Sizes, Specified: How Dimensions, Capacity and Fit Survive Production
Sourcing guide · Written by the Sales Engineering Desk · Updated September 2026
Direct answer: a bag size survives production when it is specified three ways at once — external dimensions with a measurement protocol (which points, which tension, which state), internal capacity by compartment (device fit named by inch class, usable liters separated from shell liters), and tolerances in the tech pack that state how much variation is acceptable — and programs quote from 2,000 pieces per style on FOB terms with the size spec frozen at pre-production sample approval.
A tape measure, held wrong, costs a container. Two factories measuring “the same” tote — one across the seam allowances, one inside them, one with the gusset collapsed, one expanded — produce four different products from one drawing. Dimension language looks like the simplest part of a bag program and behaves like the slipperiest, because every number silently depends on a protocol nobody wrote down. This guide is that protocol: how buyers specify size so that what ships matches what was approved. It anchors the specification chapters of the backpack sourcing hub.
How should a buyer read a bag size chart?
A trustworthy bag size chart answers three questions that most charts leave unstated. Measured what — external shell dimensions, or usable interior, and at what state (expanded, collapsed, stuffed, empty)? Measured where — across which points, with seams and gussets doing what? And varying how much — what tolerance applies to each dimension, because a chart that promises exact millimeters on a sewn soft good is promising something sewing cannot deliver. Charts that specify all three are production documents; charts that list bare numbers are marketing. When buyers compare suppliers’ custom bag sizes manufacturer quotes, the fastest filter is asking each to state its measurement protocol in one sentence — the answers separate factories that engineer dimensions from factories that round them. Buyers inheriting a chart from a catalog have a shorter test: check whether the numbers include handles in height, whether the gusset states its measurement state, and whether any tolerance is mentioned at all — three absences mean the chart was drawn for a webpage, not for a production floor, and every number on it needs re-measuring under a protocol before it touches a purchase order.
What do capacity liters actually mean — and not mean?
Backpack capacity liters and tote bag capacity liters are the most quoted and least standardized numbers in this trade. What liters measure is enclosed volume; what they do not measure is usable space — the difference being pockets, padding, hardware, seam allowances and the geometry of how a bag actually carries load. Two 20-liter backpacks can differ by a carry-on’s worth of real packing room once frame structure and pocket layout enter. The professional move is to split the claim: shell liters (the enclosed volume) and usable liters (what a packing test demonstrates), with the packing test described — because “fits a 16-inch laptop, a binder and a lunch” is a specification, while “20L” is a vibe with a unit.
Laptop bags carry 179 of 2,022 de-duplicated US B2B bag-sourcing queries — 8.9% of measured demand — and no other family is so dominated by a single dimension question: will the device fit.— YUEOU Search Desk, 2026 US Wholesale Bag Search Report
How does laptop compartment sizing really work?
Laptop compartment sizing is device-fit engineering, and the inch-class language needs translation. A “15-inch” compartment is not fifteen inches of anything — it is sized to the class of devices around that diagonal, with clearance for the population of actual laptops in that class, plus corner protection, plus the closure’s stack height. Buyers writing a laptop bag size guide 15 inch program should specify against measured device dimensions (width, depth, thickness at the thickest point including feet and hinges), not screen diagonals — and should request a fit test on the actual device models their program targets, because the difference between “fits 15-inch laptops” and “fits these three SKUs” is the difference between a spec sheet and a dispute. The family taxonomy around these programs — sleeves, messengers, backpacks — is mapped in our business carry family tree.
What belongs in a dimension measurement protocol?
A bag measurement guide used in production has four fixtures. Reference state: empty and closed, or loaded to a stated weight — because soft goods measure differently under load, and the choice must be written. Reference points: width at the widest seam, height including or excluding handles and straps (stated), depth at the gusset center, expanded or collapsed (stated). Method: flat on a surface, tension-free, corners squared — the discipline that makes two factories agree. And rounding: dimensions stated to the sensible precision sewn goods hold, which is not the third decimal place. Learning how to measure a bag for production is a thirty-minute discipline that pays for itself at the first inspection, where every dimension dispute is settled against the protocol — if one exists.
