Personalized Dog Carrier Fitting Guide

Pet carrier production desk · Updated 2026-10-07 · 14 min read

For a custom cat carrier the ventilation question is where the mesh sits, not how much there is: high side panels plus a shaded front zone keeps airflow without visual exposure.

Three Measurements That Decide Everything

Everything else in this guide is a refinement of three numbers, and if you take only three, take these. They are taken on a standing dog on a hard floor, by one person, with a soft tape held flat and snug but not tight. A tape held at an angle reads long. A tape pulled tight reads short by 2 to 4 centimetres, which is more than the entire tolerance you are working to.

The first is chest girth: the circumference of the ribcage at its widest point, which in most breeds sits immediately behind the front legs. Take it on a normal exhale, not a deep breath, and take it twice. Girth is the measurement that decides whether a dog can turn around, and it is the one most often omitted from a specification because it is harder to take than length.

The second is back length, measured from the withers — the highest point of the shoulder, where the neck meets the back — to the base of the tail, following the spine rather than a straight line over the body. On a long-backed breed the difference between the two is 4 to 6 centimetres and it matters. Back length decides whether the carrier floor is long enough for the dog to lie in sternal recumbency, which is the resting posture that actually matters for a journey of more than twenty minutes.

The third is standing shoulder height, taken from the floor to the top of the scapula, not to the top of the head. Head height is irrelevant to fit because a dog does not stand in a carrier with its neck fully extended; the shoulder line is what has to clear the closure. This measurement decides interior height, and getting it wrong produces the classic failure where the dog fits lying down but cannot be zipped without its shoulders being pressed.

Record all three in centimetres to one decimal, together with the dog's weight, breed or mix, and age. The weight is not used for sizing; it is used to check whether the girth is plausible, because a girth that implies a much heavier or lighter animal than the scale reports is a measurement error and should be retaken.

Weight Bands Lie: Why Dimensions Beat the Scale

Almost every carrier on the market is sold by weight, and almost every sizing complaint traces back to that decision. Weight is a proxy, and it is a poor one, because body density varies enormously across the shapes a carrier has to accommodate.

Take two dogs at 9 kilograms. One is a compact terrier with a deep, narrow chest and a short back; its girth might be 52 centimetres and its back length 38. The other is a leggy small sighthound with a shallow chest and a long spine; its girth might be 46 and its back length 52. Weight says they are the same animal. Circumference and length say they need interiors that differ by roughly 35 percent in plan area, and no single bag serves both well.

The consequence for a range is that a weight-banded ladder systematically misfits the extremes of each band. A band labelled "up to 8 kg" is built around the median animal in that band, which is a dog of about 5.5 kilograms, and it is too generous for the 3 kilogram end and too tight for the 8 kilogram end. The 8 kilogram dog is the one that gets returned. Building the ladder from girth and length instead — bands of 8 centimetres of girth and 6 of length — moves the median closer to the upper edge of each band, which is where the commercial volume and the complaints both are.

Weight still has a job, but it is a different job: it sets the structural specification. A floor panel, a handle root and a shoulder strap are engineered to a load, and load is mass, not circumference. So the honest specification has two independent inputs — dimensions set the interior, weight sets the structure — and any programme that uses one for both is compromising somewhere it has not noticed. The sizing work behind breed weight bands shows how the two are reconciled across a ladder.

Four Body Morphologies Worth Designing For

Breed is a weak predictor, because mixed-breed dogs dominate most markets and because two dogs of the same breed can differ more than two dogs of different breeds. Shape is a strong predictor. Four canonical morphologies cover the overwhelming majority of dogs a carrier will carry, and each one stresses a different dimension of the interior.

The compact type — short back, deep chest, moderate height — is the easiest to fit and the most forgiving of a slightly small interior, because the animal curls. It stresses height least and girth most. The long-backed type — a spine much longer than the height, a shallow chest — is the one that breaks length-based ladders, because it needs floor length that looks disproportionate to its weight. It also needs a floor that does not sag, since a long spine supported at two points flexes in the middle. We build these interiors with a stiffer floor panel rather than a larger one.

