Quick answer: Amazon changed review sharing across variation families starting February 12, 2026, rolling out gradually through May 31, 2026. Reviews are intended to be shared when child products differ only in minor ways, certain size, color, or model-fit differences, that don't change the core product experience. Variations with meaningful differences in function, formulation, or customer experience may show child-specific review counts even while remaining under one valid parent.
A skincare brand sold one moisturizer in six scents, with formula, packaging shape, and size almost identical across all of them. The parent detail page had accumulated 4,800 reviews. One morning, the lavender child showed 612 reviews, the unscented child showed 1,940, and two other scents showed fewer than 300 each. The parent-child relationship hadn't visibly broken, customers could still switch between variations. The instinct was to rebuild the family or open repeated "reviews not merged" cases. That could have made the problem worse. A family can remain a valid variation family while Amazon changes how review evidence is displayed across its children, variation structure and review sharing are related but not the same system.
What changed in 2026
Amazon announced that starting February 12, 2026, it would change how reviews are shared across product variations, rolling out gradually by product category through May 31, 2026, with 30-day email notification before changes affected specific products. The broad principle: minor differences that don't materially change the product experience may share reviews, while significant differences affecting function, formulation, or customer experience may not. Reviews continue to be shared for color or pattern variations, secondary scent variations for non-scent-focused products, and different model fitments for the same product type. The difficult part is deciding what Amazon considers minor for a particular product type.
Variation eligibility and review eligibility are different questions
A variation family asks whether these children are legitimately related under an allowed variation theme. Review sharing asks whether feedback about one child is sufficiently relevant to a shopper considering another child. A family can pass the first test and fail the second: different flavors may be a valid customer-selection relationship in a category, yet a review saying the strawberry flavor tastes artificial tells a buyer very little about the vanilla child. Amazon may keep the children together while separating the feedback.
Composite case study: size was minor until the use case changed
An outdoor brand sold the same foldable dog ramp in three lengths, 90 cm for small cars, 120 cm for SUVs, and 180 cm for vans and steep loading heights, all with the same materials and basic design. The seller assumed length was a minor size variation. Customer experience said otherwise: reviews on the longest version discussed weight, flex, and storage difficulty, while reviews on the shortest version discussed whether the incline was too steep for older dogs. When reviews separated, the seller initially argued that "size" should always share reviews, the wrong argument. The better question was whether the size difference affected function. The revised strategy improved child-level content instead: vehicle-height guidance, weight and folded-size comparison, and a comparison table showing the tradeoffs. The review split reduced the headline count, but conversion improved on the long model because shoppers finally understood what it was for.
The minor-difference test
Five questions help classify a variation. Does the child perform the same core job, a blue bottle and a red bottle likely do, a shampoo and conditioner don't. Would a defect in one child likely apply to the others, if a zipper fails on every color of the same bag, shared feedback helps, but if one flavor tastes bad it doesn't imply another flavor does. Does the variation change ingredients or formulation, a strong sign reviews may need to stay child-specific. Does the variation change fit only or function too, a phone case for two compatible models may stay functionally similar, while a "small" and "extra large" product can behave very differently even under a size theme. Would a shopper feel misled if a review referred to another child, the most useful human test of the five.
Why reviews may not share even when the family appears compliant
The rollout or catalog processing may simply be incomplete, with review counts updating at different times during a gradual change. Product type or category data may not align correctly, since browse node and product type are separate data points that can conflict. A family may use a deprecated or inappropriate theme, since Amazon removed several variation themes during 2025, leaving old families visible but prone to failure when edited. A previous abusive relationship can create a "sticky" block where Amazon prevents a cleaved child from being reattached without internal review, and repeatedly rebuilding the family can deepen that problem rather than solve it.
Do not rebuild the family first
Before deleting a parent or uploading a new variation file, capture the parent ASIN, all child ASINs and SKUs, product type for each child, variation theme and attributes, current review count and rating by child, and screenshots of the customer-facing selector. Deleting and recreating a parent can temporarily split relationships, create errors, and make it harder to prove what changed, exactly the wrong move if the real cause is a policy-based review split rather than a broken structure.
A safe variation audit
Validate the commercial relationship: would customers reasonably compare these children on one page? Validate the catalog relationship: confirm same brand, compatible product type, and complete variation attributes. Validate review relevance separately, distinguishing "valid variation" from "eligible to share reviews." Check whether the issue is structural (children detached, wrong parent, wrong theme, catalog error codes) or presentational (family remains intact but reviews display separately by design). Then contact support with the correct request, asking for review of the parent-child relationship if structure is broken, or asking the Product Reviews team to confirm review-sharing eligibility if structure is valid, not claiming that every valid variation must share reviews.
Composite case study: the seller who made a policy change look like a catalog error
A nutrition brand sold protein powder in chocolate, vanilla, and unflavored versions, and reviews split during the 2026 rollout. The seller deleted the parent, created a new one, changed the variation theme, and uploaded the children again. One child attached to an older family, another triggered a relationship error, and the product pages lost stable indexing for several days. Support later confirmed flavor differences were expected to show more product-specific review evidence under the updated approach. The original issue required a content response. The seller created a catalog crisis because the team assumed "reviews not shared" meant "variation broken."
How to adapt strategy after a review split
A lower visible review count isn't automatically a ranking disaster, it changes where trust must come from. For each child, use child-specific images, add a comparison table, address the most common child-specific objections, and monitor conversion at child level, covered further in Amazon conversion rate by ASIN. Avoid sending all traffic to the child with the largest historic review count. A family with honest, specific evidence can outperform a confusing family with a large blended rating.
EcomSanity can't merge reviews or repair a variation family, but it can help monitor the commercial effect of a review split by showing child-level sales, conversion, and Buy Box movement. A useful analysis compares the four weeks before and after the visible review change while controlling for price, stock, and ad spend. If conversion drops on only one child, the issue is likely child-specific content, not the total review count.
Frequently asked questions
When did Amazon change how reviews share across variations?
Starting February 12, 2026, with a gradual rollout by product category through May 31, 2026. Sellers received email notification 30 days before any change affected their specific products, which is why two similar competitor families could behave differently during the transition period.
Which variation differences still share reviews under the new rules?
Minor differences that don't affect the core product experience: color or pattern variations of the same product, secondary scent variations for non-scent-focused products (lemon-scented versus unscented cleaning products), and different model fitments for the same product type, like phone cases for different phone models.
Should I delete and rebuild my parent listing if reviews split?
Not as a first step. Deleting and recreating a parent can temporarily split relationships, create catalog errors, and make it harder to prove what changed. Capture the current family structure and review counts first, then determine whether the issue is structural (a broken relationship) or presentational (a valid family showing separate review counts by design).