Bag Campaign Post-Mortem: 10 Lessons From Bag Projects That Failed

Pricing was set before the bag existed

In most of the bag projects we have reviewed after the fact, the reward tier was published before anyone had written a bill of materials. The creator chose a number that made the funding goal look reachable, matched a competitor's public tier, and worked backwards from there. Every early decision, from the size of the funding target to the number of reward units promised, was then anchored to a price that no production file supported. When a real quotation finally arrived, the arithmetic had already been promised to thousands of backers.

The correction is straightforward but it costs calendar time. A tech pack should exist before the tier: exterior dimensions, capacity, shell fabric and coating, lining, foam or board structure, zip references, hardware finish, webbing, branding method, and packing rules. With that file in hand, a sourcing partner can return a cost breakdown at the locked order quantity. A campaign that skips this step usually discovers its real unit cost during the funding period, when changing the public price is expensive and changing the product is worse.

Price from landed cost per usable reward, not from product cost alone. Start with the unit price at FOB Xiamen, then add freight per packed carton, import duty and clearance, warehouse receiving, pick-and-pack, postage by zone, and a deliberate replacement allocation. Divide by the number of units that can actually be shipped to backers, not by the full 500 pcs ordered, because media units, retained references and service stock never become revenue. For a project that raised a plausible goal with a thin margin, that one denominator change can move the contribution by several dollars per reward.

Our team now refuses to sanity-check a tier that has no technical file behind it. As the programme coordinator, we ask for the drawings and the material references first, then tell the creator whether the published price survives contact with a real purchase order. The wider crowdfunding bag guide sets out the same sequence: specification, quotation, then public promise.

The sample was approved on a photograph

Photographs compress information that matters in soft goods. A studio image flattens panel curvature, hides how much a shell collapses when the bag is empty, and says nothing about zip smoothness under load, foam density, or whether a shoulder strap twists after ten minutes. Several failed projects we reviewed had approved their pre-production sample from photographs because the physical unit had not arrived before launch. The consequences arrived later, in bulk, across 500 pieces at once.

Physical approval means three artefacts held together: a signed sample, a dated measurement sheet taken at agreed datums, and a component list naming fabric, lining, zips, hardware and thread. The physical unit is the reference for the bulk run; the measurements are the reference for inspection; the component list is the reference for any substitution discussion. When one of the three is missing, disputes become opinion. A backer complaint about colour cannot be resolved against a render, and a claim about capacity cannot be resolved against a marketing page.

Approval should also include functional routines tied to the claims on the page. If the bag is sold as a three-day carry solution, pack the published list and check access at that load. If a laptop sleeve is promised, insert the maximum device envelope. If the webbing is load-bearing, hold the stated weight for a defined time and then inspect stitch lines. Sampling runs 6-10 working days per round, or 12-15 working days for complex builds, so two rounds are affordable when the structure is ambitious. The sample prototyping programme shows how to record those routines so they can be repeated at inspection.

• Sign a physical unit, not a photograph, before the deposit is released.
• Record exterior and internal dimensions at named measurement datums.
• Keep a dated component list with references for fabric, lining, zips and hardware.
• Repeat the load, zip and access routines from the marketing claims.
• Store the approved sample away from light and humidity for later comparison.
• Write the revision number on every sample, photograph and tech pack update.

One habit separates recovering projects from the ones we could not help: they treated the sample as a contract. If the bulk units did not match the signed reference, the conversation was about evidence rather than taste. That discipline is what makes an AQL 2.5 to ISO 2859-1 inspection meaningful later, because there is a physical standard to measure the lot against.

Colourways multiplied faster than demand

Colour looks free in a campaign video. Three shades of the same backpack appear to be one product with a filter applied. In production, colour touches every element: the face fabric, the lining, the webbing, the zip tape, the binding, the thread, the label, the logo ink or thread, and the carton print. Each of those items has its own material coordination. This is why the locked minimum is 500 pcs per style and per colourway: two shades of one bag are normally two commitments, not one order with a mixed box.

The failure pattern is arithmetically simple. A creator publishes three colour choices as a stretch goal at 500 pieces each, which is 1,500 units, three deposits under T/T 30/70, three balances before cargo release, three carton cubes, and three inventory positions that may not move at the same speed. If one shade sells out and another does not, the unsold colour still occupies cash and warehouse space. Several post-mortems we ran showed the second and third colours absorbing a large share of the total stock while contributing a small share of the pledges.

The safer structure is one hero colour in the first run, plus presented alternatives that are pre-priced but not promised. When a project wants a second shade, we agree the re-order specification, the material assumption and the decision deadline during the first programme, so an unlock later does not restart development. Keeping the construction identical also protects the inspection plan, because every unit in the lot follows the same approved bill of materials.

The tech pack was thinner than the marketing page

Post-mortems consistently show an imbalance: forty hours spent on the campaign page and four on the technical file. The page names materials in marketing language such as premium, rugged or water resistant, while the file says 600D polyester with a PU coating, or nothing at all. Marketing adjectives are not purchasable. A production floor can only act on measurable statements, and an inspection can only pass or fail a measurement.

