Custom Bag Prototype: Digital Mockup, Physical Sample, Golden Sample

Creators use the word prototype for everything from a Photoshop render to the sealed reference bag sitting in a QC office, and the confusion costs money. The three artefacts are genuinely different objects with different purposes, different costs and different authority. Knowing which one you are holding at any moment is the difference between a campaign that is ready and a campaign that looks ready.

Three artefacts, three different jobs

A digital mockup is a visualisation. It can be a flat drawing, a 3D render or a photographic composite, and its job is to communicate an idea to an audience or to a production team. It carries no construction information and no authority. It is the cheapest of the three by a wide margin and the fastest to change.

A physical sample is a real bag built to a specification. Its job is to prove that the design can be made and to reveal what the drawing could not: weight, hand feel, drape, stiffness, how a zip runs under load, whether a compartment actually accepts the object it was drawn for. It is the artefact that generates decisions.

A golden sample is a physical sample that has been approved and retained as the standard. Its job is not discovery but authority: it is the object the bulk lot is compared against, and it is what the AQL 2.5 to ISO 2859-1 inspection references when an inspector accepts or rejects a shipment. It is usually signed, dated and stored.

The three form a sequence rather than a choice. A campaign normally uses a mockup to test interest, a physical sample to develop the product, and a golden sample to control production. Problems start when one is asked to do another one's job, which is what happens when a creator tries to launch from a render or, worse, tries to inspect a shipment against a render.

Our bag sampling and prototyping guide places all three inside the wider programme, from first visual to retained reference. This article is about the boundaries between them and where those boundaries get crossed by accident.

The digital mockup: what it is good at

The mockup is the most underrated tool in a crowdfunding campaign when it is used for its actual purpose, which is communication. It lets you test ten colourways against an audience before committing to one, it lets you show a feature that does not exist yet, and it lets you collect email signups before any money has been spent on tooling or materials.

It is also the fastest way to align a team. A render with a few dimension callouts resolves arguments about proportion in minutes, because everyone is looking at the same picture. Trying to do the same thing in words produces three different mental images and a sample built to none of them.

• Audience testing: ten colourways, one afternoon, no material commitment.
• Crowd funding page layout: imagery before the product exists.
• Internal alignment: one picture everyone agrees on.
• Material shortlisting: a visual sense of how a fabric might read.
• Pitching to retailers, partners or a logistics provider before tooling.
• Variant planning: showing stretch goal options without building them.

For pre-launch marketing the mockup is genuinely the right tool. A landing page with a good render will collect signups, and signups are the number that decides whether a campaign is worth launching at all. There is no reason to spend on sampling before that number exists.

There is a discipline that makes renders more useful downstream: draw them to scale and put the dimensions on the drawing. A render drawn at true proportion with millimetre callouts can be handed to a pattern maker and actually used. A render drawn to look good usually cannot, because the proportions have been adjusted for the picture.

Where a render lies to you

A render is a picture of a bag that does not exist, and it is made by someone choosing lighting, material response and geometry. Every one of those choices flatters the design, because that is the point. The problem is not that renders lie, it is that they lie in specific and predictable ways that a creator can learn to spot.

Material response is the biggest one. A render shows fabric as smooth, taut and uniformly lit. Real fabric has slubs, weave texture, seam puckering and a tendency to sag under its own weight in a way no render reproduces by default. A bag that looks structured in a render frequently looks soft in person, and that gap is the single most common backer complaint we hear about reward bags.

Geometry is the second category. A render can show a curve that cannot be sewn, a corner that cannot be turned, or a panel count that would require a seam in a place the designer did not intend. These are not flaws in the render, they are consequences of the fact that sewing is a sequence of operations with physical limits.

Colour is the third and the one with the largest commercial consequence. A render is displayed on a screen calibrated to nobody in particular, and the backer's screen is different again. Then the fabric is dyed to a standard and the result is a third colour. The only fix is a physical standard: a lab dip, a Pantone reference or an approved swatch. Our render or real sample article covers the photography decision in more detail.

The physical sample: what it answers

A physical sample answers the questions a render cannot, and it answers them at a cost of 6-10 working days per round for a standard build or 12-15 working days for a complex one. That is a real investment of calendar, which is exactly why the questions asked of it should be written down before it arrives.

The first answer it gives is dimensional. Does the bag accept the object it was designed for, with the intended clearance. A laptop compartment, a camera insert or a shoe compartment is defined by the thing inside it, and only a physical piece can confirm the fit. Drawings are optimistic by nature, because a drawing has no tolerance.

