is not a black box. The sequence below is the standard path for a mid-complexity backpack, and each stage has a characteristic duration. Understanding the sequence matters because the stages that can be compressed are not the ones creators usually try to compress.

1. Pattern and specification release: the tech pack is finalised, graded and released, typically the first day.
2. Material preparation: fabric, lining, webbing and hardware are picked, checked and cut, commonly one to two days.
3. Cutting and preparation: panels are cut, interfacing applied and components bundled, commonly one day.
4. Sewing and assembly: the unit is assembled by the sample line, commonly one to three days depending on complexity.
5. Hardware and finishing: zippers, buckles and branding are fitted, commonly one day.
6. Internal checking and photography: the unit is measured against the checklist and photographed, commonly one day.
7. Dispatch: the sample is packed and sent, which is where courier time enters and is outside the working-day count.

Two things stand out from that breakdown. First, the actual sewing is rarely the longest stage; preparation and checking take comparable time. Second, dispatch is not part of the 6 to 10 working days at all, because courier transit sits outside the workshop. A creator who forgets this will find that a round advertised as eight working days takes twelve calendar days to arrive, and the gap is not a delay, it is freight.

The compressible stages are specification release and internal checking, and both are on the creator's side of the line more than the workshop's. A complete tech pack released on day one removes the most common source of slippage. A clear checklist removes the second. The incompressible stages are sewing and material preparation, and trying to compress those is how defects get introduced into a sample that then needs another round.

One more source of elapsed time deserves naming because it catches everyone: courier and customs. A sample shipped internationally passes through customs at both ends, and a sample declared incorrectly or held for inspection can add days with no warning. Our team documents samples clearly and consistently for exactly this reason, and we advise creators to plan for the possibility rather than to assume the best case.

Complex Styles and the 12 to 15 Day Window

A complex style needs 12 to 15 working days per round rather than 6 to 10, and the additional time is not arbitrary. It comes from three places: tooling or fixture preparation, additional assembly stages, and a longer internal checking sequence. A moulded back panel, for example, needs its own forming step and its own cooling and fitting check, and a custom buckle needs to be fitted and tested before the sample can be completed.

The definition of complex in this context is worth making explicit, because creators consistently misclassify their own products. A style is complex if it has a moulded or formed component, if it uses hardware that is not a standard catalogue item, if it has a printed or woven pattern requiring registration approval, or if it has a convertible or modular mechanism. Any one of those puts it in the longer window; two or three put it at the top of it.

• Formed or moulded components: back panels, structured shells, moulded bases.
• Non-standard hardware: custom buckles, branded zipper pulls, bespoke frames.
• Registered artwork: printed patterns, woven labels requiring position approval, sublimated panels.
• Modular mechanisms: removable organisers, convertible straps, expandable compartments.

The strategic question for a first campaign is whether complexity is worth it. Our general answer is no for a first project, and the reasoning is about round count rather than about cost. A complex style needing three rounds at 12 to 15 working days is roughly six weeks of sampling before launch, and if one of those rounds fails, the launch moves by more than two weeks. A simpler style gives up some distinctiveness and buys back a month of schedule risk.

Where complexity is genuinely part of the product proposition, the mitigation is to sequence it. Develop the complex component separately and earlier, and let the main bag sample run in parallel on the standard window. That way a tooling iteration on the buckle does not hold up the whole sample. Our sourcing partner routinely splits complex programmes into parallel tracks for this reason. The hardware sampling article covers the component side in detail.

Where Rounds Get Wasted

A wasted round is one that produces no decision, and in our observation they come from four causes with remarkable consistency. None of them is a workshop failure; all of them are specification or process failures that happen before the sample is made. Listing them plainly is useful because all four are preventable at zero cost.

The third cause is the most damaging and the most common. A creator receives a sample, looks at it, and reports that something feels off without being able to say what. The workshop cannot act on that, so it makes its own interpretation, and the next round is wrong in a different direction. Two rounds are consumed and the actual problem is still not identified. A checklist converts a feeling into a measurement, and a measurement can be corrected.

