Skip to main content
Card class: Standard • Category: Voice of Customer

​
At a glance

The diagnostic behind the return rate. It takes reason-coded customer returns and clusters them by ASIN into actionable buckets: sizing, quality / defect, and listing mismatch (item not as described). This card exists to replace the Brand-Analytics-gated “top complaints” view with something every seller can act on directly: it tells you which ASIN is returning for which fixable reason, so you can edit a listing, fix a supplier defect, or correct a size chart before the problem compounds into reviews and account health.
What it showsReturns grouped by ASIN, then split into reason clusters. The horizontal bars rank ASINs (or reasons) by return volume so the worst offenders surface first.
Reason clustersCustomer return reasons mapped into a small set of actionable groups, typically sizing / fit, quality / defect, and listing mismatch (not as described), plus a residual “other” (changed mind, found cheaper, no longer needed) that is usually not actionable.
Why clusters, not raw reasonsAmazon’s raw return-reason codes are numerous and noisy. Clustering them turns “47 different reason strings” into “this ASIN has a sizing problem”, which maps to a single fix.
SourceReason-coded returns from Amazon’s customer-returns data. This is genuinely available to all sellers, unlike Brand Analytics “top search terms” or voice-of-customer panels that require Brand Registry.
ActionabilitySizing cluster, fix the size chart / add fit guidance. Quality cluster, raise with your supplier / inspect a batch. Listing-mismatch cluster, correct photos, dimensions, or bullet points. The “other” cluster is noise.
Relationship to Return RateReturn Rate is the headline percentage; this card is the why behind it.
Time window30D (the selected period)
Alert triggernew top-3 reason emerging, a cluster newly entering the top three for an ASIN
Rolesowner, operations, marketing

​
Calculation

Calculated automatically from your Amazon Seller Central data. See the At a glance summary above for what the metric tracks and the worked example below for a typical reading.

​
Worked example

An apparel and accessories seller reviewing the 30D clusters on 01 May 26. Figures are illustrative.
ASINTotal returnsSizing / fitQuality / defectListing mismatchOther
Running tee (top revenue)210150202515
Travel mug950701015
Phone case6058407
Yoga mat4026428
Running tee:  sizing dominates (150 of 210). Fix = revise the size chart / add fit guidance.
Travel mug:   quality/defect dominates (70 of 95). Fix = inspect the batch, raise with supplier.
Phone case:   listing mismatch dominates (40 of 60). Fix = the photos/dimensions oversell the product.
Yoga mat:     "other" dominates (28 of 40). Mostly changed-mind; not much to fix on the listing.
Four things to notice:
  1. Each ASIN has a different, single fixable cause. The running tee is a sizing problem, the travel mug is a quality problem, the phone case is a listing problem. Three different ASINs, three different teams (merchandising, supplier QA, copy). The clustering turns a pile of returns into a clear, assigned to-do list.
  2. Listing mismatch is the cheapest to fix. The phone case is returning because the listing oversells it (photos make it look more premium than it is). That is a same-day copy-and-photo fix with no supplier involvement, and it should be the first action because it is free and fast.
  3. A quality cluster is a supplier conversation, not a listing edit. The travel mug’s defect cluster will not respond to better copy. It needs a batch inspection and a supplier discussion. Editing the listing here would just hide a real defect and invite worse reviews.
  4. The “other” cluster is mostly noise. The yoga mat’s returns are dominated by changed-mind / no-longer-needed, which is not actionable. Do not chase it. The alert fires on actionable clusters (sizing, quality, mismatch) entering an ASIN’s top three, not on the noise bucket.

​
Sibling cards merchants should reference together

This is the diagnostic. Pair it with the rate, the rating, and the health signals:
CardWhy pair it with Return Reason Clusters by ASIN
Return RateThe headline this card explains. The rate says returns are up; the clusters say which ASIN and why.
Star Rating Drift (top-50 revenue)Same root cause, different surface. A quality cluster on an ASIN usually shows up as falling stars on that ASIN too.
Review Velocity (30d)A quality-driven return surge often precedes a wave of negative reviews. Catch it in returns before it hits reviews.
Order Defect RateQuality / defect-coded returns are the ones that threaten ODR and account health.
Negative Feedback (30d)Listing-mismatch returns and negative feedback share the “not as described” theme.
A+ Content Coverage (top-50 revenue)Better A+ content (clear sizing, accurate imagery) is a direct fix for sizing and listing-mismatch clusters.

