A sizing chart is not a courtesy page at the bottom of a listing. It is the last instruction the customer reads before choosing a size, and the first document they blame when the product does not fit. Most charts fail because the customer cannot map their dog onto them, not because the measurements are wrong.
The short answer
Build the chart around one measurement the customer can take without help, show exactly where the tape goes, and make the size steps overlap so the dog between two sizes has a stated answer. Measure the approved sample, not the design drawing, and confirm the break-out for your market on the spec sheet. Samples are US$25–40 per model and take 5 days.
Most returns are a sizing decision
The customer does not return a product because it is badly made; they return it because it does not fit the dog in front of them. That decision is made on the chart, often on a phone. If the chart asks for a measurement the customer cannot take, or uses breed names as a shortcut, they guess — and a guess is a return with postage attached.
For the brand, the cost appears twice: reverse logistics and the review. That is the commercial case for treating the chart as part of the product rather than as page copy.
Pick the measurement the customer can actually take
| Product | Primary measurement | Where the tape goes | Why not breed |
|---|---|---|---|
| Belly band | Weight range | A household scale, dog standing | Breed names hide very different body shapes at the same weight. |
| Bed or mat | Body length plus sleeping posture | Collar to tail base, noting whether the dog curls or stretches | A bed sized to the floor is sized to the wrong thing. |
| Harness | Chest girth | Widest point behind the front legs, tape snug | Girth varies more than breed suggests. |
| Pad | The dog’s footprint | Where the dog actually lies, not the crate floor | Crate size and pad size are different decisions. |
For a belly band the measurement is weight, because a scale is in every home and the band is sized to the dog’s middle, not its breed. The logic is set out in belly band sizing and absorbency; the chart’s job is to repeat it in one line. For bedding, the useful measurement is the space the dog takes up lying down, which is why large on its own tells the customer nothing.
Overlaps beat gaps
Two ranges can cover the same market with the same number of sizes and produce different return rates because of where the boundaries fall. A hard boundary between steps sends the borderline dog to one side with no guidance; an overlap gives the customer a rule — if the dog sits between steps, take the larger size in a wrap product and the longer size in a bed.
The number of steps is a stock decision as much as a fit decision. More steps mean a better fit but more SKUs to hold, and every SKU carries its own order quantity. Fewer steps are cheaper to stock but push more dogs into the wrong side of a boundary. The break-out should follow your market’s size distribution, not an equal split.
The drawing does more work than the table
A table assumes the customer knows where to put the tape. A drawing removes the assumption: a dog in profile, the tape marked at both ends, an arrow at the widest point, and a caption that says measure here, while the dog stands. Two small pictures — the measurement and the dog in the product at each size — prevent more returns than a paragraph of instructions.
Put the instruction above the table, not below it, and repeat the key line in the product page’s size selector. Use both units if you sell across markets, state the tolerance, and answer the between-sizes question where the customer is choosing, not in a FAQ nobody opens.
Test the chart before you print it
Give the draft chart to a few people who have never seen the product. Ask them to pick a size and say why. Where they hesitate is where the chart fails, usually in the same place for everyone: an ambiguous starting point, a missing between-sizes rule, or a unit the customer does not use.
Then measure the approved sample. Production dimensions, not the drawing, are what the chart must describe, and shrinkage after the claimed wash belongs in the same check. After launch, read the return reasons by size — a single step that keeps coming back is a chart problem before it is a product problem, and changing the chart is far cheaper than changing the pattern.
Questions we get asked
Should the chart use weight or body measurements?
Use the measurement the product is built around. A belly band is sized by weight, a bed by the space the dog takes up, a harness by chest girth. Whichever you choose, show how to take it and repeat it at the point of selection.
How many sizes should a first range carry?
Enough steps to cover your market’s dog sizes with overlaps, and no more than you can stock. Each model carries its own order quantity, so extra steps are a stock commitment. The break-out is confirmed for your market, not copied from another chart.
Why do customers still choose the wrong size when the chart is accurate?
Accuracy is not the problem; usability is. The customer may not know where the tape goes, may measure a standing dog for a product used lying down, or may hesitate over rounding. Add the instruction, the drawing and the between-sizes rule.
Can I change the chart without changing the product?
Yes, and you should when the return data says so. But the chart has to describe what ships: measure the approved sample, including after the claimed wash, and if the dimensions themselves change, treat it as a new sample.
Send the product, the market and the sizes you have in mind, and we will tell you which measurement to lead with, where the steps should overlap and what to measure before launch. Start from the collection, or reach us through the contact page — one business day for a reply.