The second answer is weight, and this one has a direct commercial consequence. The packed weight of the sample determines the per backer shipping charge, and the packed cube determines the freight cost. A sample that lands 150 g heavier than assumed across 500 units is a measurable budget error that no render would have caught.

• Fit: does the named object go in, and come out, with clearance.
• Weight: packed grams, which drive reward shipping.
• Cube: packed dimensions, which drive freight and warehouse cost.
• Construction: can it actually be assembled in this sequence.
• Hardware behaviour: zip run, buckle release, strap adjustment under load.
• Decoration: how the method behaves on this specific material.

The third answer is the one creators value most and expect least: the sample tells you what to change. Design decisions that seemed obvious on screen look different in three dimensions, and a pocket that read well in a render is often in the wrong place once the bag is on a body. That discovery is the entire point of sampling, and it is why one round is almost never enough.

The golden sample: the reference that governs bulk

The golden sample is not a third build, it is a status applied to a build. When a physical sample is approved as the standard for production it becomes the golden sample: retained, marked, dated and stored. From that moment it is the authority against which 500 units are judged, and it is what an inspector holds when deciding whether a defect is acceptable.

Retention is the operative word. A golden sample that ships to the creator and lives on a shelf in another country cannot be used for inspection. The piece has to stay where the production and inspection happen, and the creator should receive a second piece or a detailed photographic record for their own reference.

It also has to be singular. If two approved pieces exist and they differ, there is no standard. Every revision must either supersede the previous golden sample explicitly, with a date and a revision number, or be rejected. A file with three approved samples in it has no standard at all, and that is a dispute waiting for a shipment.

1. Approve the piece in writing, naming revision and date.
2. Have it retained at the production site, not shipped away.
3. Mark it physically so it cannot be confused with a production unit.
4. Record its key measurements and references in the specification document.
5. State the inspection standard: AQL 2.5 to ISO 2859-1 against this piece.
6. Require a pre-production sample built on the bulk line for a new construction.

The golden sample also fixes the commercial baseline. Bulk production runs 35-50 days from deposit and approved sample, and the approved sample in that sentence means the golden sample. Without it the clock has not started, and a shipment inspected without it has no defensible acceptance criteria.

Choosing the right artefact at each stage

The practical question is which artefact you need right now, and the answer depends on what decision is in front of you. Working through a campaign in order, the sequence is fairly consistent, and having the wrong artefact at a given stage either wastes money or leaves a decision unsupported.

Before launch, while you are still testing interest, the mockup is correct and a physical sample is premature. During development, the physical sample is the only thing that works. At the point of bulk release, the golden sample is mandatory. After delivery, the golden sample is what settles any dispute about whether the goods matched the promise.

Note the shipping estimate row specifically. Setting a reward price and a per backer shipping charge from a mockup is the single most expensive use of the wrong artefact, because the error is multiplied across every unit and cannot be recovered after the campaign closes.

There is one exception worth knowing. A very experienced creator with a construction that is close to an existing product can sometimes estimate weight from a known comparable, and that is a legitimate shortcut. It is still a shortcut, and it should be verified against a physical piece before the reward price is published.

What each artefact costs in time and money

The cost profile of the three artefacts is not linear, and understanding it helps allocate a budget sensibly. A mockup costs designer time and nothing else. A physical sample costs a sampling charge plus a calendar window. A golden sample costs no extra money but it costs authority, which is to say it costs you the right to change your mind.

The sampling charge is normally credited against the bulk order, so the true cost of a physical sample is closer to the calendar than to the invoice. That is worth stating plainly because creators sometimes skip sampling to save money when the real saving is small and the risk is large.

• Mockup: design time only, hours to days, free to change.
• Physical sample: sampling charge, often credited to bulk, 6-10 working days, 12-15 for complex builds.
• Golden sample: no additional charge, but every later change requires a new approval.
• Pre-production sample: a further build on the bulk line, worth it for a new construction.
• Bulk: 35-50 days from deposit and approved sample, 500 pcs per style and colourway.
• Inspection: AQL 2.5 to ISO 2859-1 against the golden sample.

Budgeting three physical rounds is the norm. Three rounds at eight working days each is about twenty four working days, or roughly five calendar weeks, and that sits before the bulk window rather than inside it. A campaign that budgets one round discovers the shortfall after the launch date is public.

Commercial terms run alongside: FOB Xiamen and T/T 30/70, with the deposit releasing the production slot. Our sampling cost article sets out what the sampling charge typically covers and when it comes back to you.

Handing a mockup over for production

Most creators arrive with a render and not a tech pack, and that is fine, but something has to convert the picture into a specification. That conversion is where most of the avoidable sampling rounds are created, because an unspecified detail is resolved by assumption and the assumption is often not what the creator pictured.

