Made in Yangzhou · Shipped from Yiwu — Pet Toy Supply at the Source
Open return carton with pet toys and blank size comparison cards on a table
Home / Blog / Return Reasons

Return Reasons Decoded: Sizing, Aggression and Expectation Gaps

July 1, 2026Marketplace OpsAbout 11 min readTOYORIGIN Sourcing Team

Key Takeaways

  • Most pet toy returns are expectation failures rather than defect failures — which means most are fixable on the page and the spec sheet.
  • Sizing is the quiet leader: dimensions mean nothing to a buyer until translated into weight ranges, jaw sizes and chew styles.
  • 'My dog destroyed it' is three different problems wearing one sentence: wrong size, wrong label or wrong claim.
  • Code every return at the moment of refund — un-coded return data is a cost center; coded data is a product roadmap.
  • Feed findings into the next PO: spec lines, label wording and insert cards cost little and compound across every later order.

A returned toy costs a seller twice. The refund is the visible hit. The second cost is quieter: whatever the listing kept promising while the reason for that return sat unread in the data. This is why returns deserve post-mortems instead of pile-ups — and in pet toys the post-mortems repeat with unusual consistency. Sizing, aggression mismatch and expectation gaps account for the bulk of avoidable returns, and all three are decided upstream, in the size chart, the chew label and the claims on the page.

This piece decodes the categories, shows how to code returns so they become readable, and lists the upstream fixes that belong in your next purchase order rather than your next apology email.

What Actually Comes Back

Ask customer service why toys come back and the answers sort into a short list: too small or too big for the dog; destroyed faster than expected; the dog ignored it; it smells; the owner finds it annoying; or it arrived genuinely defective. Practitioners generally find the expectation-driven categories dominate over true defects, though exact ratios vary by range, price point and channel — treat any single number you see quoted as someone else's assortment, not yours.

The useful observation is that only one of those categories — genuine defect — is a factory conversation in the first instance. The others are translation conversations: between toy dimensions and dog reality, between material spec and marketing adjective, between what play value means to a terrier and to a basset. Returns are the trailing edge of every translation the listing failed to make.

The Size Chart Is a Translation Problem

An owner does not know what a 15-centimeter ball is. They know their dog weighs 25 kilograms, has a jaw like a nutcracker, and dismantled the last "durable" toy in a weekend. A size chart that only restates dimensions is a translation not yet performed. The working version translates in both directions: toy dimensions plus a recommended weight range, a jaw-style note (retriever carry, terrier pry, bully-breed crush), and a chew-style tier.

Return driverPage-side causeProduct-side fix
"Too small" / "too big"Dimensions listed, dog terms missingWeight range and jaw notes on chart and packaging
"Destroyed in days"Vague toughness adjectivesHonest chew-tier label tied to material and wall thickness
"Dog ignored it"Play value implied, never shownPlay video and honest "best for" scenarios on the page
"Smells chemical"Odor never mentionedLow-odor spec and airing note; see our odor control piece
"Arrived damaged"None — genuine QCAQL tightening on the failing node, batch traceability

Scale photography closes the loop the chart opens: a toy beside a shoe, in a jaw, under a tape. We keep a full shot list for exactly this purpose in our guide to pet toy product photography.

Measuring tape, small scale and plush toy arranged on a white desk for size documentation
Size documentation earns its keep twice: first on the listing, later in the returns queue that never happens.

"He Destroyed It" Is Three Different Problems

The durability complaint deserves its own section because it is really three. Problem one: the right toy went to the wrong dog — a gentle-tier plush met a bully-breed jaw, and the label never had a chance. Fix: chew-tier labeling and a size chart that filters before purchase. Problem two: the label said tough but the spec was not — a claim problem, and the dangerous one, because it repeats until the wording changes. Fix: tie every durability adjective to a measurable spec, wall thickness and material, the approach detailed in our chew-strength ratings guide. Problem three: a genuinely weak batch — a QC conversation with the factory, batch records out, retention samples compared.

The three require opposite reflexes, which is why coding matters before responding. A refund issued to problem one with a kinder label on the way is a customer saved; the same refund issued to problem two without a wording change is a customer you will refund again next month.

Code the Returns, Then Read Them Weekly

Return data becomes useful the moment it is coded at the point of refund, not reconstructed later from memory. A minimal code set:

  • SIZE-SMALL / SIZE-BIG — feeds the size chart and any variant decision.
  • DESTROYED — feeds the chew-tier label and the material spec, with jaw size noted.
  • BORED — feeds play-value content and future curation, not the factory.
  • SMELL — feeds the compound and airing conversation with the supplier.
  • DEFECT — feeds QC, with photos attached and the batch code captured.

Read the codes weekly, not quarterly. Patterns need weeks to form and quarters to become expensive; a weekly glance catches them while the fix is still a paragraph on a page rather than a revision on a purchase order.

Return shelf with sorting bins of toys organized into labeled totes in a small warehouse
Coded returns, sorted by reason, are the cheapest product research department a toy brand can operate.

Fixing It Upstream in the Next PO

Coded data earns its keep when it changes the next order. The usual shortlist: split a hero SKU into two honest sizes instead of one optimistic one; rewrite chew labels as tiers tied to material and wall thickness; add an insert card that sets expectations per toy — what it is for, what it is not for; and, where the same soft compound keeps meeting the same heavy jaws, change the material rather than the marketing. That last decision is a materials conversation, and our materials matrix lays out the candidates with their trade-offs.

Close the loop with the supplier in writing: the failure themes, the spec changes agreed, and which inspection line items will verify them. Our quality pages describe how batch testing and retention samples support exactly this handoff. A return code that survives into a spec line is the difference between paying for a failure once and subscribing to it.

Honesty note: the temptation when returns climb is to promise harder — "indestructible," "unbreakable," "puppy-proof." Resist it. Inflated durability claims convert a sizing return into a trust failure, and trust failures leave reviews that outlive the SKU. The durable fix is always a better translation: size charts in dog terms, honest chew tiers, and claims a spec sheet can back within twenty-four hours.

Frequently Asked Questions

FAQ

What is a normal return rate for pet toys?
There is no industry-wide number worth quoting; rates vary widely by channel, price point and how honestly the listing sets expectations. The useful benchmark is your own trend: rate by SKU, rate by return reason, and the direction of both over time. A rising share of destroyed returns is a different problem than a rising share of too small.
Should I refund a toy the dog destroyed?
In most cases yes, and without a fight. Unconditional refunds are cheaper than the reviews an argument produces, and the destroyed toy is data: log it against the SKU and the chew tier. If the same tier keeps failing, the fix is a better label or a tougher spec, not a harder conversation with customers.
Do size variants complicate the SKU count too much?
Less than the returns they prevent. Start with one extra size on hero SKUs only, defined by weight range and chew style rather than a second color. Two honest sizes consistently outperform one optimistic size that serves every dog badly.
How fast should page fixes follow return data?
Page-level fixes belong in the current content cycle: size charts, images and wording can change within days. Labels and insert cards move at the next print run, and spec or material changes move at the next purchase order. Matching each fix to the fastest vehicle that can carry it is the whole discipline.

Returns clustering around sizing or durability?

Send the return themes — we reply with size variants, honest label wording and FOB ranges for the corrected range.