A usable technical file for a 500-piece bag run contains dimensioned drawings with tolerances, a material sheet with references and weights, a trim and hardware list, stitch density and seam construction notes, branding artwork with coordinates and Pantone references, packing instructions, and the acceptance criteria for inspection. It also states what the bag must survive. Without tolerances, every later argument becomes a negotiation: a body 2 cm narrower is either within specification or it is not, and only the file can say which.

The practical fix is to write the file backward from the claims. For each promise on the page, ask what measurable statement supports it and what test would prove it. A claim about organisation becomes pocket count and internal dimensions. A claim about weather becomes a fabric and coating reference plus a seam note. A claim about weight becomes a target plus tolerance in grams. Our production team reviews this exercise with creators before any quotation, and the bag manufacturing overview explains how the file becomes a cutting marker, a sewing sequence and a packing rule.

There is a second benefit. A complete file makes quotations comparable. Three responses to a vague brief will each contain different hidden assumptions, and the cheapest is usually the one that assumed less. Three responses to the same technical file can be compared line by line, and the differences that remain are informative: material origin, trim references, packing density, sampling discipline and inspection inclusion.

The deposit calendar ignored payout timing

Several failed projects were solvent on paper and illiquid in practice. The funding total looked sufficient, but the cash arrived on a different schedule from the one the production programme required. Under T/T 30/70, the 30% deposit is due when the pre-production sample is approved and the purchase conditions are confirmed; the 70% balance is due at the agreed inspection gate before cargo release. If the deposit date falls in a week when cleared funds are not yet available, the reserved production slot moves.

The gap comes from three sources that campaign calculators routinely ignore: platform payout delay, failed or reversed pledges, and tax plus marketing deductions. A project can display a funded total and still have a materially smaller cleared balance. When the shortfall hits the deposit week, the options are all bad: delay the slot, pay late and lose priority, or reduce the order below the quantities the reward tiers promised.

The control is a weekly cash calendar from sample approval through to final-mile dispatch. Mark sample fees, the 30% deposit, packaging or tooling charges, the 70% balance, freight booking, duty and clearance, warehouse receiving, and postage. Add a contingency line. On a 500-pcs order at a known unit price, the deposit is a four- or five-figure obligation landing on a specific week, and the balance is roughly two and a third times that amount. Knowing both dates before launch is the difference between a schedule and a surprise.

We also ask creators to separate the marketing budget from the production reserve in their own accounts. Ad spend is discretionary and can be paused; the 70% balance is contractual and cannot. Projects that treated the reserve as a general pool were the ones that arrived at the inspection gate without the funds to collect their goods. That situation is expensive: storage at origin, a missed seasonal window, and backer updates written under pressure.

The clock was counted from campaign close

The production 35-50 days, and it starts only after two conditions are met: the deposit has been received and the pre-production sample has been approved in writing. Campaign close starts nothing. Projects that promised an eight-week delivery from the final day of funding had, in effect, promised a schedule with no sampling, no approval, no inspection, no ocean transit and no fulfilment time inside it. Those promises then had to be walked back in public.

The realistic sequence has more stages than the production window itself. Sampling takes 6-10 working days per round, or 12-15 for complex builds, and two rounds are common. Creator review and courier transit sit outside those working-day figures. After bulk production there is inspection, cargo release, ocean or air transit, import clearance, inland delivery, warehouse receiving, and pick-and-pack in the destination country. A delivery month published to backers should be calculated from the far end of that chain, not from the near end.

Our advice on public dates is deliberately conservative: publish a month, not a day, and hold the later part of that month in reserve. A timeline can absorb fast approvals and stock materials; it cannot absorb removing inspection or compressing a sewing sequence. Campaigns that promised tightly and delivered late damaged the relationship with backers more than campaigns that promised a conservative month and delivered early. The QA and QC process is the stage most often squeezed first, and it is the one that protects the 70% payment decision.

Freight was estimated from a render

Freight is quoted on chargeable weight, which for bags is usually volumetric rather than actual. A backpack held open by foam and board, or a structured duffel, occupies carton space that a render says nothing about. Projects that estimated freight from a three-dimensional drawing, or from a rough guess of 20 pieces per carton, later discovered a cube perhaps 30% larger, which multiplied across every carton in the shipment.

The control is a packed trial carton built from the approved sample: pieces per carton, exterior dimensions, gross weight and net weight. Those four numbers drive the freight quotation, the pallet plan, the warehouse receiving cost and the pick-and-pack rate. Ask for them before bulk production, not after, because the carton specification can still be adjusted while the packing rules are being written. Changing carton dimensions after 500 pieces are boxed is a manual re-handling exercise.

Under FOB Xiamen, the origin-side responsibilities are handled up to the agreed on-board point; the buyer arranges main freight, insurance, import clearance, duty and inland delivery. Therefore, the comparison a creator needs is not FOB against FOB, but total landed cost per usable reward under each mode. Air freight can rescue a small urgent group of units; sea freight is the economic default for the whole lot; a split can serve both. Compare like with like, and require every quotation to state what it excludes.