​
Reconciling against Amazon Seller Central

Where to look in Seller Central: The closest Amazon-native views are:
Reports → Fulfilment → Customer Returns for FBA reason-coded returns per ASIN, and Voice of the Customer dashboard (Performance) for the per-ASIN customer-experience health and the verbatim return comments where available.
Amazon’s Voice of the Customer dashboard shows per-ASIN CX health (Excellent / Good / Poor / Very Poor) with the top return and complaint reasons. This card reorganises the same reason data into action clusters and ranks by volume, which is faster to triage across a large catalogue. Timing, settlement, and reporting-lag table:
TopicDetail
Reason captureThe reason is the customer-selected code at return initiation. It reflects what the customer said, which may differ from the true cause (a “defective” code can sometimes be buyer’s remorse, and vice versa).
Clustering mappingRaw reason codes are mapped into the action clusters. Where Amazon adds or renames reason codes, the mapping is maintained, but a brand-new code may briefly land in “other” until mapped.
FBA vs FBMFBA returns carry structured reason codes promptly. FBM returns depend on how you log the reason; free-text or missing reasons reduce cluster precision for FBM.
Comment availabilityFree-text return comments are not always present. Clusters built on reason codes are robust; comment-level detail is supplementary.
Why our number may legitimately differ from Seller Central:
ReasonDirectionWhy
Cluster vs raw reasonDifferent shapeAmazon shows individual reason codes; the card groups them. A seller comparing one raw code to a cluster will not see a one-to-one match.
Voice of Customer scopeDifferent populationThe VOC dashboard blends returns with other CX signals (negative feedback, A-to-z, chargebacks). This card is returns-reason only.
FBM reason qualityOurs coarser for FBMIf FBM return reasons are logged inconsistently, FBM ASINs cluster less precisely than FBA ASINs.
Window alignmentEdge effectsA return’s reason is dated to the return event; sales-side dating differs, so per-ASIN counts can disagree at the window edge.
Cross-connector reconciliation against other connectors the same seller may run:
CardExpected relationshipWhat causes legitimate divergence
amazon.return_rateThis card decomposes that one. Summing cluster volumes per ASIN reconciles to the returns feeding the rate.Goodwill refunds with no physical return appear in some refund views but not in reason clusters.
shopify return cardsSame product, different reason mix. A DTC store with its own return form may capture richer reasons than Amazon’s fixed codes.Different return policies and audiences shift the cluster mix even for the identical product.

​
Known limitations / merchant FAQs

Why clusters instead of Amazon’s raw return reasons? Amazon’s raw reason codes are numerous and noisy, and across a big catalogue you cannot triage them all. Clustering maps them into a handful of actionable buckets (sizing, quality, listing mismatch) so each ASIN’s dominant problem maps to a single fix and a single owner. Doesn’t Brand Analytics already give me this? The Brand Analytics “top complaints” and search-term reports require Brand Registry and are gated. This card is built on customer-returns reason data that every seller can access, which is why it works whether or not you are brand-registered. How accurate is the reason coding? It reflects what the customer selected, which is usually right but not always. A “defective” code can occasionally be buyer’s remorse and the reverse happens too. Treat a dominant cluster as a strong signal to investigate, not as proof. Cross-check quality clusters against Star Rating Drift (top-50 revenue). What do I do with each cluster? Sizing, fix the size chart and add fit guidance. Quality / defect, inspect a batch and raise it with your supplier. Listing mismatch, correct the photos, dimensions, and bullets so the listing stops overselling. The “other” cluster (changed mind, found cheaper) is usually noise, do not chase it. Why did an ASIN suddenly enter the alert? The alert fires when a new actionable reason breaks into an ASIN’s top three. That usually means something changed: a new supplier batch (quality), a listing edit that overstated the product (mismatch), or a size run that does not match the chart (sizing). The change point is the clue. Does this work for FBM? Yes, but the quality depends on how cleanly FBM return reasons are logged. FBA returns carry structured codes automatically; FBM clusters are only as good as your logging discipline.

​
Tracked live in Vortex IQ Nerve Centre

Return Reason Clusters by ASIN is one of hundreds of KPI pulses Vortex IQ tracks across Amazon Seller Central and 70+ other ecommerce connectors. Nerve Centre runs the detection layer; Vortex Mind investigates the cause when something moves; Ask Viq lets you interrogate any number in plain English. Start for free or book a demo to see this metric running on your own data.