The broad-chested type — high girth relative to length, heavy forequarters — is the most common cause of a "too small" return on a bag that measures correctly. The animal fits the length and cannot rotate, because the girth envelope is circular and the interior is rectangular; the diagonal of the interior is what has to clear the circumference, not the width. The fix is a wider floor rather than a longer one, and a gusset that lets the side panels bow outward under load.

The tall type — leggy, with shoulder height close to or above back length — is the hardest, and it is where a structured carrier fails and a soft one succeeds, because the shoulder needs vertical clearance at the closure rather than floor area. For these animals the specification is interior height plus a closure that opens past the shoulder line, which usually means a full-perimeter zip rather than a top flap. We cover the dedicated geometry in long-bodied and tall breed interiors.

A fifth group exists and should be acknowledged: brachycephalic breeds, whose airway anatomy makes ventilation and thermal behaviour a fit issue as much as a design one. For these, the interior can be correct and the carrier still wrong, which is a material question rather than a dimensional one.

The Fit Test Protocol We Run Before Approval

A fit test is not an opinion and it is not a photograph. It is a short, repeatable procedure with pass criteria, run on the pre-production sample with a live animal of the specified measurements, and recorded. We run five steps and we require all five to pass before a size is released to bulk.

Step one is entry. The dog enters the open carrier without being lifted, or is placed with a single smooth motion, and does not resist. Resistance here is usually a floor-height or a lip-height problem rather than a volume problem, and it is fixed by lowering the entry sill, not by enlarging the bag. Step two is the turn: the dog rotates 180 degrees inside the closed carrier within 15 seconds. This is the test that catches the broad-chested misfit, and it is the single most informative step in the protocol.

Step three is the down. The dog moves from standing to sternal recumbency — lying on its chest with its legs folded — and holds that position for 60 seconds without the side panels pressing into the ribcage. Step four is the rise: with the carrier closed, the handler lifts it to carrying height and the dog's centre of mass stays within the footprint of the floor panel. A dog that slides to one end on lifting is in a bag whose floor is too long for its body, which is a real and frequently missed failure.

Step five is the settle: three minutes in the closed carrier on the floor, with the handler out of sight, and no persistent panting, scratching at the closure or vocalisation. This is the welfare gate rather than a dimensional one, and the criteria we use are consistent with guidance published by the American Veterinary Medical Association. A dog that passes steps one to four and fails step five is telling you something about ventilation, sightlines or thermal load, and the answer is never to make the bag bigger.

Recording and reusing fit data

Every test is recorded on a one-page sheet: the three measurements, the morphology classification, the pass or fail on each of the five steps, and one photograph per step. Over a programme that produces a library of maybe thirty animals, and that library is what lets you answer the next customer's sizing question in minutes instead of guessing. It is also the only defensible evidence if a retailer asks how the size ladder was derived.

Where Personalization Changes the Fit Itself

Personalization is usually treated as something applied to a finished product, and mostly it is. But three personalization decisions change the interior geometry, and they need to be made before the pattern is cut rather than after, because they cannot be retrofitted.

The first is a removable liner. A personalized bag frequently ships with a liner carrying the name, and a liner consumes interior volume: a quilted liner with a 8 millimetre loft removes 16 millimetres from both width and length. That is more than the tolerance on most ladders, and a bag that fits without its liner will not fit with it. The rule is to dimension the interior with the liner fitted, and to publish the bare interior separately if the liner is optional.

The second is a name panel placed inside the animal's space. An embroidered panel with a bonded cover adds 2 to 3 millimetres where it sits, which is negligible on a flat wall and not negligible at a corner radius, where three panels meet and the cumulative build-up narrows the usable floor. We keep interior name panels on flat wall sections, at least 40 millimetres from any corner seam, and we mark the exclusion zone on the pattern.

The third is the divider. A personalized programme for two small animals — a name on each side of a divider — halves the floor and changes the shape from one wide rectangle to two narrow ones, and two narrow rectangles do not fit the same animals as one wide one, because the turn test is performed in half the diagonal. If a divided configuration is part of the brief, the divider position has to be tested as a separate size rather than assumed.

Interior Geometry: Floor, Height, Headroom

Three interior dimensions do all the work, and they are not the exterior dimensions on the specification sheet. The gap between the two is the padding, the board and the liner, and it is typically 30 to 50 millimetres per side on a structured carrier and 10 to 20 on a soft one. Publishing exterior dimensions without publishing interior ones is the most common source of a sizing complaint that is not the manufacturer's fault and is still the brand's problem.