Two further estimates are routinely missing: duty rate by destination, and the receiving plus storage rate at the 3PL. A duty classification error can be the largest single surprise in the whole programme, and it is the buyer's responsibility under the locked trade term. The rewards fulfilment overview covers the handling side; our team asks for the trial carton data and the destination before quoting, so the freight line in the budget is an estimate with a basis rather than a hope.

Inspection was treated as a formality

The locked inspection standard is AQL 2.5 to ISO 2859-1. AQL is a statistical lot-acceptance framework: it defines how many units are drawn from the lot and how many defects in each category the lot may contain before it is rejected. It is not a promise that every bag is perfect, and it is not a substitute for a product-specific checklist. Projects that treated inspection as a rubber stamp had no checklist, no physical reference on the line, and no agreement about who pays for rework.

A working inspection for a bag run names the sampling plan and the defect categories first: critical, major and minor. Critical defects include safety or functional failures such as a broken load path. Major defects are visible or functional problems that a backer would reject, such as a misplaced logo, a failed zip or a crooked panel. Minor defects are workmanship issues that do not affect use. Without those definitions, inspectors and creators argue about the same unit from different standards.

• Agree the sampling plan and defect classification before the inspection is booked.
• Use the signed sample and measurement sheet as the physical reference on the line.
• Add an in-line check when the first 50-100 pieces are finished.
• Record responsibility and timing for rework and re-inspection in the purchase order.
• Reserve replacement stock separately from units intended for later sale.
• Keep the inspection report with the 70% payment decision, not in an inbox.

Inspection also decides money. Under T/T 30/70, the 70% balance falls at the inspection gate before cargo release, so the report is the document that authorises the largest payment in the programme. A campaign that skips it has paid in full for goods it has not verified. For a structured review of how inspection fits the wider production programme, the Kickstarter QA and QC guide is the companion to this article.

Fulfilment was designed after packing

In several post-mortems, the goods arrived at the warehouse before anyone had decided how a backer's parcel would be assembled. Kitting decisions, barcode placement, inserts, polybag requirements, carton labels and bundle rules were improvised on the receiving dock. Every improvised touch has a cost, and 500 small handling charges can exceed the saving won from a material negotiation months earlier.

The fix is to write the packing and fulfilment rules in the same document as the tech pack. State pieces per carton, units per polybag, label position and content, insert placement, and whether accessories are kitted at origin or at destination. If the reward is a bundle, say whether the warehouse receives a finished kit or two items to combine. A production line can apply a barcode label in seconds during packing; a 3PL applies the same label at a per-unit handling rate after the goods have crossed an ocean.

Data discipline matters as much as physical packing. Survey responses must close on a published date, addresses must be validated before dispatch, and SKU counts in the warehouse system must match the purchase order. Projects that shipped before their survey closed paid twice for the same addresses. The rewards fulfilment guide explains the sequence in detail; our production team simply asks for the packing rules early enough to build them into the carton specification.

There is a cash consequence too. Fulfilment is the longest tail in the budget: postage by zone, parcel count, exception handling, reshipments and customer service hours. A project that spends its reserve on production and arrives at dispatch with nothing left is the pattern we see most often in reviews. Reserve the final-mile money at the same time as the deposit, not after the balance has been paid.

No plan for the shortfall or the overrun

The last lesson is a planning symmetry. Projects prepare for success and not for the two outcomes that actually follow a 500-piece first run: the campaign lands short of the quantity needed, or it lands well above it. In the short case, the reward tiers have promised more units than the project can fund, and the minimum order still applies at 500 pcs per style and colourway. In the overrun case, demand exceeds the first run and a re-order must be placed while the first is still on the water.

For the short case, decide before launch what happens: reduce the reward cap, extend the campaign, or fund the minimum from other sources and deliberately allocate the balance across late pledges, wholesale samples and later direct sales. The worst outcome is discovering the shortfall after the tiers are public. For the overrun case, prepare the re-order file during the first run: signed sample, tech pack with revision history, component list, artwork, colour references, carton specification and the inspection checklist. A re-order that starts from a complete file avoids a second development cycle.

Re-order economics are not the same as campaign economics. Demand after a campaign is usually smaller, more spread out, and more sensitive to delivery time, while the minimum stays at 500 pcs per style and colourway. Plan the channel before committing: post-campaign pre-orders, wholesale conversations, corporate programmes or direct stock. A project that orders 1,000 pieces because the first 500 sold out should be able to name the customers for the second 500 before the deposit is sent.

Our role in both cases is to keep the options visible early. As the campaign bag specialist on the file, we hold the specification, the sample archive and the pricing logic, so a re-order or a reduced first run can be quoted from evidence rather than reconstructed from memory. That is also why we ask for the same records at the start: they are cheap to keep and expensive to recreate.

Frequently Asked Questions

People Also Ask