The fourth cause, material availability, deserves special attention because it is easy to prevent and expensive to discover late. If the specified fabric is not in stock and a substitute is used for the sample, the sample is not representative of production, and approval based on it is not meaningful. Our team confirms material availability before releasing a round, and we tell creators immediately when a substitution has been made rather than hoping it will not be noticed.

There is a fifth cause that is less about process and more about governance: too many decision makers. A sample sent to a creator, a co-founder, a designer and a marketing lead produces four sets of feedback, frequently contradictory, and no decision. Our recommendation is to name one approver in writing before sampling starts, with others consulted but not deciding. The sampling communication article covers the questions to settle in advance.

The Approval Decision Framework

Approval should be a decision against criteria, not a judgement against taste. The framework below is what our team uses, and it has two parts: a checklist of measurable items, and a rule for what happens when an item fails. The rule matters as much as the checklist, because without it, every failure becomes a negotiation about whether it is acceptable, and those negotiations are how rounds get wasted.

Dimensions

Measured against the tech pack, with stated tolerance

Stitch density

Counted per inch at defined points

Hardware function

Every zipper, buckle and adjuster operated repeatedly

Colour match

Compared under standard lighting against the reference

Print registration

Position measured against the approved artwork

Load behaviour

Straps and attachment points loaded to the stated weight

The rule for failures is simple and worth stating explicitly: a critical failure, anything affecting safety, function or a dimension that changes fit, requires a new round. A minor failure, a cosmetic issue that does not affect function, can be carried into production as a correction with written agreement. Most disputes arise because that distinction was never established, so establish it before the first sample arrives.

Approval should also produce a physical artefact: a sealed reference sample. Once approved, one unit is retained and not used, sent to a photographer, or given away. That unit becomes the standard against which the first article of mass production is checked, and without it, the inspection has nothing to compare against. Our team treats the sealed sample as a controlled item with a recorded location rather than as a spare.

Finally, approval should be written. A short document recording the date, the checklist results, the accepted minor corrections and the identity of the approver takes ten minutes and prevents the most expensive failure in bag programmes: a production run that differs from what the creator believes was approved, with no record to arbitrate. The sample approval checklist article gives the measurement sequence.

Communication Cost per Round

Every round carries a communication cost that does not appear on any invoice and is frequently larger than the sampling fee. It consists of the time spent writing the brief, reviewing the sample, writing feedback, clarifying the feedback and confirming approval. For a first-time creator working alone, that is commonly several hours per round, and it falls at exactly the point in the schedule when the campaign page also needs writing.

This is why reducing round count matters more than reducing per-round time. A programme that needs two rounds instead of four saves perhaps two or three weeks of calendar and a substantial amount of creator attention, and that attention is the scarcest resource in a first campaign. The interventions that reduce round count are all cheap: a complete tech pack, approved artwork, a written checklist and a named approver.

• Batch your feedback: collect every comment into one document rather than sending them as they occur.
• Rank your feedback: mark each item critical, major or minor so the workshop knows what must change.
• Attach a photograph: mark up an image rather than describing a location in words.
• Confirm in writing: restate what was agreed at the end of each exchange, in one message.
• Set a review window: commit to returning feedback within a stated number of days.

The last item is the one that most directly compresses the calendar. If the creator commits to a 48 hour review window, three rounds consume six days of review time. Without the commitment, review time commonly stretches to a week per round, and three weeks of the schedule disappear into nobody's inbox in particular. Our programme coordinator asks for that commitment at the start of sampling and holds to it.

What Sampling Rounds Cost

Sampling has a direct cost and an indirect one, and creators usually budget only the first. The direct cost is the sample fee, which varies with complexity and material. The indirect cost is the calendar time, and for a campaign with a fixed launch date, that cost is measured in campaign weeks rather than in currency. The sampling cost article covers the fee side in detail.

The way to think about the trade is to compare the cost of an extra round against the cost of launching with an unrepresentative sample. An extra round costs a week or two and a fee. Launching with a sample that does not match production costs reputation, because the first thing backers do when units arrive is compare them against the campaign photography. In our observation the second cost is always larger, which is why our team argues for the extra round whenever there is genuine doubt.

There is also a cost to skipping rounds that shows up in production rather than in reputation: the change order. A creator who approves quickly and then requests a change during mass production is requesting the most expensive kind of change, because it affects material already cut and units already assembled. Every change requested after approval is a round that was skipped, and it costs several times more than the round would have.