The floor is the dimension that governs comfort and it is the one most often short-changed, because a longer floor means a bigger bag and a bigger bag means a higher freight cost. But the floor is where the dog lives for the whole journey, and a floor that is 4 centimetres short forces a curled posture that a dog will hold for twenty minutes and then fight. We do not compress floor length to hit a freight bracket.

Interior height is the dimension that governs the closure, and the clearance over the shoulder — the gap between the top of the scapula and the underside of the closed top panel — is the number that decides whether the zip closes without pressure. Three to six centimetres is the working range, and less than three is a product that will be returned by anyone whose dog is at the top of the band. Headroom above the shoulder beyond that range is wasted volume: it adds height, adds freight cube, and does nothing for the animal.

Finally, the floor has to be flat under load. A dog standing on a floor panel that deflects 15 millimetres in the middle is standing in a hammock, and it will not settle. The deflection criterion we hold is 8 millimetres maximum under the rated load, tested on a supported frame with the load applied over a 200 millimetre disc rather than distributed, because a dog's weight is carried on four points.

Fit Failure Modes and the Return Data Behind Them

Fit failures cluster into a small number of modes, and each has a distinct signature in the return reason and a distinct fix. Reading the return data by mode rather than by size is the fastest way to improve a ladder.

"Too small" is almost always girth rather than length, and it presents as the dog fitting lengthwise and failing the turn. The fix is 3 to 4 centimetres of floor width and a gusset that permits the side walls to bow, not a move up a size, which would add length the animal does not need and freight cost the brand does not want. "Cannot close" is height, and it presents as a correct floor with a shoulder pressed against the zip; the fix is clearance, and if a size cannot deliver it, the closure geometry has to change.

"Dog will not stay in" is the mode most often misread as behavioural. It is usually thermal or visual: a carrier with a dark interior and a single mesh panel at one end runs 4 to 6 degrees hotter at the animal's level than the same bag with cross-ventilation, and an animal that cannot see out will not settle. The dimensional fix — a bigger bag — makes this worse. The correct intervention is ventilation area and sightline, and the test protocol in ventilation testing quantifies both.

"Slides around" is a floor-length overshoot, and it presents on a long-backed animal in a bag sized by weight. It is fixed by reducing floor length and adding a shaped bolster, not by adding padding everywhere. "Chewed out" is not a fit failure at all and should not be treated as one; it is an escape-behaviour problem and belongs to a different specification conversation entirely. under-seat dimensional limits impose their own constraints on the travel end of a range and should be read alongside this guide.

The commercial point is that three of these five modes are fixed by a dimensional change costing nothing per unit, and two are fixed by decisions that reduce cost. None of them is fixed by moving up a size, which is the default response and the most expensive one.

Turning Fit Data Into a Size Ladder

A size ladder is a commercial instrument before it is a technical one. It has to be small enough to hold in stock and large enough that no customer is pushed to a competitor, and those two pressures resolve differently depending on channel.

For a direct-to-consumer programme, four sizes is the working answer: XS, S, M and L, covering girth envelopes from under 44 to 80 centimetres. Four sizes is the most a customer can reliably self-select from a chart, and self-selection accuracy is the metric that matters, because a customer who chooses correctly does not return. Adding a fifth size improves coverage by a few percentage points and reduces self-selection accuracy by more than that.

For a retail programme, three sizes is usually right, because shelf space is the binding constraint and a retailer will take three facings of a bag before it takes five of the same bag in different sizes. The three should be S, M and L, with the ladder shifted so that the median animal sits at the bottom of M rather than the middle — retailers sell the middle, and the middle should be the size most customers need.

Every ladder needs an overlap and a stated tie-break. Adjacent sizes should overlap by roughly 4 centimetres of girth, and the chart must say what to do in the overlap: for a dog between sizes, choose by back length if the animal is long-backed and by girth if it is broad-chested. Without that sentence, a customer in the overlap guesses, and half of them guess wrong. The published chart should carry interior dimensions, not exterior, and should carry the weight range last and labelled as guidance rather than as a rule. Our companion note on weight classes inside a range covers the structural side of the same decision.