The conversion has a fixed list. Every visible panel needs a material and a colour. Every visible piece of hardware needs a size, type and finish. Every dimension needs a number. Every closure needs a type. Every seam needs a construction. Every logo needs a method, a size and coordinates. That is the whole job.

1. Supply a dimensioned drawing, not only a perspective render.
2. Name each panel material with weight, coating and finish.
3. Name each hardware item by size, type and finish.
4. State the construction for the main seams and the strap anchors.
5. Fix the colour to a numbered or physical standard.
6. Give decoration coordinates measured from a stated edge.

Our campaign bag specialist will normally take a render and return a first draft specification with the open questions marked, and that draft is the fastest route from picture to sample. Creators who send a render plus a written list of the things they care about most get better first drafts than creators who send a render alone.

One warning about 3D files. A 3D model is useful for proportion but it is not a production file, because it has no seam allowance, no material thickness and no assembly sequence. Sending one is helpful context, not a substitute for a pack. The 12 section tech pack guide shows what the production document needs to contain.

Using all three in campaign marketing

The three artefacts are not only engineering tools, they are marketing assets with different credibility values, and using them in the right order on a campaign page builds a narrative that backers find convincing. The sequence that works is honest about what exists at each point.

Early renders establish the concept and the colour range. A first physical sample establishes that the thing is real, and photographs of it carry far more weight than any render. A golden sample establishes that production is controlled, and a photograph of the retained, marked reference on a QC bench is a surprisingly effective trust signal.

• Renders for concept, colour options and stretch goal variants.
• First sample photographs for the main gallery and the feature explanations.
• Golden sample images for the production and quality control section.
• Material swatch close ups for the material story, taken from real fabric.
• Hardware macro shots for durability claims, taken from the real build.
• Packed carton shots for the fulfilment section, taken with real packing.

The trust signal works because experienced backers know the difference. A page where every image is a render reads as pre-production. A page with a real sample handled, opened and loaded reads as a product that exists. We have supported more than 20 crowdfunding campaigns and the pages that show real objects consistently field fewer sceptical questions in the comments.

There is one rule that protects you: never present a render as a photograph of a real product. Backers notice, and the accusation of misrepresentation is far more damaging than the honest caption concept render. Our pre-production sample guide explains what a production-intent piece is and when you can legitimately show one.

Common mistakes with each artefact

Each artefact has a characteristic failure mode, and all three are avoidable. Naming them is most of the work, because once you know what to look for the mistake announces itself.

With mockups the mistake is treating them as evidence. A render that has been tested with an audience proves that people like a picture, not that a bag can be built at a price. Treating positive feedback on a render as product validation is how a campaign commits to a design that cannot be made at the reward price.

• Mockup: treating audience approval of a render as product validation.
• Mockup: drawing to look good rather than to scale, making the file unusable downstream.
• Physical sample: reviewing by feeling instead of by measurement.
• Physical sample: changing the artwork mid round, forcing an extra build.
• Golden sample: shipping it away, so it cannot be used for inspection.
• Golden sample: keeping three approved pieces, which means having no standard.

With physical samples the mistake is the soft review. A creator looks at the bag, likes it overall, notices one thing, and sends feedback with one item. The structural issue that would have been caught by measuring against the pack goes unnoticed, and it resurfaces in bulk across 500 units where it is expensive.

With golden samples the mistake is casualness about status. An email that says approved without naming the revision, or a second approved piece created after a small change, quietly destroys the standard. Then a shipment arrives with a defect and nobody can say whether it matches, because there are two answers to the question of what was approved.

The remedy across all three is the same: write it down. A dated approval, a named revision, a measured review and a retained physical reference cost an hour each and remove most of the arguments that otherwise consume weeks.

A decision table for creators

When you are unsure which artefact to ask for, the question to answer is what decision you are trying to make. The table below maps the common decisions to the artefact that can actually support them, and to the consequence of using the wrong one.

Read down the middle column and the pattern is clear: renders support decisions about desire, physical samples support decisions about reality, and the golden sample supports decisions about conformity. Mixing those categories is where campaigns get into trouble.

The practical plan for a first campaign is therefore short. Build a mockup, test demand, and collect signups. Then convert the mockup into a tech pack and run three physical sampling rounds at 6-10 working days each. Then approve one as the golden sample and release production 35-50 days, inspected to AQL 2.5 to ISO 2859-1 at 500 pcs per style and per colourway.

That plan is not exciting and it is also not negotiable. Every campaign we have supported that skipped a step paid for it later, and the payment was always larger than the step would have cost. The artefacts are cheap; the consequences of confusing them are not.

Frequently Asked Questions

People Also Ask