The recurring mistakes are few and expensive. Quoting screen diagonal as compartment size — the compartment must fit the device, not the marketing number. Measuring height with handles up because the photo looks better — then failing the retailer’s fixture. Copying a competitor’s chart without its protocol — inheriting numbers measured under rules nobody in your chain follows. Stating gusset depth at full expansion — shipping a bag that stands poorly at the state customers use. And the quietest one: revising a dimension verbally after sample approval, so the golden sample, the tech pack and the memory disagree by the inspection date. Every item on this list is prevented by the same object — a written protocol attached to numbers.
How do tolerances keep dimensions honest at volume?
Tolerances are where dimension talk becomes contract. A professional tech pack carries a tolerance table — the dimensions our documentation chain confirms at sampling — stating the acceptable band per measurement class: structural dimensions held tighter, soft geometry (gussets, gathers) held looser, and elastic elements specified by behavior rather than by rest-state size. The bands exist because sewing is a process with real variation; pretending otherwise just moves the variation into arguments. Programs should expect backpack dimensions guide-grade documentation from any factory quoting at volume, and should test it early: ask what tolerance the factory proposes for your most critical dimension before you state yours. A factory that proposes sensible bands unprompted has done this before; a factory that accepts “exact” has not understood the question.
| Dimension class | Tolerance philosophy | Consequence when missed |
|---|---|---|
| Device-fit compartments | Tightest band; fit-tested per device class | Major defect — the product’s promise fails |
| Structural shell dimensions | Firm band, protocol-measured | Fit-in-packaging and shipping failures |
| Soft geometry (gussets, gathers) | Looser band, state-stated | Capacity claims drift |
| Elastic and stretch elements | Behavior-specified, not rest-size | Function degrades silently |
How do tote programs specify the dimensions that matter?
Totes concentrate their fit questions in three numbers. A tote bag dimensions guide worth quoting from separates them: the mouth opening (what fits through, which governs document and box formats), the base footprint (what stands upright inside, which governs retail and packaging programs), and the carry clearance (handle drop, which governs shoulder vs hand use). Tote bag capacity liters claims follow the same split-audit as backpacks — shell volume versus demonstrated usable space — with one tote-specific honesty: gusset depth is the number catalogs exaggerate most, because an expanded gusset measures impressively and carries awkwardly. Specify gusset at the state the program actually uses, and the number becomes engineering instead of aspiration.
Which cross-market size differences catch programs out?
Size is a local language. Document formats differ — Letter markets and A4 markets put different demands on the same “document tote.” Device populations differ — laptop classes popular in one region skew a half-inch in another, which matters exactly at compartment clearance. Trolley systems differ — travel programs specifying carry-on compliance must name the airline standard they target, because the honest chart lists the rule it follows. And packaging differs — a bag sized to stand upright in a US retail fixture may lie awkwardly in a EU planogram, a detail that belongs in the program brief, not in a post-shipment discovery. Programs serving multiple markets treat size as a per-market attribute of one construction, which is cheaper than treating it as a surprise.
How does the size spec travel through the order?
The same trail every spec travels, with dimensions riding shotgun. At quotation: dimensions and protocol stated, so the price reflects the size actually being built. At sampling: the pre-production sample measured against the protocol, the numbers recorded in the approval, the golden sample retaining those numbers as the order’s definition. At inspection: final random inspection under the standard AQL plan — Critical 0, Major 2.5, Minor 4.0 — with dimensions outside tolerance classified by consequence, because a laptop compartment that misses its class is a major defect and a decorative pocket that runs a centimeter long is not. The full sampling chain is mapped in our first-order guide inside the trade terms hub; the size discipline simply rides the rails already there, from 2,000 pieces per style at quoting to a measured container at loading.
For a first order, the whole size discipline compresses into one week. Day one: send the reference product or device models with your target dimensions and ask for the factory’s protocol and proposed tolerances — the reply quality is the qualification. Mid-week: measure the returned spec against your own numbers, reconcile every gap in writing, and freeze the tolerance table into the tech pack. At sampling: measure the pre-production sample with the agreed protocol, record the numbers on the approval itself, and let the golden sample carry them forward. From there the inspection plan enforces what the week built — and dimension disputes, which once consumed emails, become arithmetic against a document both sides signed.
Specifying sizes for a program? Send the dimensions that matter — device models, format requirements, capacity expectations — through the quote form — the reply includes a proposed measurement protocol and tolerance table with the quotation, so the first sample is measured against something both sides wrote — which is the entire art of size in this trade, where a millimeter argued early is a container saved late, reduced to one habit: numbers that carry their rules with them, from the first email to the container door, in every market the program serves, at every quantity the calendar demands.