For context on the other side of approval, mass production runs 35 to 50 days, inspected to AQL 2.5 to ISO 2859-1, shipped on FOB Xiamen terms with T/T 30/70 payment. Compared with that, sampling is a small fraction of both cost and time, which is the strongest argument for spending it properly rather than compressing it.

Compressing Rounds Without Cutting Quality

Compression is possible, but only in specific places. The sequence below is what our team actually does when a campaign is behind, and it works because it attacks the parts of the cycle that are process rather than craft. It does not touch sewing time, material preparation or checking, because compressing those produces a sample that cannot be trusted and therefore cannot be approved.

1. Overlap artwork approval with the first round: approve print and label artwork separately and in parallel.
2. Pre-book material before the tech pack is final: reserve fabric and hardware so preparation is not waiting on stock.
3. Review on video rather than waiting for the courier: a first look can happen the day the sample is finished.
4. Commit to a 48 hour feedback window: the single largest compression available to most creators.
5. Accept minor corrections into production: do not spend a round on a cosmetic item that can be fixed in the run.

The video review point is worth expanding, because it is the least intuitive and often the most valuable. Waiting for a sample to arrive internationally can add several days to each round. Reviewing it on video the day it is finished lets the creator give feedback immediately and lets the workshop start the next round before the physical sample has even landed. The physical sample still arrives for confirmation, but the clock has moved.

What should never be compressed is the checking stage. A sample that has not been measured against the checklist is not a sample, it is a prototype, and approving it transfers risk into mass production where it is far more expensive to correct. Our team would rather tell a creator that the launch date must move than approve an unchecked sample to protect the date.

One final compression lever is rarely used and very effective: reduce the number of variants in the first round. If the programme has three colourways, sampling one construction in one colourway first, then approving colour separately, separates the two decisions. Construction changes are expensive and colour changes are cheap, and mixing them in one round means a colour rejection can delay a construction approval.

Sampling Data You Should Log Yourself

Everything in this article is an observational range, and the most useful thing a creator can do is replace it with their own data. The log below takes minutes per round and, after one programme, produces a dataset that is far more predictive for that creator's product type than any category benchmark. Our team maintains a version of it for every programme we support.

The field that most often surprises creators is the feedback return date. Many creators believe their sampling is slow because the workshop is slow, and the log frequently shows that the workshop turned the sample around inside the stated window while the sample sat waiting for review. That is an uncomfortable finding and an extremely useful one, because review time is the easiest thing in the whole programme to change.

The defect classification field has a second use. If most defects are specification issues, the tech pack is the problem and the fix is upstream. If most are execution issues, the workshop needs a clearer brief or a different process. Those two diagnoses lead to completely different remedies, and without the log a creator will guess and usually guess wrong.

From Sample Approval to Production Release

Approval is not the end of sampling; it is the handover point. What follows is a short sequence that converts an approved sample into a released production order, and skipping any step in it reintroduces exactly the risk that the sampling rounds were spent removing. The sequence is short, but each item has a specific failure it prevents.

1. Seal the reference sample and record its location, so the first-article check has a standard.
2. Issue the written approval document with checklist results and accepted minor corrections.
3. Freeze the specification: any change from this point is a change order, not a revision.
4. Release material and book the production slot, against the confirmed quantity.
5. Run the first-article check at the start of production against the sealed sample.
6. Inspect the completed run to AQL 2.5 to ISO 2859-1 before release.

The freeze step is the one that most needs emphasis, because it is where campaigns quietly come apart. A creator who has finally approved a sample is often still thinking about improvements, and a suggestion sent during production is a change request whether or not it is framed as one. Our team requires change requests to be identified as such, costed in time and cost, and approved deliberately rather than absorbed silently.

The release sequence then runs on its own clock: 35 to 50 days of mass production, inspection to AQL 2.5 to ISO 2859-1, and shipment on FOB Xiamen terms with T/T 30/70. A creator who has logged their own sampling data can now predict the whole sequence with reasonable confidence, which is the practical value of keeping the log in the first place. The lead time analysis sets out how that production window breaks down.

Frequently Asked Questions

People Also Ask